The Basement: Donald Hoffman | Consciousness, Math, and the Case for Letting Go
AJ Gentile's two and a half hour Basement interview with cognitive scientist Donald Hoffman, rebuilt in full. The preacher's son who coded fighter jet displays in machine code at Hughes, studied under a dying David Marr at MIT, and sat in Francis Crick's secret Helmholtz Club lays out why physicalism has explained zero conscious experiences, the fitness beats truth theorem and the jewel beetle, observers as Markov chains, the trace logic Chetan Prakash proved at Heathrow, and how missing a yellow light becomes time dilation. Then DMT avatars, UAPs as Grand Theft Auto coders, and a story he had never told publicly: two figures at his hospital bed. Seven custom visuals, every reference linked.
Published Sep 14, 20262:27:48 video116 min readAdded Sep 16, 2026Open on YouTube →
At a glance
For two and a half hours in The Basement, the long form interview room of The Why Files, host AJ Gentile sits across from Donald Hoffman, the UC Irvine cognitive scientist whose TED talk told millions of people that we have never once seen reality as it is. The first act is the man: a fundamentalist preacher's son raised on a 4,000 year old earth, a seventeen year old who decided to find out whether we are just machines, a UCLA undergraduate hand coding a complete fighter jet flight simulator into 64K of machine code at Hughes Aircraft, the last student cohort of a dying David Marr at MIT, and the young professor whom Francis Crick pulled into a private consciousness club that met a five minute walk from his house. The second act is the argument: forty years of brilliant physicalist neuroscience that has not explained a single conscious experience, a theorem in evolutionary game theory that puts the probability our senses were shaped to show us the truth at exactly zero, and a simple reason there can never be a theory of everything. The third act is the new mathematics: observers as Markov chains, agency as a recursion of Markov chains over Markov chains, a "trace" relation that his longtime collaborator Chetan Prakash proved yields a logic on all Markov chains while sitting at Heathrow, and a counter on every chain that Hoffman says becomes Einstein's time dilation when one observer cannot see all of another's states. Then it goes where The Basement always goes, into DMT entities as avatars, a mathematician's rotating tesseract, UAPs as the programmer who wrote Grand Theft Auto, and a story Hoffman says he had never told in public: the night his heart ran at 190 beats a minute for 30 hours, the goodbye text he sent his wife, and the two men standing beside his hospital bed at three in the morning.
age 5 Walking to kindergarten, he finds a bush covered in butterflies at eye level, stops to enjoy it, brings one to class, and is punished for being late.
childhood Sunday is a fundamentalist church and a 4,000 year old earth, taught by a father who holds a master's degree in chemistry. Monday is school, and billions of years.
12 or 13 After an Ice Capades outing the family takes up figure skating. He falls for weeks, then it clicks, and he works up to a double toe loop and an axel.
17 He picks the question of his life: are we just machines?
1976 to 1978 Undergraduate at UCLA while writing fighter jet cockpit display software in raw machine code at Hughes Aircraft in El Segundo. As a senior he reads a Marr paper in an AI class and is electrified.
1978 to 1979 A year off before graduate school, full time at Hughes.
1979 to 1983 MIT, on a Hughes funded free ride: the AI lab, a seminar in Marvin Minsky's home, and about eighteen months with David Marr before leukemia kills him at 35.
1983 to 1984 He turns down directing the Hughes AI lab in Malibu for a UC Irvine job at a quarter or a fifth of the pay. Francis Crick invites him to the Salk Institute, then into the Helmholtz Club.
about 1984 Mathematician Chetan Prakash starts working with him. Forty two years later, he still is.
1986 or 1987 The mathematics "slaps him in the face": we are creating everything we see, like a VR headset.
1989Observer Mechanics, with Bruce Bennett and Prakash. John Wheeler cites it in his "It from Bit" paper. Then roughly twenty years in the wilderness.
1992 At a Helmholtz Club meeting Crick tells him about the book he is writing, The Astonishing Hypothesis.
about 2 years ago Hoffman realizes the trace relation is a partial order and a logic on all Markov chains. Prakash says it is too pretty to be true, then proves it at Heathrow.
5 months ago What he calls a big step in the model, after forty years at it.
now, at 70 Tentative proofs of special and general relativity from Markov chains sit with his mathematicians. The Trace Institute's goal: prove nine conjectures in two or three years.
Figure 1. The path as Hoffman tells it in this conversation, from the butterfly bush to the Trace Institute. The dates are his; the relative ones ("two years ago", "five months ago") are as of the September 2026 recording.
The cold open
AJ lays out the guest in a few lines. Don is a cognitive scientist who spent decades at the heart of mainstream science. He wrote fighter jet software for Hughes Aircraft in pure machine code. He trained at MIT. He sat for years in a private consciousness club that Francis Crick cofounded. That last one triggers Hecklefish, the show's resident talking fish: "Francis Crick, the guy who discovered the building blocks of life. I discovered you can't return underwear at Target. Both took real courage."
AJ carries on. "Then his own math convinced him that we have never once seen reality as it is. He seems very calm about it." The episode will cover the beetle that fell in love with a beer bottle and why that matters for evolution (Hecklefish, dryly: "A beetle dated a beer bottle."), why Don says space and time are a headset we are wearing, what that means for UAPs and for the entities people meet on DMT, and, near the end, a story Don has never told before. "It's about the night he texted his wife goodbye from a hospital bed. You might know that part, but you don't know the rest." For once the fish has nothing to add: "Okay, no notes on that one. That one's real."
AJ promises to come back after the interview and break the conversation down, "which is not going to be easy, but I'll be here. Let's go down to the basement."
Ice Capades, not hockey
Hoffman grew up in San Antonio and spent most of his life in Southern California, so AJ opens with the item on his résumé that makes no sense: ice skating.
It started when he was about 12 or 13. The family went to an Ice Capades show or an Ice Capades Chalet rink, his parents liked it, and they decided the kids should try. He was 12 or 13, his brother a year younger, his sister four years younger. He hated it at first. "You just fall down. Everything that you try to do is just wrong. All your normal reactions are wrong." He fell and fell and fell, and did not want to go back, and his parents made them keep at it for several weeks.
Then one day it clicked. "You realize you don't walk like you normally walk. Heel toe, heel toe. You push with a side and you glide." When you really discover that for yourself, he says, it opens up a whole new world, and once you are in that world gliding is fun. He eventually landed a double toe loop and an axel, the jump with the extra half turn. He still goes back to the ice now. Not for double jumps ("I'm 70, so it's not smart"), but carefully stroking and gliding is very good exercise.
It was figure skating for the whole family. He has never actually been on hockey skates. AJ says he is picturing Hoffman in the full Lycra outfit, and that he looks great, by the way.
A preacher's son and a 4,000 year old earth
His father was a fundamentalist preacher. So AJ asks the obvious question: what would that man make of his son starting an institute premised on nothing being real?
In his later years, Hoffman says, his father did hear about the work, and they talked about consciousness being fundamental. His father liked that part, because it lined up with his own views. "He thinks he knows what that consciousness is and it's his god and not other people's gods." He liked the non physicalist direction and he would have liked that there was a mathematical model. What he wanted, though, was for it to prove that his particular denomination of Christianity was the truth and all the others were not. It does not show anything like that. What Hoffman thinks it shows is a much deeper level of consciousness, of which the religions, Christian, Buddhist, Hindu and the rest, are all perspectives. Each gets a piece of the puzzle and each misses pieces, as any perspective would.
AJ wonders whether that is what planted the seed: one story on Sunday and a different one on Monday. "That's exactly right."
The Sunday story was severe. The earth was 4,000 years old. And his father believed it while holding a master's degree in chemistry, not just a bachelor's, and working at high levels in industry using that chemistry. "He had decided to choose the religious view over what chemistry had shown." Evolution was simply off the table, anathema. The Monday version was all the science saying the earth is billions of years old and that we evolved.
AJ cannot square a chemistry master's with that and asks what drew him to faith over the science. Hoffman finds it hard to explain; it seems irrational to him. The experiments are very clear, and he chooses his words with care: "If you're going to use the language of space and time and chemistry in that framework, the earth is four billion years old," not 4,000. So he had to decide between the two views on his own.
AJ: "So Catholics don't go to heaven then?" Hoffman: a lot of people, even in other Christian denominations, might not be going to heaven by those lights. It was austere, my way or the highway. And the worst of it was the fear. "They would talk about God being love, but in fact you were trained to be afraid. You had to be really, really afraid and cry every Sunday and repent." He calls it real psychological control. When you are raised in it, it is all you know, and it has taken him decades to recover.
Are we just machines?
Around 17, he chose a specific question to answer with his life. "Are we just machines? That's my question." To answer it he would have to understand what machines can do, and whether there is anything humans can do that machines cannot.
That took him to MIT in 1979, into the artificial intelligence lab, which is the "what can machines do" half. He had the luck of taking a class with Marvin Minsky, who founded the field together with John McCarthy. The class met for a semester in Minsky's home, where the students could argue with him about the philosophical foundations of artificial intelligence.
From 1979 to 1983 he was in the AI lab. You cannot just talk about AI in general, he says; as a scientist you pick a problem and try to solve it. He picked "how do we see in 3D?" Machine vision: how to build visual systems that see in three dimensions, the kind of thing self driving cars now do. "We were pioneering it back then." At the same time he was in what is now the Department of Brain and Cognitive Sciences, studying human neuroscience, what humans can do, and trying to piece the two halves together.
He kept building mathematical models of vision after MIT, and it was around 1986 or 1987 at UC Irvine, working with two "extremely talented mathematicians," Bruce Bennett and Chetan Prakash, that the story turns. (The auto captions render the second name "Jayton Pash" and later "Chayon"; it is Prakash throughout.) He calls them both geniuses, and he is still working with Chetan.
Coding fighter jets at Hughes Aircraft
AJ stops him there. While all this was going on, wasn't he working at Hughes Aircraft on vision systems for missiles?
He was. As an undergraduate at UCLA from 1976 through 1978, and then for a whole year from 1978 to 1979, he worked at Hughes, the last stretch full time. The Hughes people liked his computational vision work so much that they paid his entire way through MIT. "I got a free ride through MIT from Hughes." There were no strings attached, but everyone assumed he would come back and become director of the Hughes Aircraft artificial intelligence laboratory in Malibu. He assumed it too. "I was about to be a very rich guy living in a very nice place."
He worked out of El Segundo, on fighter jet cockpit software for displays. It was the moment the old mechanical cockpit gauges were going digital, but the microprocessors were not fast, so the displays had to be programmed in machine code. He and two or three other people were the coders.
AJ asks whether they did not even have Fortran. Hoffman had learned Fortran as a sophomore, but these were special purpose, interrupt driven processors, brand new in 1976, and all they came with was machine code. He names the machine as something the captions spell "anuk 30, an k 30" ("you can look that up"); it reads like a designation in the military AN/UYK computer series, but we could not find public documentation for that exact model, so take the name as heard. What he remembers is the craft. "I knew all the ones and zeros. If I wanted to multiply by two, I didn't multiply by two. I shifted." Divide by four, shift right twice. You found all the tricks.
With those tricks the team wrote an entire, complete flight simulator in machine code, programmed by hand bit by bit, and fit it into 64K. AJ: "You can't type one word in an email for 64K." Hoffman: "We knew all the bits." It was delivered to Wright Patterson Air Force Base in 1978.
Because he was one of perhaps three or four people in the world who knew how to do that kind of work for the new electronic cockpit displays, the undergraduate was flown around to military installations and contractors to help. "I was a cold warrior as an undergraduate," and for that full year in 1979 as well. Then he went to MIT, on Hughes money, with the directorship waiting.
Money or freedom
Toward the end of MIT he realized he was facing a choice between money and doing the research he wanted. Even as director of the Hughes lab there would be directives from above, and he would not be free to explore whatever he wanted, only whatever might lead to a product. He is careful to say this is not right or wrong; some people lean one way, some the other. He wanted complete autonomy.
So he took a pay cut, a large one. As a new professor at the University of California, Irvine he made less than he had made as an undergraduate at Hughes, "maybe a quarter or a fifth" of what he would have made there. AJ points out he could have gone back at any time. True, Hoffman says, but once the research got going it took on a life of its own, and he knew he wanted to pursue it all the way.
The butterflies that made him late for kindergarten
AJ goes back to "young Don," to a story that might be the seed of everything, including the interest in vision: the day five year old Hoffman was late to kindergarten.
He had learned the route and was confident he could get there. On the way he passed a bush in bloom, right at his eye level, covered in butterflies. "It just was obvious to me that kindergarten was nowhere near as important as enjoying life. I mean, this is a miracle right in front of me." The right thing to do was obviously to relax and play, to enjoy and explore, because nature was showing him something amazing. "You're here to enjoy, to observe, and to learn." He is careful to add that he is putting adult words on it. At five it was not an intellectual position, just an emotional, childish certainty that this was a whole thing to explore.
AJ draws the line forward: leaving Hughes was the grown up version of stopping at the bush. Hoffman agrees. Life is about exploring, going where your heart wants you to, and having fun. "It's really a play." And he means that about reality itself. "We're here. It's not that serious." He reaches for the Hindu idea of Leela, that the world is a game, and calls it a really deep insight. He thinks the greatest insights, technological ones included, come when you just play, and that reality somehow rewards it. "Only when you let go of your uptightness do you start to relax into new dimensions of reality."
The five year old butterfly chaser, though, "got strangled." The kindergarten teacher, instead of rewarding it, told his parents he had been late, and he was punished. AJ asks whether he told the teacher why. Of course he did. He brought a butterfly in to share with everybody, not understanding that this was not good for the butterfly. He meant no harm. It was simply obvious that something this joyful should be shared with all his friends, and finding out it was not accepted was a slap down. AJ: "That's how you create a heretic right there that day." Eventually, Hoffman says.
Unforming the formative
He was shaped into the mold until he could leave home at 17 or 18, and those are the formative years. "Unforming the formative" takes a long time, and he says he is still decompressing. In that decompression he has come to see that he was trained to be afraid and to go by the rules, and he now thinks true intelligence is play. "True intelligence is not assuming. It's assuming that I don't know anything." Everything of interest is still ahead of him. As much as he knows right now is trivial. "Everything I've learned is 0%," to be dropped the moment he gets something deeper.
The interesting thing is how deep the emotional programming goes. In meditation he sees himself having to face the old program: follow the rules, God is mad, buckle under. AJ asks if he can observe that while meditating. He can. He sees the pain of it and the trapped energy. What that programming does is close you up to reality, tighten you. As he opens up, the tied up energy flows out, and the old tied up person is afraid of that. "That's why it's painful. That old person is dying. It's really a death of that old." And it is also the birth of someone ready to play, to relax into life, and to explore intelligently, because only then can you get outside your box.
"The only way to make something new is to let go of the old." That does not make old knowledge bad. Often old knowledge takes you to the frontier. It did its job, and once you are standing at the new frontier you have to let go of it and dive into the unknown. "And so that's fun for me."
Theories are just theories
AJ says his audience has followed Hoffman since the TED talk, and nobody has ever seen him lose his temper in a debate. He has released papers that drew ten rebuttals and answered that maybe the critics are right.
Hoffman thinks most scientists would agree that is simply how science is done. People are human and some get rattled and personal, but dispassionately, most scientists would say theories are just theories and you should not be identified with them. Be as clean and rigorous as you possibly can, gather all the evidence you can, make the strongest possible presentation, and then let it go. He has many brilliant friends who disagree with him, and it never has to be ad hominem, and it never is on his side, because they are brilliant. Even when he thinks they are wrong, they push him in new directions that help him think outside his own box. And sometimes he is the one who is wrong, and they point it out, "and hey, good to drop it as soon as you can."
David Marr changed his life
"Theories are just theories" gives AJ his opening for David Marr, who more or less launched computational neuroscience, died very young, and taught Hoffman at MIT. AJ sums Marr up with two positions: without math behind it a theory is just a theory, and vision is designed to bring you the truth. Hoffman built his career on the first and has used it to knock down the second. How would that argument go between them?
Hoffman starts with what Marr did for him. As a senior at UCLA taking a class on artificial intelligence, he read a vision paper by Marr and, he thinks, Tomaso Poggio ("Tommy Poggio" in the captions' spelling), and he was electrified. "I realized that's it. This is really rigorous." They were doing neuroscience with mathematical models and not waving their hands: build a working system, and if it does not work, you are wrong. "There is no nonsense here. This is all serious stuff. This is how you make progress."
He went looking for the author and found him in what was then MIT's psychology department, now Brain and Cognitive Sciences, and also in the AI lab. He had not even imagined psychology at MIT; he thought of engineering and math. It was a long shot, but he applied and flew out for the interview. He had never been to the East Coast and had no idea how cold it was. It was February, and he brought a light jacket. "One trial learning."
They accepted him. AJ asks what it was like to walk onto that campus. "It's a who's who." He took a class with Noam Chomsky. Jerry Fodor was there; there was one class where Fodor and Chomsky were the instructors, and a figure the captions identify only as "Thomas" whose class he also sat in. The fifty or sixty graduate students around him became a who's who of the field. He took a class with the neuroanatomist Walle Nauta. "I had no idea how lucky I was." At the time, all he knew was that David Marr was there.
His co advisor was Whitman Richards, whom he calls an absolute gem. Richards encouraged him to think outside the box, treated him as an equal, and bounced ideas back and forth without pulling any punches, all of it friendly. "He really taught me how to be a gentleman and yet a researcher that doesn't pull any punches." Marr had gathered an incredible group: Berthold Horn (Hoffman took his class and calls it stunning), Eric Grimson, Ellen Hildreth, John Hollerbach and many more.
Marr's last cohort
When Hoffman arrived, the department already knew Marr was sick. Hoffman had him for maybe a year and a half. Marr had leukemia, and it was tragic to watch him die over those eighteen months or so. He kept teaching. He still came to the research meetings, holding something over his mouth because he was bleeding. AJ: "Oh, this is heartbreaking."
Then Hoffman tells a story he clearly has not forgotten. A graduate student he knew, though not well, and whom he will not name, defended his PhD and killed himself the very next day. It was a shock to all the graduate students: the man had worked so hard, and why the day after? The next day Hoffman was walking through the AI lab toward one of the Lisp machines to do his research when Marr saw him and waved him into his office. Marr knew about the suicide, and said: "Look, if you have any feelings about doing that, life is worth living. Come talk with me first." This was a man at the edge of death himself. Marr died at 35.
"The guy was a complete genius." When you sat in a room with him he was the intellectual leader. He knew the neuroscience (his PhD was a theory of the cerebellum) and he knew the AI, and he commanded the room. Everyone looked up to David. Hoffman was in his last cohort, one of perhaps one or two. AJ imagines what Marr could have done with another thirty years. "Oh, can you imagine?" Hoffman feels very lucky to have known him. Without Marr he would not have gone to MIT at all. He would have gone to UCLA, which is a great school, "but MIT at that point was unique in the world" for AI.
Francis Crick and the secret Helmholtz Club
David died, and a year and a half or two later Hoffman graduated and took the UC Irvine job instead of going back to Hughes.
While Marr was dying, Francis Crick, of Watson and Crick and the Nobel Prize for the structure of DNA (AJ: "I've heard of him." Hoffman: "Everybody's heard of him"), was at the Salk Institute by UC San Diego. Crick knew Marr well and tried to save his life, using all of his very substantial connections to get Marr the latest medical technology. Marr got the best there was, which Hoffman says was clearly the right thing to do, because Marr was inventing modern vision science. "He reinvented the whole field."
Crick knew Hoffman had been Marr's student, and almost as soon as Hoffman arrived at Irvine, Crick invited him down to the Salk to spend time with him and talk. Hoffman did not know then that Crick was interested in consciousness, and in 1983 and 1984 it was not really kosher for a scientist to talk about consciousness. It would be another six or seven years before it was, "because Francis said it was kosher."
AJ asks whether John Wheeler had talked about consciousness yet. Wheeler had talked about observers, Hoffman says, and may have discussed the fundamental nature of observers before, but the famous "It from Bit" paper was 1989, just after.
Then Crick invited him into the Helmholtz Club, named for Hermann von Helmholtz (the club's history is written up here). It was a private group, and it met at UC Irvine because Irvine sits in the middle of Southern California, with USC and UCLA to the north and UC San Diego and the Salk to the south. Hoffman lived on campus, so the meetings were literally a five minute walk from his house. AJ: "You could walk to the club." Everybody else had to drive.
It was secret, "not for any nefarious reasons," but because if anyone knew Crick was on campus, nobody would get any work done; people would want autographs. "He was as big a scientist as they come." They met in private at the University Club, one Tuesday every month at one o'clock, and had lunch together. There were maybe ten or twelve core members. Each member could invite one guest, and as a group they would invite two speakers from anywhere in the world whose work interested Crick and the club, and fly them in. The speakers had lunch with them, and then the club grilled them all afternoon. "It was no holds barred." Speakers could not get through their talks.
AJ asks what Crick was like when he disagreed. Always a gentleman, Hoffman says; he never once saw Crick be impolite. But he did not suffer fools, and he wanted answers. He was an older man by then. AJ: "I don't have much time. Let's get to it." Exactly. And the goal was consciousness. Crick had demystified life with DNA and he wanted to demystify consciousness. He wanted the latest neuroscience because he wanted to find whatever finding would break open the door to consciousness the way the double helix broke open the door to life. "He was on a mission." Polite, a gentleman, but nobody was going to get in his way, and if he had a question it was coming out.
AJ wishes he could have been a fly on that wall. There would be a dozen or fifteen sharp people in the room, and the conversations were unbelievable. For Hoffman the lesson was watching how focused Crick was: a clear goal, looking for any clue anywhere, bringing in people from all over to chase it.
In 1992, at one of the meetings, Crick told him about the book he was writing, The Astonishing Hypothesis. They talked it through. That book is what made it official that serious scientists could talk about consciousness. It was also thoroughly physicalist: somehow neural activity, or some aspect of neuroscience, would turn out to cause conscious experience. AJ notes that it is inverted from Hoffman's view. "Completely inverted from mine," Hoffman says, "but in line with what 99% of my colleagues in neuroscience and computer science would say."
Why the physicalist bet made sense
AJ observes that Crick's direction is a lot easier for a scientist to wrap their mind around than the reverse. Hoffman agrees, and defends the bet on its merits. Physicalism with mathematics had done very well since at least Galileo. Spiritual ideas had been around for thousands of years, but all of our technology came from physicalism with mathematically precise models. Until the spiritual traditions produce a mathematical model of their own, "there's no beef for a scientist to go after," unless the scientist takes those ideas and turns them into math.
So the story went like this, and he stresses he is describing the story, not endorsing it. We got the secret of life physically: A, C, G and T, the letters of DNA, and the mathematics of biology. The bet was that we would get the secret of consciousness the same way, from some neural process at some level of some size of system that causes experience. "That's what we were after, and it was fun for me to see the pros go after it."
Batting zero
"But you know, we never got it." To this day, AJ adds, it is still the hard problem of consciousness. Thirty or forty years of good, hard neuroscience, artificial intelligence, computer science and information theoretic attempts. AJ mentions Penrose and Hameroff. Hoffman knows most of the players, calls them brilliant, and calls them friends and colleagues. And the fact remains: there are several theories out there, trillions of conscious experiences to explain, and no physicalist, neuroscientific, AI or computational theory that can explain even one specific conscious experience. Not the taste of chocolate. AJ offers the smell of mint. Not that either.
To put that in perspective he offers an analogy. Suppose Hoffman, a cognitive scientist, walks into a room of physicists and announces a new theory of particle physics. They smile, and ask the natural question: "So, Don, what specific particle interaction does your theory explain, and how does it do it? Photon electron interactions? What is it?" And he answers that he has a general theory of particle interactions that cannot explain any specific particle interaction. "Would I be taken seriously?" They might be kind. They might pat him on the back and say let's have a beer, then go your way, man. They would not take him seriously. "And that's where we are with neuroscience and physicalist approaches to consciousness and AI." Not a single specific conscious experience.
He made exactly this point, he says, two or three weeks earlier at a meeting in Venice. One person there, a good friend whom he deeply respects and so will not name, replied that there might be one: integrated information theory might have a theory of the experience of space. First, Hoffman says, that makes his point for him: trillions of experiences, and we might have one. But he goes further. The theory his friend meant is the integrated information theory of Giulio Tononi and Christof Koch (the captions render them "Toni and Ko"), and it is very clear about its own standard. It says certain kinds of causal structure give rise to specific conscious experiences, and that the rigorous way to write down a causal structure is to write down a Markov matrix: a square matrix with n columns and n rows, with a probability measure in each row.
AJ, who has done his homework, flags that they will get to the traffic light example later. Sure, Hoffman says, that will be fun.
So IIT's own claim is that for every conscious experience there is a causal structure, and you will know you have it when you write down its Markov matrix. "And when you look at the paper that they published, there is no Markov matrix." (The paper in question appears to be the IIT account of why space feels the way it does.) What Hoffman wants is the matrix. How many rows? A hundred? Then it has a hundred columns, which means ten thousand numbers. "What are those numbers, and why must those numbers be the causal structure that gives rise to the taste of mint," or in this case the perception of space? "There's nothing on the table. And of course not. And there never will be."
Again, he says, these people are not dumb. They are absolutely brilliant, and they are doing good work. Even without solving consciousness they are learning a great deal about neuroscience and its structure, so the work is not wasted. "We're just not learning about consciousness, except what we're learning is it doesn't come out of neurons or physical structures." Geniuses trying their best to get consciousness out of physical or computational structures, and failing, is itself the lesson.
AJ thinks it is better that the scientists disagree, because a truth is not a truth until it is known. Hoffman takes that further. If you are going to argue that consciousness is fundamental, the failure of geniuses on the physicalist side is a real acid test, and those same geniuses will be your harshest critics, which is exactly what you need. "This is not about patting people on the back." If he puts forward a theory in which consciousness is fundamental, the physicists should come after him with everything they have, "and then of course we'll go out and have a beer afterwards." Nothing personal; professionally, no holds barred. Often he learns something. "Oh wow, didn't think about that. I need to go look at that. So it's all good."
Simulation theory is physicalism through the back door
Most of the audience, AJ says, knows Hoffman from the TED talk, which got passed around under titles like "Don Hoffman proves there's a simulation." That is not quite right, is it? Simulation is still physicalism through the back door.
Right, Hoffman says. Nick Bostrom is known for the simulation argument. In Bostrom's picture we are quite likely in a simulation, probably coded by some teenager at a lower level on their own computer, and that teenager and their computer are probably a simulation at an even lower level, and so on down until you hit a bottom.
Hoffman likes part of it: the idea that we are not seeing reality as it is, that space and time are not the final reality. He disagrees on two points.
First, the bottom. The assumption, by Bostrom and almost everyone, is that at the very bottom there is again some kind of physicalist spacetime reality. "I think that that's not going to work. I think that spacetime is doomed," and they will get to why. The bottom is not physical and not a spacetime.
Second, consciousness. To the extent that Bostrom and others think our conscious experiences result from programming at the lower level, then since it is a simulation, the explanation is computational. AJ: so it's not really consciousness, is it? Hoffman declines to say that. He would say only that we have no theory that could explain how it could happen. Computational theories of consciousness have yet to explain a single conscious experience, "and I think it's principled. The failure again is principled." So Bostrom is wrong if he says the bottom is spacetime ("if he doesn't say that, I'd be interested to see what he does say"), and wrong to say computational systems can give rise to conscious experiences.
Illusionism does not get you off the hook
Then there is the escape hatch. Some people, and he names Michael Graziano, say there are no conscious experiences, only the illusion of conscious experiences (illusionism). AJ calls it unfalsifiable. Hoffman is happy to play the game, because making hypotheses is what science does. But if you play it, you owe the same thing. "What specific computational system must be the illusion of the taste of chocolate and could not be the illusion of the taste of mint?" There are no candidates on the table. So the illusion hypothesis has no empirical evidence for it whatsoever, and he predicts it never will. "Just moving from consciousness to the illusion of consciousness doesn't get you off the hook." You still have to put up specific theories and specific experiments for specific cases, and there are none. "So illusionism gets you nowhere on this."
"You have to be hard nosed," he says, and it is not ad hominem. He is the first to say Graziano is a brilliant guy. All these people are. "But this just won't work. But I love the disagreement. That's the only way forward."
Physicalist neuroscience
Simulation theory
Illusionism
Hoffman
Examples he names
Crick's astonishing hypothesis; IIT (Tononi, Koch); Penrose and Hameroff
Nick Bostrom
Michael Graziano
Conscious agents, recursive trace logic
What is fundamental
Physical objects in spacetime
A physical spacetime at the bottom of the stack of simulations
Physical or computational systems; experience is not real
Consciousness. Spacetime is a headset it builds.
Where experience comes from
Some neural process, or a causal structure written as a Markov matrix
The computation at the level below
Nowhere. Only the illusion of experience.
It is the starting assumption: experiences that change
What he agrees with
Good science. It is teaching us a lot about the brain.
Space and time are not the final reality
A legitimate hypothesis to make
His own theory is "just a theory" he expects to replace
Where it fails, in his view
Zero specific experiences explained; IIT has never written down the matrix
Spacetime is doomed, so it cannot be the bottom; computation explains no experience
Zero specific illusions explained, chocolate versus mint
Must derive relativity, quantum field theory, the Born rule and the Big Bang, or it is wrong
What it owes
The taste of mint, from neurons
The same, from code
The illusion of mint, from code
Einstein and quantum physics, from Markov chains
Figure 2. The four positions as Hoffman himself contrasts them in this conversation. His standard is symmetric: each side owes a specific derivation from its own starting point, and he spends the second half of the interview on what his side owes.
Observer Mechanics, and twenty years in the wilderness
AJ turns to history. Hoffman and Chetan wrote Observer Mechanics in 1989, made their bet early, and were more or less out in the wilderness for twenty years. What was it like to have nobody paying attention?
There were three authors, Hoffman corrects: Bruce Bennett, "a genius mathematician," who died in the early 2000s, and Chetan Prakash, also a genius mathematician, still alive, still collaborating, and a good friend. They worked hard through the 1980s, four or five years of it, to put out the book. And when it came out in 1989, John Wheeler cited it in his "It from Bit" paper. "If you look at his paper, one of the citations is to Observer Mechanics." AJ: that's a fun phone call to get. It really is, Hoffman says. Wheeler knew that some cognitive scientists and mathematicians had taken conscious observers as fundamental and were trying to show how to build spacetime from them.
The career cost was real. Coming from MIT to Irvine he had been on the fast track, getting grants left and right, the head of AI research and vision science. Then he started moving in this direction. What was really interesting to him was the moment around 1986 or 1987 when he realized that we are creating everything we see. "You can't walk away from the whole physicalist view in one moment," he says, but there was one moment "where the mathematics all of a sudden just slapped me in the face, and I realized that we're creating all this stuff, like it's a VR headset."
AJ holds that thought for after the break, when they will get into the math.
The jewel beetle and the beer bottle
After the break AJ asks for the dinner party version. Assume everyone has seen the TED talk; give him sixty seconds on the beer bottle beetle.
The TED talk, Hoffman says, took on a very deep belief: that evolution shapes us to see the truth. Darwin's theory says evolution shapes organisms to be fit, and shapes their sensory systems to make them fit. Technically that means it shapes sensory systems to make you good at reproduction; the fitness payoff is how many offspring you have. And most of us assume, informally and even formally (experts included), that sensory systems shaped to be fit are fit because they were shaped to show you the truth. Clearly, it would seem, senses that show you the truth will keep you alive longer than senses that do not. "That's a deep, deep intuition. And it's wrong." Very, very brilliant people believe it, and it is wrong. He will start with an example, then give the principle underneath, and the examples, he notes, are not new.
In the outback of Western Australia there is a jewel beetle. The beetles are dimpled, glossy and brown. The males fly and the females are flightless, so the males fly around looking for a female, and when one looks right, the male alights and mates.
In that same outback there were men drinking beer from bottles Australians call stubbies, and the bottles were also dimpled, glossy and brown, apparently exactly the right brown. When they tossed the empties into the bush, the male jewel beetles flew down onto them and swarmed them, trying to mate. And the remarkable part is that they did not land, notice it was not a female, and leave. They crawled all over the bottles with full body contact, as much contact as they could get, and persisted. AJ: "Until they die." That's right, Hoffman says. They do not give up, and it could drive the species extinct.
What that shows is that the males do not know much about what a real female is. "A female is anything dimpled, glossy and brown. The bigger the better." The room laughs. "And some females might tend to agree that males don't have much insight into it."
The point is what evolution was actually doing when it built that sensory system: giving the beetle a solution good enough for reproduction in its niche. In the niche, dimpled, glossy, brown, bigger is better, was enough. The male did not need to know anything more about a female because nothing was going to fool him; there were no stubbies. It is what's called a satisficing solution: good enough, not the best possible. AJ: "This is 'don't worry about the man behind the curtain,' right?"
Supernormal stimuli, and red lipstick
And it is not a one off. Nature is full of these tricks and hacks. If you want to read about it, Hoffman says, look online or "do an AI on" supernormal stimuli, and you will find all sorts of fun cases of organisms being tricked by a stimulus.
We use the same trick on ourselves in design all the time. Makeup is a supernormal stimulus. The red of lipstick is supernormal: in nature almost no woman would have lips that red. But men carry a program for sexual attractiveness, and up to a point, redder lips tickle that algorithm. Push it too far and you are into clown territory, and the effect falls off a cliff. So in these cases, our sensory systems have not evolved to show the truth. They are tricks and hacks.
Marr's exception for humans
This is where Marr comes back in. As a graduate student Hoffman watched Marr, who knew all about this. Marr knew flies had tricks and hacks in their sensory systems; Marr and Poggio had studied the fly visual system and would have said so. But Marr drew a line at us. The human visual system, he said, is estimating the true shapes of surfaces. With our more sophisticated visual systems and our billions of neurons, we can do something the fly cannot: we have been shaped by evolution to see the true shapes of real objects in space and time. "So he was very much a physicalist."
AJ: "But he was wrong about that." "I think that he was wrong." And now Hoffman gives the technical argument, as accessibly as he can, because there is "a nice, clean technical argument that I think takes us apart very, very quickly."
Fitness beats truth, as a theorem
Darwin was not a mathematician. His theory was deep and brilliant but not mathematical. John Maynard Smith and others made it mathematical in the 1970s with evolutionary game theory, so that you can now state Darwin's ideas with precision, prove theorems, and look at the details.
In game theory there are payoffs. If you are playing a game and you take certain actions, you get certain rewards, and what you get depends on who your opponent is, what state the world is in, and what action you take. That is standard von Neumann game theory, and the rules that assign those rewards are called payoff functions.
Now the claim that evolution shapes us to see the true structure of the world can be stated very clearly. The only thing evolution uses to shape you is the payoffs. So if a payoff function does not depend on the structure of the world, and does not somehow carry that structure, it cannot possibly shape you to know that structure. "So it's all about these payoff functions."
What structure? Mathematicians have precise names for the kinds of structure a world could have: topologies, metrics, total orders, partial orders, and so on. Which gives a clean, technical question, "not just a handwave question": what is the probability that a randomly chosen payoff function carries information about the structure of the world, whichever structure you like? That is the probability that a payoff function could even in principle shape you to see the truth.
"And the answer, the probability, is zero. 0% of the payoff functions hold information about the structure of the world."
AJ: "Fit wins." Hoffman puts it the other way around. The version of the story where seeing the truth is what makes you fit, "that version of fitness loses. Big time."
AJ raises a challenge: a group at Yale reran the evolutionary contest with different payoff functions. What did they get right or wrong? Hoffman says they did not address the question he is raising. When you go into the weeds, you can always find particular cases where something nice happens, and he has seen plenty of "little papers" of that kind: in this particular case, sure, you can build a case like that. That is exactly why he goes to the big picture.
The big picture is this. Evolutionary theory, in its current form, does not restrict the class of payoff functions. Any possible function is legitimate. It would be perfectly fine for someone to propose a new version of evolutionary game theory that says only certain classes of payoff functions are permissible, for principled reasons. "But we do not have such a theory right now." So take the set of all possible payoff functions and ask what fraction of them are, in the technical term, homomorphisms of the world's structure: of its total orders, partial orders, metrics, topologies, whatever you pick.
AJ stops to catch the audience up. A homomorphism is a subway map against the subway: the stations on the map correspond to the real world. Hoffman loves it. "The map is a faithful representation of the structure of the subway system. That's a great way of talking about homomorphism. I'll use that in the future." AJ: "I just got to catch them up. It's hard to keep up with you."
So the stunning answer is that the fraction is 0%. What is the probability that we have been shaped to see the truth? 0%. "It's that simple." All his other arguments are optional; this one is the whole case "in one 30 second clip." The set of payoff functions is big. The ones that are homomorphisms have probability zero. Therefore we do not see the truth.
Figure 3. Fitness beats truth, as Hoffman argues it here. Left, the example: the male beetle's perception tracks a cheap proxy that paid off in its niche, not what a female is. Right, the principle: evolutionary game theory places no restriction on payoff functions, and of all of them, the ones that faithfully carry the world's structure (a line inside an area) have probability zero. The wiggly curves are illustrative stand ins for arbitrary payoffs, not data.
"You shot yourself in the foot"
Then Hoffman argues against himself, with the objection he gets constantly. "Don, this is all high falutin math. It sounds really great, but here's the fact. You shot yourself in the foot logically." He started with Darwin's theory, which assumes physical objects in space and time, organisms and resources competing with each other. That is a physicalist framework. Then he used that theory to show there are no organisms in space and time. He used the theory to refute its own foundations. "So you should just go learn some logic, Don. This is stupid."
AJ has seen this. Hoffman gets it in publications, in philosophy journals, and constantly in YouTube comments. AJ: "Don't read those. Stick with the journals."
The reply, he says, is very straightforward. Every scientific theory starts with assumptions, and no scientific theory proves its assumptions. It assumes them. "Those are the miracles of the theory." A theory says: if you grant me these assumptions, please, then I can explain all this other wonderful stuff. If it is a good theory it will, and it will give you the mathematical tools to explore its scope. But that scope is finite, because the theory does not explain its own assumptions. "So no scientific theory is a theory of everything, because no scientific theory explains its own assumptions." AJ: there can never be a theory of everything. "Just that simple. The argument is drop dead simple."
A good theory gives you tools to explore its limited scope. A great theory goes further and gives you the tools to see the limits of its own assumptions, to show that its assumptions are not the final word, which we knew had to be true anyway.
His example: Einstein's theory of spacetime together with quantum mechanics, which gives quantum field theory. Among its assumptions is that space and time are fundamental; quantum fields are defined over spacetime. And when you work through the mathematics, spacetime itself falls apart at the Planck scale, 10 to the minus 33 centimeters and 10 to the minus 43 seconds, where it has no operational meaning. The theory starts by assuming quantum fields in spacetime and then proves spacetime cannot be fundamental.
"No one comes along and says Einstein and the quantum guys should have learned some logic. Those stupid guys." Nobody says that. It is viewed as a breakthrough: science so rigorous and precise that it shows you the limits of its own assumptions. "That's how we make progress."
And that, he says, is what he is claiming about evolution. Darwin started with physical organisms in space and time fighting over resources. Looked at through John Maynard Smith's mathematical version, Darwin's theory turns out to be brilliant enough to show that its own starting assumption, physical objects inside space and time, is not fundamental. "So it's a brilliant theory. That's the way science works."
Always 0% of a theory of everything
So there is infinite job security in science, in principle. There will always be deeper assumptions. What we should avoid is the trap Max Planck pointed out, that science progresses one funeral at a time. Better to let our theories die, and to know up front that our assumptions are just assumptions.
"In fact, my own view is we are 0%. Always 0% of a theory of everything. That's all the science knows." And he thinks it matters to have that zero, rigorously. It should be very humbling, and it means the next generation never has to worry that the older generation did it all. There will always be plenty to do.
AJ asks whether he wants the new generation to start from "theory of everything: 0%" and go from there. Not quite, Hoffman says. What they have to do first is take the current theories very, very seriously: study them, do the homework, know them backwards and forwards. At this level of sophistication "you don't have a prayer of doing something new" until you have spent several years mastering what we have. Only then do you let yourself think outside the box about deeper assumptions.
He gets emails all the time from people who have not done their studies and have a new theory, and it is clear they need ten years of study before they can even begin. "And using an AI is not going to help you. You can't make this gap up." You have to know the theories and have grokked them yourself, or the AI will mislead you into thinking you have a new theory of everything, and it is a waste of time. If you have done the homework, then he thinks it is safe to use AIs as an assistant, because you can step back and evaluate, and let the AI push you around. If you do not know the current theories, you cannot tell when the AI is taking you down a dead end. AJ: AI is great for pattern recognition, but there is no intuition there; you already have to know the subject. That's right, Hoffman says, and if you do, it is a helpful tool.
A mathematical model of consciousness
Before trace logic, photons and the "aha" moment, AJ wants the foundation: what are conscious agents in Hoffman's framework?
If he is going to say consciousness is fundamental and still be a scientist, Hoffman says, he has to have a mathematical model. "And that's no small order." The idea that consciousness is fundamental has been around for thousands of years, and in all that time there has not been a single mathematical model of it with scientific merit, anywhere. AJ asks whether that is really so stunning; everyone seems surprised he has one. What stuns Hoffman is how simple the model is, in retrospect. And even so, he says, he discovered a big step in it just five months ago, after forty years of working on it.
AJ says it took him some reading to grasp the math, "trace A implies B and all of that," and the idea that the universe ticks at a certain rate, but he finally got it. Good, Hoffman says, and starts from the bottom. He and his team, Chetan and Bruce and now many others, needed a mathematically precise model of consciousness, and they wanted the simplest thing possible, "not a Rube Goldberg device," the most cut down model they could find.
Experiences that change
Start with just a conscious observer; agency comes in a minute. What is the minimal thing you would need?
Observers have experiences they can have, say seeing red, green or blue. So list the experiences this observer can have. The other thing that seems necessary is that experiences change. "I'm seeing red now. Maybe I'll see blue next, or red next, or green next." That is the minimum he can imagine. There are experiences, and they change.
What is the most general mathematical object for that? List the experiences down a column. Then next to each one, write a row: if I am having this experience, what is the probability that I will go to that experience, or that one, or that one? You just list all the transition probabilities. "That's it." That little table is a Markov matrix, which Hoffman dates to Andrey Markov in 1905.
AJ asks for the traffic light, because that example is what made it click for him.
A traffic light has red, green and yellow. If you are seeing red now, the probability you will see red next is zero, green next is one, yellow next is zero. So the first row is 0, 1, 0. For green: red next is zero, green next is zero, yellow next is one, so 0, 0, 1. For yellow: red next is one, green zero, yellow zero, so 1, 0, 0. That is the whole matrix for that simple case. It is a special kind, a cyclic matrix. Most Markov matrices are not cyclic; they are more complicated. But it gives the idea.
All possible observers
Now Hoffman makes the move that turns a table into a theory. All possible observers are represented by all possible Markov matrices. One might have a thousand experiences, tastes and colors and smells. Some might have a trillion. Some might have a googol. It goes off to infinity, so there are infinite matrices too. "Just imagine, if you can, and it's hard, the space of all possible matrices of all possible dimensions. That's the space of all possible observers."
The idea is very simple, and the reason for including everything is that he has no reason to exclude any set of observations or any transitions. So all possible observers are all possible matrices, and it is an infinite space.
Think of each matrix as an observer window: a way of looking, a way of seeing. The traffic light matrix is a way of seeing, and if you look at a traffic light, that is what you see, going around and around. Other things, like this room, are much more complicated and would need matrices with trillions of experiences and complicated probabilities. But all of these are passive observers. You are sitting there, looking through the window, not taking any action. You are just watching.
Agency, by recursion
How do you get agency? A clean way to think about it, Hoffman says: I am looking through this window, and now I want to change and look through that window, or that one.
Model it the same way as before. With observations, he listed the possible observations and wrote down a matrix for how he moved around on them, red now, green next. Now he wants to move around on windows. There is an infinite number of observer windows, and he has a policy for which one to look through next: if I am looking through this window, here is the probability I will look through that window, or that one. "That's another Markov matrix."
AJ: but it is also another matrix, isn't it? A single policy is a Markov matrix, Hoffman clarifies, but the set of all possible policies, all the different ways of moving through all the windows, is an infinite collection of Markov matrices, not itself one matrix. And then he can change policies: have this policy, now switch to that one, walking around on policies. "You can see this goes off to infinity. This is what we call recursion."
That recursion is where agency comes from, in degrees. Even the passive windows have a very minimal kind: take an ergodic Markov matrix, and no matter what start state you give it, it always settles into the same long term behavior. "In some sense that's a weak notion of agency. It's always trying to get to the same place." Policies, moving around on windows, give a little more agency. Being able to change policies gives more flexibility still, because now you are changing how you move around. "As you go meta, meta, meta, you're getting ever more sophisticated agency, off to infinity." It unpacks agency from the most trivial, the mostly passive observer windows, out to infinite flexibility.
What he finds beautiful is that science does not have to take on all of agency at once. Go through the recursion once, really understand what policies can do, master that, then move to meta policies, and so on, from least complex to most complex. "But I haven't told you the most fun part."
Figure 4. Hoffman's two building blocks. Left, the traffic light he and AJ build row by row: the entire observer is one Markov matrix. Right, the recursion that turns observers into agents: every level up is another Markov matrix whose "states" are the matrices of the level below, so agency comes in degrees rather than all at once.
Leibniz was right, and Newton did not have the math
Before the fun part, Hoffman tips his hat. Around 1700, in his Monadology, Gottfried Leibniz was saying something like this. Hoffman does not want to put words in his mouth, but suspects Leibniz would like the approach. Leibniz said science needs to start with perceiving entities, observers, which he called monads. He was quite religious, and he said there had to be some kind of pre established harmony, set up by God, so that the monads are coordinated rather than random observers, completely disconnected, each doing whatever.
At the same time Isaac Newton was trying to find a mathematical foundation for science, and Newton was also quite religious. He wrote more theology than physics. AJ: "Did more alchemy than he did physics." "More alchemy. That's right." Hoffman thinks Newton might have liked to put consciousness and observers at the foundation, "but the math just wasn't there." So Newton went with what he could do. He could write down F = ma, and F = G m1 m2 over r squared for gravity, and get going. That set of equations gives you a physicalist, machine kind of universe, and science got started in a machine universe "because that's what we could do with the mathematics."
"I think Leibniz was right, and I think Newton would have liked to go that direction, but the math just wasn't there to do that." And now, he says, he can show AJ something that looks like the pre established harmony Leibniz was looking for.
The trace: ten colors, three visible
Forget agency for a moment and take just the observer windows.
Suppose I am looking at ten colors, and there is a matrix governing how I see those ten colors. That is my window. Now suppose that while I am looking through it, someone somehow shuts off seven of the colors, so I can only see three. I am still looking through the window, but now I effectively get a 3 by 3 matrix on those three colors, induced by the big 10 by 10.
AJ: so there is all sorts of hidden stuff going on in those seven that you can't see, but it's there. "They're hidden, but they do influence what I'm seeing in the 3 by 3." That induced matrix is the trace.
For the mathematically sophisticated, Hoffman adds a clarification. The more common meaning of trace for a matrix is adding up the diagonal entries, 1,1 plus 2,2 plus 3,3 through n,n. "I'm not talking about that trace." This one is more sophisticated, and for mathematicians it is the Schur complement: take a big matrix and induce a matrix on the subset you can see. (In standard Markov chain notation, if A is the visible set and B the hidden set, the trace chain's transition matrix is P_AA + P_AB (I − P_BB)⁻¹ P_BA: go directly between visible states, or leave, wander among hidden states for any number of steps, and come back. That is the "exit, hidden world, re entry" structure Hoffman returns to later when he talks about UAPs.)
AJ: like a trace element, a small part of it. Right, Hoffman says, and the small part is in effect reflecting the whole, only through a smaller window. "It's almost like the smaller one is observing the whole, but through the smaller window." And that is the trace.
He is explicit that he did not invent this. The notion of the trace of a Markov chain has been known for maybe 50 or 60 years.
A logic on all Markov chains
What he discovered, about two years ago, was that the trace relationship is a partial order on all Markov chains. "It gives you a logic," and for those who know some mathematics, the logic is not Boolean. Boolean logics are the ones we are most comfortable with, where you can take ands and ors and complements.
AJ: but Boolean fits into your theory at the local level, doesn't it? It does. This logic, which Hoffman calls trace logic, is not Boolean, but it contains an infinite number of Boolean sublogics. Take any matrix, and look at all the matrices that are traces of it: all of those together form a Boolean sublogic. "It's really, really pretty." AJ: "It's elegant." So locally, everywhere, you have Boolean logic; you have infinitely many local Booleans. How they get tied together into the whole is the part the team is still working out. "This is nasty math. We're still trying to grok this."
The theorem at Heathrow
He still remembers the conversation two years ago. He said to Chetan Prakash: "Chetan, I believe that this trace relationship will give us a logic. It's a partial order and it gives us a logic." Prakash's response: "Don, that's too pretty to be true."
Then Prakash went off and proved it. What he had to prove was that the relationship is transitive, and AJ puts it in his own shorthand: trace A implies B. In plain terms: the trace of a trace is a trace. Take a 10 by 10, trace it down to a 5 by 5, then trace the 5 by 5 down to a 3 by 3, and you get exactly the same answer as tracing the 10 by 10 straight down to the 3 by 3.
AJ asks what Prakash said when he called from Heathrow after proving it. Pretty matter of fact, Hoffman says. It's true. Hoffman had believed it was true, "but when Chetan says it's true, then I know it's true, because Chetan is brilliant and he doesn't usually make mistakes." He was very pleased.
And now you can see how far it reaches. The trace logic applies to the observer windows. But the policies, the first level of agency, are also a whole set of Markov kernels, so there is a trace logic on them. And on the meta policies, and the meta meta policies. "So you get a recursion of these trace logics, all the way out to infinity."
"And I want to propose that this is the pre established harmony that Leibniz was looking for." All these observers, tied together by one beautiful recursive mathematical logic. They are still trying to understand it; there is a lot of mathematics left to do.
Figure 5. The trace, and the theorem. A small observer window is the trace of a bigger one: its effective dynamics fold in every excursion through states it cannot see. Hoffman's observation is that "is a trace of" orders all Markov chains into a logic; Prakash's proof of transitivity is what makes it an order. The same block structure (visible, exits, hidden, re entry) is what Hoffman later uses to talk about things that leave and re enter our headset.
"Just logic, so what?"
AJ asks how the establishment is responding. The theorem is true, Hoffman says; there is no problem there. AJ: and if Chetan proved it, others can check it. Right, and for a mathematician, "this is falling off a log." Prakash proved it while he was waiting at Heathrow. A brilliant mathematician can do it pretty quickly.
"So that's just logic," AJ says. "So what? How do we bolt your philosophy onto it? That's what makes it controversial."
Hoffman agrees there are two claims. One is that he found a new recursive structure on the set of all Markov chains. That is a contribution to mathematics, and as a contribution to mathematics he does not think it is controversial at all. The other is that it applies to consciousness, "and that's very controversial."
"A rookie mistake"
Here is the kind of thing that gets said publicly, he says. A very prominent person with a YouTube presence (he declines to name them) called it complete nonsense. Markov chains are fine, this person said, but they are used for standard things like predicting the weather and the stock market. AJ adds nuclear fission and Google's PageRank. Exactly, Hoffman says; they are used for all of this, and to say they apply to consciousness is nonsense, because nothing in the mathematics says it is about consciousness. "For Hoffman to even say that it's about consciousness is just a rookie mistake."
AJ has seen that, but never seen anyone explain how it is a mistake. They don't, Hoffman says. They just say it. So he explains why it isn't.
Mathematics never tells you what it can be applied to. There is nothing in Markov chains that says they apply to stock markets. There is nothing in them that says they apply to weather. Once you see that, "the claim you can't use it for consciousness is silly. It's just plain silly," because the same argument would forbid every other application. The math is a structure. Using it in science means saying: I think this arena might profitably be modeled by this structure. Maybe stock markets, maybe weather. In his case, consciousness.
What he thinks is really going on is that the critics simply do not like consciousness. What they are really saying is "consciousness is nonsense, and you're trying to attach nonsense to Markov chains." And he grants the possibility. "You could be right. Maybe consciousness is nonsense. We'll see." But right now physicalist approaches cannot explain a single illusion of a conscious experience, much less a conscious experience. "If in 50 years we're still batting zero, I would say it's over for physicalism. It's just over." And if you can start with a theory of consciousness modeled by Markov chains and do real work with it, like building up spacetime, that would not prove consciousness can be modeled by Markov chains, "but it sure makes it a pretty interesting scientific hypothesis. Certainly not nonsense."
From Markov chains to consciousness
"Let's build that bridge," AJ says. How do we get from the chains to consciousness?
What drew Hoffman to Markov chains is that at the most elementary level, what we have is experiences, of colors, of shapes, of more complicated things like three dimensional objects, and they change. At that minimal level the Markov model fits.
He raises the first objection himself. A Markov chain has a finite memory: the next transition depends only on the current state. Some people say that is far too limited. It is not, he says, because it is well known in Markov chain theory that you can build bigger, more complex states. Take a series of twenty states and make it one state. You can build as much history as you want into the states, you can have an infinite number of states, and the individual experiences can be as complicated as you like. He uses red, green and blue because they are simple, but a state could be a particular three dimensional shape, a sphere, a cube, or something far more complicated.
So he thinks it is a good hypothesis to use Markov chains, organized by what he calls the recursive trace logic (the recursion of chains, policies and meta policies, with the trace logic on every level), as a theory of observation and agency, because it is the most general and comprehensive option with the fewest assumptions. "There are experiences and they change. It's amazing. That's all I assume. Literally, that's it." The best way to model that is Markov chains, and "the recursive trace logic just falls out of that. It's just a theorem." That is what he loves about it as science: a minimal assumption (start with consciousness, model it as experiences that change, use the minimal model, Markov chains), "and oh, by the way, no one noticed it, but there's this partial order, the trace logic, on all this, and it's recursive, and this gives us a theory of agency. It's just that simple."
AJ tries a summary: so consciousness is just a state that can change. The states change, Hoffman says, but the policies and meta policies change them in an agentic way. In effect they say: given that I am looking at the world through this window now, these are the ways I want to look next. That brings in agency, and it also brings in a notion of time, "because you can't choose what next window you're going to go to until you have a current window." It is a very observer centered, agent centered notion of time.
The second objection: these are just Markov chains, with no deliberation, no reasoning, no goals, no payoff functions being maximized. Hoffman's answer is that those are different levels of description of the same thing. Give him a decision theoretic description of an agent (I have these goals, I can take these actions, I model these possible worlds, so I can work out what each action does toward my goals in each world) and when you actually write it down, you get a Markov chain: given that I am in this world, what is the probability I will do this, or go to that world. He prefers the Markov chain level because it has the recursive trace logic, "and also because it's looking to me like we'll be able to show how we can build spacetime and quantum theory from just this recursive trace logic."
What his side owes
Then he turns his own standard on himself. He has been very hard nosed about the physicists in this interview: they start with space, time and physical objects as fundamental, so they owe a precise account of conscious experience, or of its illusion, with no handwaving. "Show me how we get the illusion of the taste of mint from neurons. Show it to me, or why should I believe you?"
Now turn it around. His colleagues say: "Don, great. You've got this nice mathematics and you're claiming it's consciousness. Where then do spacetime and quantum field theory and the Born rule and all this stuff come from?" What he owes them is special relativity, general relativity, quantum field theory, eventually quantum gravity. Nonlocality. The Born rule. The Big Bang. "Can you give us this?"
His answer: he just gave his colleagues Niffe Hermansson (whom the captions call "Nifa") and Chetan Prakash tentative proofs of special relativity and general relativity. They look very plausible to him. And he is working on the Born rule.
AJ asks whether that means integrating time dilation and length contraction. It does, and Hoffman offers an intuition for why it might work.
Einstein's train
One thing Einstein taught us in special relativity: suppose AJ is on a train going past Hoffman, who is sitting at the station. Each has a clock and a meter stick. When Hoffman looks at AJ's clock, it is running too slow, and AJ's meter stick is too short. And AJ, looking at Hoffman from the train, says no, Don's clock is running too slow and Don's meter stick is too short. That was stunning, very counterintuitive, not accepted right away, and experiments have shown it is correct.
AJ: does Minkowski spacetime factor into this? That is exactly where Minkowski spacetime comes from, Hoffman says, the case without gravity. Einstein did that in 1905, and it was 1915 before he had curved spacetime; it took him ten years to master the mathematics. AJ: "Oh, is that all?" A Herculean job, Hoffman says. Truly impressive.
The counter
So why should anyone believe Markov chains could give you Einstein? Here is the piece he has not mentioned yet. It is standard in Markov chain theory to attach a little counter to a chain. Every time there is a transition, you increment it. See red: one. Now green: two. Now blue: three. You just keep counting. That is what they call an enhanced Markov chain, and he stresses that it is textbook material, "not our invention."
Now go back to the big 10 by 10 window with ten colors. Every time any of the ten colors changes, its counter clicks. AJ asks: only on a change of state? On every transition, Hoffman says, including red to red; a state can transition to itself.
Now take the 3 by 3 sub window that only has, say, red, green and blue. Its counter only clicks when red, green or blue happens. It never catches yellow, purple, pink or the rest. "So notice its counter isn't going to go as fast as the big counter." The 10 by 10 counts every change among all ten colors; the 3 by 3 is only getting "30% roughly" of the counts.
AJ wants it concrete: give me colored glasses, and let's click our counters. Keep it simple, Hoffman says: three colors versus two. We are at a traffic light, red, green, yellow. Every time it changes, red to green, green to yellow, yellow back to red, you click your counter. Now suppose you cannot see the yellows. All you see is red and green, so you click when red changes to green and when green changes to red, but you never get a yellow. "You're missing a third." AJ: one out of three counts. "And so one clock is only going at two thirds the speed of the other clock. And that is where we're going to get Einstein's time dilation."
AJ: so time is moving slower for me, because your counter is slower? The way Hoffman puts it: the reason he would see AJ's counter as running slow is that his Markov chain is not completely intersected with AJ's. He is not counting all the things AJ sees, so from where he sits AJ's clock does not advance the way AJ's own counter does. And it is symmetric: from AJ's point of view, Hoffman's counter is too slow, because AJ is missing some of what Hoffman sees. "So that's why we get the time dilation."
Figure 6. The intuition Hoffman gives for his tentative relativity proof, rebuilt from his traffic light example. The blind observer's chain is the trace of the full one onto red and green, so green to yellow to red collapses into a single green to red click. The derivation itself is unpublished; this is the picture, not the proof.
No seconds, no frame rate
AJ asks the natural follow up. What is the minimum unit? Is that a Planck time? What is our universe's frame rate?
In the recursive trace logic, Hoffman says, there is no minimum, because time is a spacetime notion. There are no seconds in it at all. If we are talking about the Planck time, that is 10 to the minus 43 seconds, but the very notion of a second is alien to the recursive trace logic. It only has counters. What he has to show is that he can use it to build Einstein's special relativity.
AJ pushes: don't you need some kind of base measurement to count? You have to be counting something. There will be something like a fundamental clicker, Hoffman agrees, but no notion of seconds tied to it. Instead of 10 to the minus 43 seconds, it could be 10 to the minus 43 trillion seconds, or 10 to the minus infinity. It can be as small as you wish. "But our spacetime is stuck at 10 to the minus 43 seconds," which, if you think about it, raises the question of why not 10 to the minus 43 trillion. And spacetime falls apart at 10 to the minus 33 centimeters. We think that is small. Why not 10 to the minus 33 trillion? Why should spacetime fall apart there? "That's a fairly shallow data structure."
So what he has to show is how the recursive trace logic, which has no seconds, just Markov chains with counters, can create what he calls a headset.
The headset
AJ had been waiting for that word, because until now, he points out, all of this could still be physicalism. Markov chains do not care what they describe. "Oh, sure. Absolutely," Hoffman says.
The headset is what lets him go from saying "the recursive trace logic models consciousness" to "and here is how we get what we call the physical world out of it." Building a headset means starting with universal consciousness, the recursive trace logic, mathematically described, "with the pre established harmony that Leibniz wanted," and showing how Einstein's special and general relativity and quantum field theory come out as a special model within it.
And he wants to be very careful about one thing. He does not want to prove that the recursive trace logic forces us to see the world through Minkowski space, general relativity or quantum field theory. "It doesn't. All I need to do is prove that it allows us to build those structures." In fact he would be "bitterly disappointed" if it forced them, because he wants the flexibility to show that our spacetime is one of an infinite number of headsets consciousness can build.
"My view is ours is one of the most trivial and simple spacetime headsets that's available. We have the training wheels version, really dumbed down." Our view of ourselves as the top of the food chain and the top of intelligence is exactly backwards. He plans to show it, and says he has a version of the proof right now, "but it's not real until it's published." Chetan and Niffe are looking at it. It is not real until they say it is real, and even then, AJ finishes, not until peer reviewed publication.
Infinitely many headsets
AJ: if the headset is how we perceive reality and ours is basic, that implies something more advanced. The mathematics makes it very clear, Hoffman says, that there is an infinite, unbounded number of far more interesting headsets. You could think of ours as a trace of much bigger headsets.
Even among three dimensional headsets there is variety. Presumably mice see in three dimensions, and some birds, and their headsets will differ from ours: still three dimensional, but with different features. So the plan is to first prove they can get a generic spacetime headset, not human, not mouse, not bird, and then look at the neuroscience of humans specifically.
AJ: if there are infinitely many headsets above, must there be infinitely many below? There could be quite a few below, Hoffman says, but it is hard to go below three dimensions if you are counting dimensions: three, two, one. AJ: so dimensionality matters. For that way of measuring headsets, yes. But you could get rid of dimension as well, and have headsets with topologies and no dimensions at all. "You can't think big enough. There's an infinite number of different kinds of headsets."
AJ: humans don't like infinity. "We don't like infinity. We like to think that what we're seeing is the truth." And what he is saying is that spacetime, which science took to be the final reality with everything inside it, is one of the most trivial headsets you could build out of the recursive trace logic, with an infinite number of more complicated ones beyond it. "And I think that we are the consciousness that's capable of understanding those headsets."
AJ: you mean our species? Our species, Hoffman says, is a headset representation of the consciousness that is actually able to do all of this. Part of our joy here is to enjoy this headset, then wake up and realize it is just a headset, like Leela. This is a game. "We let ourselves get lost in this game, and we can smile once we wake up and realize, oh, we thought this was the whole thing." As beautiful, complicated and completely engaging as this world is, it is trivial compared to what you can do. So relax and play with it, learn from it, be open, explore, and realize that you infinitely transcend it. The recursive trace logic says that mathematically: there are infinitely many other headsets you can build.
AJ asks for something practical, "for my monkey brain": what might those other headsets see? One simple direction, Hoffman says: go from three dimensions of space to four, five, fifty, a billion. AJ: dimensions are infinite as well? "Why not? Why not build dimensions off to infinity?" And that is only one direction. Then let go of dimension altogether and try different topologies. Mathematics helps here, because all the structures mathematicians have discovered open you up to the possibilities and infinities. "I study mathematics because it really helps push me out of my little boxes," out of his thought dead ends.
Figure 7. The whole architecture as Hoffman lays it out across the second hour. The top two layers are his hypothesis about what is fundamental; the middle row is the claim that spacetime is one buildable headset among unboundedly many; the bottom left is the work he says the Trace Institute must finish or be wrong; the bottom right is how familiar things look from inside.
Who gets him out of trouble
AJ quotes Hoffman back to himself: he knows enough math to get into trouble but not enough to get out. So who gets him out?
His good friend Chetan, who has worked with him since 1984 or so, "so it's been 40, 42 years he's put up with me." Bruce Bennett worked with him until the early 2000s, and Bruce was brilliant. Now he also works with Niffe Hermansson, who is an expert in Markov chains, "so this is his area." And at the Trace Institute they are looking to bring in more mathematicians. "The more the better."
AJ says he is building a nice roster, and asks what a Tuesday is like at the Trace Institute. Right now it is distributed, Hoffman says; they meet by Zoom, because Niffe is in New Zealand and Chetan is about an hour and a half away from him. Robert Prentner is in Shanghai. Prentner is a former postdoc of Hoffman's, now a professor in Shanghai, and he is going to be the leader of the Trace Institute. "He's a younger guy. I'm 70. He's in his 40s. So it's time for me to make sure that someone younger is in charge."
The Trace Institute's nine conjectures
What is the mission? Hoffman admits he has forgotten the exact wording of the mission statement, so AJ asks what they are generally trying to accomplish.
The idea is to explore the recursive trace logic until it is completely understood mathematically, and then prove the nine conjectures the Trace Institute has published on its site. He lists them as he remembers them:
One: they can build Minkowski space, Einstein's special relativity.
Two: they can build general relativity.
Three: they can get the Born rule.
Another: they can get the Big Bang.
Another: they can get quantum nonlocality. AJ: "You can get nonlocality out of this?" "Absolutely."
And ultimately, everything in quantum mechanics.
"If we can't, then we're wrong, by the way." These conjectures are what anybody would require of him.
AJ recalls that someone (the captions render the name "Kristen") challenged Hoffman to derive the Schrödinger equation this way. That is among the conjectures, Hoffman says, and more: not just the Schrödinger equation but all of quantum field theory, completely, out of this. AJ jokes that the Nobel committee is probably listening.
Is this something they can release piece by piece, "hey gang, conjecture four is done, we're moving on"? That is exactly the goal, Hoffman says, to get the nine conjectures proven in the next two or three years. AJ: quantum field theory within three years? "That's our goal." And again the careful framing: they do not have to show the recursive trace logic uniquely gives quantum field theory. They only have to show it can give quantum field theory, and then that it can give infinitely many other structures that are probably much more interesting.
What will impress scientists, though, is the mathematics and the spacetime physics they already know: general and special relativity, quantum field theory, the Born rule, the Big Bang. And he gives one brief reason he is sure they can do it: the space of all Markov chains is computationally universal. Anything that can be computed by any Turing machine can be done by Markov chains.
Why call it consciousness at all?
AJ is candid. The math makes complete sense to him. The jump to consciousness he cannot get his mind around yet. He does not see why "consciousness is fundamental" is required for any of this to work; it seems like it could work with physicalism just fine.
Hoffman grants the point outright. You could say, I don't like the consciousness stuff, I just want this to be a theory of observers and agents without the consciousness label, and it would work perfectly fine. AJ: so consciousness could just be a label? It could, Hoffman says. Here is why that might end up being uncomfortable.
"If you hit me on the thumb with a hammer, I feel there's something I can't ignore." It is a real experience, an unpleasant one, and "if anything is real, that painful thumb is real. And if that's not real, I don't know what is. And if the taste of chocolate isn't real, I don't know what is." The abstract structure we call spacetime may or may not be real. But his experience right now, as he looks around, is real. Einstein gives us beautiful mathematics that nicely describes it, but all he knows first person is that he is experiencing colors, distances and smells. Einstein's work is great and it works well, but he is not sure it is the final reality; it might just be a description of his particular kind of experiences. And what about a shark, or a bat using echolocation? Why should its world be anything like his? There are all these different sensory worlds.
Ants and hands
AJ asks him to explain that using ants and hands, an example he has heard Hoffman use: from an ant's perspective, you could reach down and kill it at any time.
Hoffman admits he does not know much about ants beyond chemical signals, then takes the example. If he sees an ant crawling around on his dining table, he can kill it, and it will not even know what is about to happen. He can just put his finger on it. (His wife, he notes, probably would not want him to; she would put it on a piece of paper and carry it outside, to be kind. The ant would not know in either case.)
That is his perspective on the ant. From where he sits, it does not seem terribly bright. It does some chemical tracing, and apparently some dead reckoning (path integration), wandering around and then finding its way back, which is pretty smart. But it looks simple compared to a human.
Now flip it. From the ant's point of view, how much would the ant know about AJ? AJ: nothing. Almost nothing, Hoffman says, and it might not even know AJ is there. And if it did, maybe its representation of AJ would be as simple as his representation of an ant. "So by symmetry, I have to ask myself: what I think of as an ant, maybe that's just because of the limitations of my own headset." Maybe his headset is dumbing things down so much that he is actually interacting with an incredible intelligence, far greater and more capable than he is. AJ: and you just have a trace matrix of it. "That's right. I'm just seeing a trace of this thing, and I'm getting an ant." Because that is all your headset allows you to see.
It also means, he says, "and this is where UAP stuff comes in," that there could be other consciousnesses, or other agents or observers if you do not like the word conscious, in much bigger headsets than ours. Our headset might be to them what the ant's headset is to us. "We can go down and smash the ant anytime, and they could come down and smash us anytime, because we're trivial compared to them." The headsets go off to infinity.
Are DMT aliens real?
After the break, AJ notes that Andrew Gallimore was sitting in that same chair not long ago, and that he and Hoffman are now working together, which AJ loves. Their first project is literally titled "Are DMT aliens real?" (It became the paper Traces of the Other, and its public launch is rebuilt on this site as Traces of the Other: Are DMT Entities Real?.)
So, are they? Hoffman thinks he and Gallimore would both say the emphasis on "real" is a little misguided, "but it catches attention." AJ: "It gets funding." "That's right."
The idea runs like this. Many kinds of headsets are possible. Ours is not the final reality; spacetime is not the final reality. There could be headsets much bigger than ours that can play with ours, and entities in higher headsets could play with us the way we play with ants. So the question is: does DMT just screw you up and make you hallucinate? Or does it somehow let us modify our headset, open it up to more dimensions, perhaps more dimensions of space, perhaps whole new kinds of conscious experience?
And if so, could the entities people meet in DMT space be avatars? When Hoffman talks with AJ, he is not in direct contact with AJ's consciousness. He is seeing an avatar we call the human body, which lets him, with his headset, interact with AJ's consciousness, and the Hoffman avatar lets AJ interact with his. And even these avatars do not exist when they are not perceived. We render them on the fly. "So the AJ that I'm seeing is in some sense not real, because that AJ is gone" the moment he stops looking. AJ: "Gone." "Completely gone." AJ's consciousness presumably is not gone. Hoffman's avatar of AJ is gone, while AJ is just fine.
So the real question is whether the DMT entities are genuine avatars. "We put real in because it's catchy." One property of an avatar is that you can share information through it. Hoffman can tell AJ something, ask AJ to tell someone else, and then check with that person whether they heard exactly what he said. That is the kind of question they can ask about DMT entities: could they send two people into DMT space? There are certain entities that seem to come up again and again.
Higher resolution
"They do," AJ says. "I've been in that space." Hoffman asks if it feels more real than this. It does. Hoffman has heard that the resolution seems higher, that by comparison this world seems low resolution. AJ: "This seems black and white."
"And the funny thing is, that's what the recursive trace logic is telling me." This is one of the cheaper headsets. AJ had not connected those two things. "When I say this is a cheap headset, I really mean it. We got the cheap version." And probably, Hoffman adds, even what AJ sees on DMT is cheap compared to other things.
AJ offers a personal data point. There are certain psychedelics after which he does not need reading glasses for about a week. He can just see; colors are sharper. It drives his wife nuts ("You can read that?"). "It's like superpowers for about a week," something to do with neuroplasticity. Hoffman wants to see that too. That is exactly the kind of thing the recursive trace logic could let them really understand. If they understand how this headset is built, and what DMT is doing to change it, they can reverse engineer it.
The tesseract
Niffe Hermansson, Hoffman says, has done some DMT. He is a mathematician, and he told Hoffman that he went in and saw a tesseract, a four dimensional cube, rotating rigidly in four dimensions. "Again, it's not proof, but it's good evidence from a reliable mathematician" that he was at least in a four dimensional space and was seeing a rigid tesseract, for a fraction of a second. He did not hold it for a whole second, but he saw it.
To Hoffman that makes the question worthwhile. "That's not proof. I mean, hard nosed scientist, that's not proof. But it's very suggestive" that DMT, a chemical in our headset, is a headset representation of a tool outside space and time that lets us change parameters of our headset.
The brain is a headset representation of how the headset is constructed
What they have to do, then, is prove the nine conjectures and build the human headset, and that is going to take a lot of neuroscience. They are going to have a neuroscience team. There are 86 billion neurons and trillions of synapses to understand.
"And I'll be very slow here. The brain is a headset representation of how the headset is constructed." AJ: "That's a big concept." Hoffman repeats it: the brain, the nervous system, is a headset representation of how the headset is constructed. "So I'm not getting rid of neuroscience. I'm saying we need more neuroscience, and it's going to be much harder than my colleagues in neuroscience think it is." Understanding the 86 billion neurons and trillions of synapses is only the first step. The hard step is reverse engineering them to understand the software, from the recursive trace logic, that is being used to build the human spacetime headset.
Once they understand that, "and it won't be next week," they can ask the technical question: what exactly is DMT doing to the construction of that headset? Is it just taking us from three dimensions to four, or to more? What else is it doing?
The wizard and the geek
"Once you know the software that's building a VR game, you can do miracles in the game." You are no longer bound by its rules, because you are writing them.
In Grand Theft Auto, the player who is a wizard is wonderful; he can play the whole game by the game's rules. "But the geek who wrote the code can take the gas out of the car of the wizard." He can do anything he wants.
That, Hoffman says, is what the recursive trace logic offers once the nine conjectures are proven and they begin to understand how the headset is designed. "You cannot think big enough about the technologies that will come out of this. It will make everything that we've got seem like firecrackers." When people see the technologies, "then it'll be game over for physicalism." AJ: "Game over, physicalism. Wow. These are big goals."
AJ asks whether replicating the brain as a headset means building an artificial neural network, or doing it with proofs. They are already working from very clean microscopic sections, Hoffman says, piecing together what we can see of the neural structures inside spacetime and assembling all the physiology, a really complicated process that has to continue. But then they have to think outside of spacetime completely. AJ: because even if you got the brain perfect, you are not seeing everything. "No, you're only seeing a compression." There is some really complicated software "out here, so to speak," that gets compressed in this headset into what we call the brain. The brain looks complicated from inside, 86 billion neurons and trillions of synapses, "but it's trivial compared to what's outside." The headset restricts, funnels down, loses information. "So we need more money for neuroscience, not less. A lot more."
AJ: "We're living the compressed data stream. We're not getting a lossless data stream." That's right, Hoffman says, and AJ's own DMT experience fits: from there, this already feels like the lossy interface.
UAPs
AJ brings back something from earlier: UAPs. Is that what they are, objects from a higher headset?
"I don't know." But the recursive trace logic leaves that possibility open and gives a rigorous way to start thinking about it. One thing about the trace logic, he reminds AJ, is that it is about attention. If there is a 10 by 10 matrix of colors and you only see the 3 by 3, you are only paying attention to three of the colors, "and there's all sorts of magic that can happen outside, in the other seven colors."
Look at the trace construction. You have your visible states and their matrix. In the big matrix there are transitions from the visible into the invisible. Then there are transitions among the fully invisible part, the 7 by 7. Then there are transitions back in. "So you have the exit, the external world, and the re entrance." AJ: so that could explain everything from UAP teleportation to particles appearing out of the vacuum (virtual particles). That's right, Hoffman says. It shows how something could exit our spacetime headset into an entire world infinitely more complicated than our headset, and then re enter.
That is one aspect. Another is simply the magician's trick. A lot of stage magic works by manipulating attention: I do this, you automatically look over there, and meanwhile I do something over here. "So the recursive trace logic is the logic of attention." One way these things could happen is just distract and change. Once you are higher up in the recursive trace logic, there are all sorts of tools you could use to hoodwink consciousnesses stuck with smaller headsets.
Inside spacetime, by contrast, we have Einstein's theory of gravity, and we are working on quantum gravity, and with gravity you cannot do what these UAPs seem to do. AJ runs the list: hover with no apparent propulsion, move at Mach 40 instantly, at 466 g, go into the water with no displacement. "That's just not possible with our physics. But it's certainly possible if you know the software of our headset and you play with the software."
AJ: "So we're the ant, then." "We're the ant." If they understand the software of our headset, it is like the geek in Grand Theft Auto. The wizard will be stunned. It does not obey his rules or his laws. He will say it is impossible, and it is not impossible, "because the geek is not stuck in the headset. The geek is making the headset." So with UAPs we may be dealing with higher levels that understand how our headset is built and can play with its rules.
AJ: it feels like they are doing that. So a UAP doing Mach 40 is just hitting that clicker faster than we can perceive? That's right, Hoffman says, or it could be distracting us. AJ: there's a trickster element to it, for sure. There could be a trickster, Hoffman says, and that goes back to Leela. It is very clear from their technology that if they wanted to destroy us, we would be completely helpless. There are cases that seem to show some kind of hostility, "but if they were really hostile, we wouldn't last five seconds." So he gets the feeling of a playful aspect: let's explore more, let's play with these.
"And if it really is the one consciousness looking at itself through various headsets, it could be the one consciousness saying: I put myself on these really stupid human avatar headsets. I want to give them a little prod from a little higher headset, to sort of wake up." The one source consciousness, playing with itself at various levels. And with the recursive trace logic, he says, we could begin to see scientifically how that is being done.
A humble note
Then he puts "a humble note on the whole thing." The recursive trace logic is just a scientific theory. It makes its own assumptions, and he looks forward to replacing it at some point; we will need to go beyond it. For now he thinks it is a good next step. "But whatever the fundamental reality is, what we might call the source, it transcends any scientific description, including mine."
What the recursive trace logic really is, he says, is "the logic of an infinite number of perspectives that the one source can take on itself." Each perspective is consistent on its own. Each Boolean sublogic is a consistent view, but different Boolean logics can contradict each other, and they are all useful perspectives on the source, which transcends all of them. AJ: "That's beautiful."
AJ asks whether there is a catch. If we evolved not to perceive all this, is that an impediment to the research? Does the source just say, I'm not going to let him figure this out? Hoffman's answer: "You and I are just the source looking through a particular headset." The game is such that very few of us are Einsteins and Schrödingers, and we are grateful for all of them. But the Einsteins and Schrödingers are trivial compared to what is out there to be discovered.
"And yet you and I are fully that source." In silence, we go into the space of infinite intelligence that it is. All of us, in some sense, are that infinite intelligence, and we have chosen to play in this avatar, to allow ourselves limitations, to take this perspective very, very seriously, to look at the source through this particular lens and to play. "We take ourselves very seriously. It's all very, very serious. But really it's about playing with this perspective," really enjoying it, and then letting it go and realizing it was just a perspective, and that you infinitely transcend it.
The goodbye text
AJ says he thinks he has come to understand where Hoffman's philosophy comes from, and it runs through his heart scare after COVID. His heart ran at 190 beats a minute for 30 hours. He was in the hospital. There were two surgeries. And he texted his wife goodbye. What was in that text?
It was very, very short, Hoffman says. He was in a horrible place, exhausted, and could hardly think. He texted his wife, and then, he thinks, his daughter separately. "I just said: I don't think I'm going to make it. I love you. Goodbye." It was that simple. The one to his daughter was something like: my heart's been beating 190 beats a minute for 30 hours. I don't think I'm going to make it. I love you. Goodbye.
"As soon as I did that, I laid back in the bed to die. Just to wait for it to happen."
AJ asks whether he was at peace in that moment or scared. "I was scared." He was exhausted and his heart was racing, and a heart racing at 190 beats a minute makes you feel scared to death all by itself. "Even though it's just atrial fibrillation, it feels to you like you're scared to death."
Two figures at 3 a.m.
"Now, I've only been recently starting to talk about this."
At the very moment he leaned back, at about three in the morning in the hospital, he saw two male figures standing next to his bed. He had not seen them before. One of them looked at the other, smiled briefly, lightly, and nodded his head.
"And then all of a sudden I felt an incredible warmth in my heart, and my heart started beating normally."
AJ: "What?"
He looked down, and looked up, and they were not there anymore. Neither of them had been wearing hospital garb.
AJ has never heard him tell this part. "No, I haven't done it publicly before." Who does he think they were?
He did not talk about it for several years, because it was so strange. But in the last couple of years he has reasoned through it. First, if they were hospital staff, it would have been malpractice to wait 30 hours to give a patient a drug they knew would fix him. That did not make sense. Second, when he felt the warmth, he did not feel it come in through the IV; he felt it directly in his heart, and nothing in his arm. Third, he never saw those two men before or since.
"So I don't know what to make of it. I'm just stating what I saw. And I'm here, because my heart flipped from 190 beats a minute for 30 hours to, instantly, a feeling of warmth, and it was done."
AJ asks what he said to the doctor who came in to ask what had happened. He was too exhausted to even grok what was going on. All he knew was that he was so glad he was not dead, so glad his heart was not beating like that anymore, and that he needed rest. He could not wrap his head around it. He tried to think outside the box, "but that was so far out of the box that I just let it go for three or four years. I didn't even think about it."
"I believe you," AJ says.
Just for Laughs, and the ego
AJ connects the story to something lighter: from this, to Just for Laughs: Gags. "That's your show that you love, right?" Hoffman is delighted. "How'd you know that? I love Just for Laughs. It's one of my favorites."
Why? Because people get surprised. They are set up to expect one thing, they get upset and tense, and then in many cases they end up laughing, because they realize: "Oh, wait, wait, wait. I don't need to be upset at all. That was just a gag the whole time." For Hoffman the show is a good picture of what he thinks the whole thing is. "The whole point is to relax and enjoy the game."
He thinks about his research progress the same way. The more he can relax, enjoy, and be open, the better. Let go of everything he thinks he knows, with no egoic attachment to it. "The ego is the biggest, biggest impediment to growth and discovery. So let go of the ego, let go of attachment, let go of all that stuff, and be open like a little child." He quotes Jesus: unless you become like a little child, you cannot enter the kingdom of heaven. "And I really think that that's right." You have to be open enough to say: I don't know anything. As much as I have learned, I know 0%. Let me know more. "And play the game. I think that's what it's about."
AJ: and don't take it too seriously; when it's over you might be scared, but when it's finally over you can just have a big laugh. "And even being scared was part of the whole deal," Hoffman says. You can go back and laugh even at being scared, but for some reason we set it up so that we want to go through the scared part too. "I was scared. I mean, there's no... I'm no hero. I was scared to death, and I was really sad to say goodbye."
Then he gives his best current answer to the biggest question. For the one, the source, to really know itself, it has to take an infinite number of perspectives. And taking a perspective does not mean casually trying it on. "It jumps in with both feet, all in, on that perspective, to really believe it's that." It lives it out, and slowly realizes it was just a perspective, and wakes up from it. That is how it learns that the perspective, as rich as it was, is something it infinitely transcends. "That's the best story I can tell right now about how the one knows itself, and why it does this."
"I think it's perfect," AJ says.
The monarch butterfly
AJ's last question. Given everything Hoffman has done and everything he knows, when he sees a monarch butterfly, is it more beautiful, or less?
"It's very beautiful." There is a preschool just down the street from his house with a big poster of a child looking at a monarch butterfly. "And every time I go by, I go: that was me at five years old."
His message: don't lose the love of surprise, and the joy of discovering all the beautiful stuff around you right now that you are just not seeing. He loves that picture of a boy looking at a butterfly. "It's there for me, reminding me: relax, Don. Don't get uptight."
AJ: "It's just a game." "That's the message."
AJ thanks him. Hoffman has a flight to catch; AJ hopes he will come back and calls him a treasure. "I'd love to," Hoffman says. "This has been a great pleasure, AJ."
AJ's breakdown
As promised, AJ comes back alone.
"That was Donald Hoffman. He ended on a butterfly. So let me start with the other bug." The beer bottle beetle is real. The entomologists Darryl Gwynne and David Rentz watched males in the outback mating with stubbies until they died, and their 1983 paper won a prize. (The captions say "a Nobel Prize"; it was the 2011 Ig Nobel Prize in Biology, for "Beetles on the Bottle: Male Buprestids Mistake Stubbies for Females.") Australia changed the bottle for the beetle. "Evolution built them to like dimples, not truth. That's Don's whole argument in one insect. And I love it."
Now the new math. Don says proofs of special and general relativity, built from consciousness alone, are sitting on his desk. "Nothing is published, so there's nothing to check. He said it himself: it's not real until it's published. I can wait."
In the hospital, Don's heart raced out of control for 30 hours. He texted his wife goodbye and lay back to die. Two strangers stood by the bed at three in the morning, neither in hospital scrubs. One smiled and nodded, and his heart flipped back to normal. "He never told that story publicly before. I had never heard it."
"Forty years of math to prove reality is a headset. And the takeaway he cares about most is a preschool poster of a kid staring at a butterfly. That part I can verify. I watched him mean it."
Don's book is The Case Against Reality, and the new work is at traceinstitute.org. "Until next time, be safe, be kind, and know that you are appreciated." The episode closes over The Why Files theme song, a rock number about loving UFOs and paranormal fun, conspiracy theories that turn out true, a faked moon landing on a Kubrick film set, Project Stargate and the Dark Watchers, "we're in a simulation, don't you worry," and a chorus about dancing with the fish on Thursday night, because all anyone ever wanted was to hear the truth.
Key takeaways
The question behind the whole career is from age 17: are we just machines? Hoffman attacked it from both sides at once, what machines can do (the MIT AI lab, machine vision, Hughes Aircraft) and what humans can do (neuroscience under David Marr).
Physicalism is batting zero, by his count. Thirty to forty years of brilliant work has not produced a theory that explains one specific conscious experience, the taste of chocolate versus mint. Integrated information theory says the answer is a Markov matrix and, he says, has never written one down. Illusionism owes the same derivation for the illusion.
Simulation theory is not his view. He agrees with Bostrom that spacetime is not final, but rejects a physical spacetime at the bottom of the stack and rejects computation as the source of experience.
Fitness beats truth is a one line argument about payoff functions. Evolutionary game theory, as it stands, allows any payoff function. Of all payoff functions, the fraction that are homomorphisms of the world's structure (a subway map faithful to the subway) is zero. So the probability evolution shaped us to see the truth is zero. The jewel beetle and red lipstick are illustrations; the theorem is the argument.
Using Darwin to undercut Darwin's assumptions is not a logical error, he says, it is what great theories do. Quantum field theory assumes spacetime and then shows spacetime has no operational meaning at the Planck scale. No theory explains its own assumptions, so there is no theory of everything, ever. "We are 0%."
His model of an observer assumes only that there are experiences and they change. The most general math for that is a Markov matrix. All possible observers are all possible Markov matrices, of any size, up to infinity.
Agency is recursion. A policy is a Markov matrix on observer windows, a meta policy is one on policies, and so on without end, giving degrees of agency that science can master one level at a time.
A trace is what a smaller window sees of a bigger chain. Shut off seven of ten colors and the three you can see still carry the hidden dynamics. Hoffman noticed that "is a trace of" is a partial order, a non Boolean logic on all Markov chains containing infinitely many Boolean sublogics. Chetan Prakash proved the transitivity (a trace of a trace is a trace) at Heathrow. Hoffman proposes it as Leibniz's pre established harmony.
Time dilation, in his intuition, is a counter mismatch. Every chain clicks a counter per transition. An observer who cannot see some of another's states clicks less often, like missing the yellow at a traffic light and counting two clicks for every three. He says tentative proofs of special and general relativity are with his mathematicians and unpublished, and that "it's not real until it's published."
Spacetime is a headset the math allows, not one it forces, and ours is "the training wheels version," one of the most trivial of an unbounded number. The brain is "a headset representation of how the headset is constructed," so he wants more neuroscience, not less.
The Trace Institute has two to three years to prove its conjectures, among them Minkowski space, general relativity, the Born rule, the Big Bang, nonlocality and all of quantum field theory. "If we can't, then we're wrong."
He concedes the math works without the word consciousness. He keeps it because a hammer on the thumb and the taste of chocolate are the most certainly real things he knows.
DMT, UAPs and entities become questions about headsets. Are DMT entities genuine avatars that can pass information between two people? Is DMT a tool from outside spacetime that changes a headset parameter (Niffe Hermansson reports a rigidly rotating tesseract)? Are UAPs the geek who wrote Grand Theft Auto, using exits and re entries in the trace structure, or simple misdirection of attention?
The closing is personal. After 30 hours at 190 beats a minute he texted goodbye, lay back to die, saw two men who were not in hospital clothes, felt warmth in his heart, and his rhythm returned to normal. He does not know what to make of it. The lesson he draws is the butterfly one: relax, let go of the ego, and play the game.
Chapters
0:00:00 A Preacher's Son, MIT, and Coding Fighter Jets at Hughes Aircraft
0:13:20 The Butterflies That Made Him Late for Kindergarten
0:20:13 David Marr, Francis Crick, and the Secret Helmholtz Club
0:34:20 Why Neuroscience and Simulation Theory Can't Explain Consciousness
0:46:44 The Jewel Beetle and the Beer Bottle: Why Evolution Hides Reality
0:53:04 Fitness Beats Truth, and Why There's No Theory of Everything
1:05:52 Conscious Agents, Markov Chains, and the Traffic Light
1:16:04 Trace Logic: The Theorem Chetan Prakash Proved at Heathrow
1:26:01 From Markov Chains to Einstein's Time Dilation
1:38:34 The Headset: Ours Is the Cheapest Version of Reality
1:45:43 The Trace Institute's Nine Conjectures, Ants, and Hands
1:53:43 Are DMT Entities Real? Avatars and a Rotating Tesseract
2:02:45 UAPs, Grand Theft Auto, and Hacking the Software of Reality
2:10:05 The Goodbye Text and the Two Figures at His Hospital Bed
2:13:36 Just for Laughs, the Ego, and the Monarch Butterfly
2:17:30 Outro
Notable quotes
"Are we just machines? That's my question."
Donald Hoffman, on the question he chose at 17 (chapter at 0:00:00)
"If I wanted to multiply by two, I didn't multiply by two. I shifted."
Hoffman, on hand coding cockpit displays in machine code at Hughes (0:00:00)
"It's just obvious that the right thing to do is to relax and play. Relax and enjoy and explore."
Hoffman, on the butterfly bush at age five (0:13:20)
"Everything I've learned is 0%."
Hoffman, on how he tries to hold what he knows (0:13:20)
"Look, if you have any feelings about doing that, life is worth living. Come talk with me first."
David Marr, dying of leukemia, to Hoffman the day after a fellow student's suicide, as Hoffman recalls it (0:20:13)
"He had demystified life with DNA and he wanted to demystify consciousness."
Hoffman, on Francis Crick and the Helmholtz Club (0:20:13)
"There is no physicalist, neuroscience, AI, computational theory, none of them, that can explain even one specific conscious experience."
Hoffman (0:34:20)
"Just moving from consciousness to the illusion of consciousness doesn't get you off the hook."
Hoffman, on illusionism (0:34:20)
"A female is anything dimpled, glossy and brown. The bigger the better."
Hoffman, on the male jewel beetle's model of a mate (0:46:44)
"The set of payoff functions is big. The ones that are homomorphisms, probability zero. Therefore we don't see the truth. Just that simple."
Hoffman, the whole argument "in one 30 second clip" (0:53:04)
"Every scientific theory starts with assumptions. No scientific theory is proving its assumptions. Those are the miracles of the theory."
Hoffman (0:53:04)
"My own view is we are 0%. Always 0% of a theory of everything."
Hoffman (0:53:04)
"There are experiences and they change. That's the minimum I could imagine."
Hoffman, on the starting point of his model (1:05:52)
"Don, that's too pretty to be true."
Chetan Prakash, before proving the trace relation is transitive, as Hoffman tells it (1:16:04)
"The math never tells you what it can be applied to."
Hoffman, answering the "rookie mistake" critique (1:16:04)
"One clock is only going at two thirds the speed of the other clock. And that is where we're going to get Einstein's time dilation."
Hoffman, on the traffic light counters (1:26:01)
"That's a fairly shallow data structure."
Hoffman, on spacetime falling apart at 10 to the minus 33 centimeters (1:26:01)
"I would be bitterly disappointed if it forced me to build those structures."
Hoffman, on wanting the math to allow spacetime rather than force it (1:38:34)
"We have the training wheels version, really dumbed down."
Hoffman, on our spacetime headset (1:38:34)
"If we can't, then we're wrong, by the way."
Hoffman, on the Trace Institute's conjectures (1:45:43)
"If anything is real, that painful thumb is real."
Hoffman, on why he keeps the word consciousness (1:45:43)
"The brain is a headset representation of how the headset is constructed."
Hoffman (1:53:43)
"The geek who wrote the code can take the gas out of the car of the wizard."
Hoffman, on Grand Theft Auto (1:53:43)
"The geek is not stuck in the headset. The geek is making the headset."
Hoffman, on UAPs (2:02:45)
"I don't think I'm going to make it. I love you. Goodbye."
The text Hoffman sent his wife from the hospital (2:10:05)
"I'm just stating what I saw, and I'm here."
Hoffman, on the two figures at his bed (2:10:05)
"The ego is the biggest, biggest impediment to growth and discovery."
Hoffman (2:13:36)
"Relax, Don. Don't get uptight."
Hoffman, on what the preschool's butterfly poster says to him (2:13:36)
"Nothing is published, so there's nothing to check. He said it himself. It's not real until it's published. I can wait."
AJ Gentile, in his breakdown (2:17:30)
Resources mentioned
The people in the room
Donald D. Hoffman, Professor Emeritus of Cognitive Sciences at UC Irvine and founder of the Trace Institute (Wikipedia)
Natural selection and veridical perceptions, Mark, Marion and Hoffman, Journal of Theoretical Biology, 2010, the earlier evolutionary game simulations (background, not discussed in the episode)
Traces of the Other: Are DMT Entities Real?, the public launch of the Hoffman and Gallimore DMT project mentioned here, with the fullest technical walk through of the qualia kernel, traces and the time dilation outline
We Live in a Simulation, with Tom Bilyeu, Hoffman's longer earlier case for spacetime as a headset, fitness beats truth, and why he is not a simulation theorist in Bostrom's sense
The Basement: Thomas Campbell, the same interview room with a physicist who does think reality is a computed virtual reality, a useful contrast to Hoffman's "physicalism through the back door" critique
Where it stands
This section is the only place on the page that steps outside Hoffman's frame, and it separates three kinds of claim he makes, sometimes in the same breath: published mathematics, conjecture, and personal experience.
Published and checkable. The fitness beats truth result is a real, peer reviewed theorem (Prakash and colleagues, 2021), and the conscious agent formalism has been in print since 2014. The trace logic result is in a preprint. The transitivity at its core is, as Hoffman himself says, elementary for a mathematician; the novelty he claims is reading the trace relation as a logic over all observers. What is contested is not whether these theorems follow from their premises but what they license. On fitness beats truth, critics argue that "a randomly chosen payoff function" is the wrong question: real payoffs are not drawn at random from all functions, and which measure you put on an infinite space of functions largely decides what "probability zero" means. The Yale simulations AJ raised found that when an organism must serve many independent goals with one perceptual system, the gap between interface and veridical perception closes. Hoffman's reply in this episode is that current evolutionary theory places no principled restriction on payoffs, and that particular cases do not change the general picture; that is a fair statement of his position, not a resolution of the dispute.
Conjecture. Everything from spacetime onward is, by his own account, unfinished. The relativity proofs are unpublished and unchecked, which AJ rightly flags. The traffic light intuition shows how two observers with different visible states will click counters at different rates, but as given it is one directional (the full observer is not missing anything of the blind one), while Einstein's effect is reciprocal and depends on relative velocity through the Lorentz factor; producing exactly that is what the promised proof has to do. The argument that Markov chains are computationally universal cuts both ways: it makes it very likely that known physics can be encoded in them, and for the same reason an encoding alone would not show the framework is the right one; the stronger evidence would be a derivation that is natural rather than engineered, or a new prediction. The Trace Institute's research page currently lists eight physics conjectures rather than the nine Hoffman recalls. AJ's objection that the mathematics works equally well without the word consciousness is one Hoffman concedes outright; his reason for keeping it (the undeniable reality of pain and taste) is a philosophical commitment, not a consequence of the math. His characterization of integrated information theory is also his: IIT's papers do specify transition probability matrices for model substrates, though not for real brains, and whether that meets his demand is exactly the argument between them. A small historical note: Markov's first paper on the chains that bear his name is usually dated 1906 rather than 1905.
Report and experience. The rotating tesseract is a secondhand report of a sub second DMT experience. AJ's week without reading glasses is his own anecdote. The UAP performance figures AJ lists (Mach 40, 466 g, water entry without displacement) come from contested analyses of military encounters and are not established. The hospital story is the most personal thing in the episode, and Hoffman tells it without claiming to know what it means; for what it is worth, new episodes of atrial fibrillation frequently convert to normal rhythm on their own, which is the ordinary medical reading, while his three reasons for doubting a mundane explanation are his to weigh. AJ's closing line is the right calibration for the whole program: the beetle is real, the theorem is published, the new physics is not yet, "and it's not real until it's published."
Full transcript
========================================
Today I'm talking with Donald Hoffman.
Don is a cognitive scientist who spent
decades at the heart of mainstream
science. He wrote fighter jet software
for Hughes aircraft in pure machine
code. He trained at MIT. He sat for
years in a private consciousness club
that Francis Crick co-ounded.
>> Francis Crick, the guy who discovered
the building blocks of life. I
discovered you can't return underwear at
Target. Both took real courage. Then his
own math convinced him that we have
never once seen reality as it [music]
is. He seems very calm about it. Today
we're covering the beetle that fell in
love with a beer bottle and why that
matters for evolution.
>> A beetle dated a beer bottle. Huh?
Life worse.
>> Why Don says space and time are a
headset we're wearing and what that
means for UAPs and the entities people
meet on DMT. Near the end, Don tells a
story that he's never told before.
>> [music]
>> It's about the night he texted his wife
goodbye from a hospital bed. You might
know that part, but you don't know the
rest.
>> Okay, no notes on that one. That one's
real.
>> Once we wrap up, I'll come back and
break down the conversation, which is
not going to be easy, but I'll be here.
Let's go down to the basement.
>> Professor, thanks for doing this. I
appreciate you. Thanks a lot, AJ.
>> Um,
first thing that kind of
was I found interesting about your
background is you grew up in San
Antonio, spent most of your life in
Southern California. Where did you start
ice skating? How did that even how does
that part of your resume?
>> I think that that started when I was
about 12 or 13 and we saw maybe an ice
capes chalet or ice capes show or
something like that. And u my parents
liked it. We we liked it and they
decided to try us ice skating. I was 12
or 13. My brother was a year younger. My
my sister was four years younger. So we
went out there. I didn't like it at
first cuz I you know you just fall down.
You get out there, you try stuff.
Everything that you try to do is just
wrong. All your normal reactions are
wrong. So I fell fell fell. And I
basically didn't want to do it. And and
they sort of forced us to do it for
several weeks. And then all of a sudden
one day it clicks. You realize you don't
walk like you normally walk. Heel toe,
heel toe. You push with a side and you
glide. And when you really discover that
for yourself, it opens up a whole new
world. And once you then enter that
world, then it's it's fun to glide. And
eventually I got to the point where I
could do a a double toe loop and an axle
jump and so forth. So I actually did a
double jump and and a one and a half
turn jump and so forth. So it was it was
a blast and it's very very good exercise
and even now I'll go you know go back
and ice skate really I'm not doing
double jumps but you know you know I'm
70 so it's not not smart but if I'm
careful and just you know just stroke
and glide on the ice it's it's good
exercise yeah
>> I thought it was great okay so ice
capades not hockey that's fine
[laughter]
>> right yeah it was it wasn't it wasn't
hockey it was and and we and so I did
figure skating and all of us did figure
skating I've never actually been on
hockey skates
>> Wow
I'm picturing you in the whole Lyra
outfit. You look great, by the way. Um,
so dad was a fundamentalist preacher.
>> Yes.
>> What do you think he would make of you
starting an institute that nothing is
real?
>> How would that conversation go?
>> Well, in in his later years, he did
>> hear about the work I was doing. He did.
>> We did have conversations about
consciousness being fundamental.
He liked that actually because that
aspect of it is, you know, on board with
his his views. I mean, he he thinks he
knows what that consciousness is and
it's his god and not other people's gods
and so forth. But but so he he sort of
liked the non-physicalist stuff. So so
he was all on board with that. But when
when it wasn't just directly into, you
know, fundamentalist Christianity, then
he wasn't so excited about it. Um, so I
would say that he was glad and and would
be happy about the idea that
consciousness is fundamental and he he
would like it to show that his
denomination of Christianity was the
truth. And and it it doesn't show
something like that. It just shows that
there's a much deeper level of
consciousness that I think the different
religions Christian, Buddhist, Hindu and
so forth are all perspectives on a
deeper consciousness and they all get a
piece of of the puzzle and they miss
them as as as any perspective would.
>> Is this um is that what planted the seed
for you to go into this research was you
were getting one story on Sunday and a
different one on Monday. Does this kind
of reconcile it? That's that's exactly
right. I got one the the story I got is
a pretty
severe story from the the Christian
view. It was the earth is only 4,000
years old and uh
>> Oh, he was that.
>> Oh, no. He was that that's right. And he
had a master's degree in chemistry. So,
it was
>> really he had decided to choose that the
the the religious view over what
chemistry had shown. And so, it was
quite quite stunning. So, he was he was
quite into it. Uh so he he took that
point of view. Um
yeah. So he he wouldn't have liked
me going beyond the 4,000 years which
which which I have. I think you know
that the earth is well over 4,000 years
old. And he wouldn't have liked um I
think he likes that there was a
mathematical model. So he would have
liked that. But um he also wanted it to
come out that his particular brand of
Christianity was the truth and and all
the others weren't the truth and so
forth. So so so yeah it was I got one
version on Sunday but then I also had
you know the Monday version which was
there was all this science and I would
like to um that says that we're billions
of years old and we evolved and he
didn't believe in evolution at all. That
was an aathema to him. The notion of
evolution was just off the table. I
can't square the masters with chemistry
with with that. What what drew him to
faith instead of the science?
>> I
you know I think that
it's hard to explain. It was it seems
irrational to me. The the science is
very very clear. Uh the experiments are
are are quite clear. If you're going to
use the language of space and time and
and chemistry in that framework, the
earth is four billion years old or and
and
[snorts] not 4,000 years old. So, I
think he just chose to reject his ma a
master's degree, not just a bachelor's,
a master's degree in chemistry and and
and working at various companies uh in
high levels for you using chemistry. So,
he so so I had to on my own
then decide between the the spiritual
view that I was getting I'll say the
Christian view particular kind of
spiritual view a Christian view
>> uh and in fact a fundamentalist
Christian view on Sunday
>> so Catholics don't go to heaven then
>> uh well a lot of other people even in
other Christian denominations might not
be going to heaven right
>> okay
>> so this this was um it was it was pretty
austere right my way or the highway and
and and and it was
>> bad thing about it was that But God was
not a really they would talk about God
being loved but but in fact you were
trained to be afraid. You had to be
really really afraid and and cry every
Sunday and repent and and and you know
it was it was it was quite a scene. So
it was real psychological control and it
took when you're raised in it that's all
you know right that's and so it took me
it's taken me decades to um recover from
that.
>> Last week I was stuck at the airport
with a delayed flight. my phone, watch,
and headphones, all dangerously low.
Instead of fighting over one outlet, I
used my Ridge 5 in-1 travel power bank.
It's one device with five [music] ways
to charge. Mag Safe, Apple Watch
charger, lightning, USBC, all built
right in. No separate cords, nothing to
lose, nothing [music] to untangle. And
it's not messing around on power either.
20 watts, charges [music] fast, 10,000
milliamp hours in the tank. So, you're
talking three full phone charges before
you even need an outlet. Slap it right
on the back of your iPhone magnetically
and just let it ride with you while you
charge. This is the last power bank
you'll ever need. Ridge backs it the way
they back everything. Free shipping,
99-day trial, lifetime warranty. One
thing to pack, five ways to power. You
can find Ridg's power bank at Best Buy,
or our listeners can get 10% [music] off
at ridge.com by using code basement at
checkout. Just head to ridge.com and use
code basement and you're all set. One
more time, one thing to pack, five ways
to power. You can find ridges power bank
at Best Buy or our listeners get 10% off
code basement and you're all set. After
you purchase, tell them we sent you.
>> So, I I decided as a teenager around age
17 that I had a specific question that I
would try to answer. Are we just
machines? That's my question. Are we
just machines? And to do that I would
have to understand what machines can do
and is there anything that humans can do
that machines can't? That would that was
sort of the idea. So I ended up
ultimately trying to answer the
question. I went to MIT and I was in the
artificial intelligence lab. So that's
going after what machines can do. So I
started in 1979 and had the good fortune
of um actually interacting with Marv
Minsky. Took a class with Marv Minsky
the guy that one of the guys that
founded the field McCarthy. The two of
them founded the field. So I got to
argue with him at his home. We we had
his that class for a semester in in
Minsky's home and we could just argue
about the foundations of the
philosophical foundations of artificial
intelligence.
>> And so from 79 to 83 I was there in the
AI lab trying to build you can't just
talk in general about AI. You need to
pick a problem and try to solve it as a
scientist. So I picked the problem of
how do we see in 3D? It was machine
vision I went after. How do we build
visual systems that can see in 3D? The
kind of stuff that we now have in
self-driving cars. We were we were
pioneering it back then. So I did that
as my concrete mathematical how do we
see in 3D? And
then I was also in the brain and
cognitive sciences department. Um and
there I was studying human neuroscience.
So that was you can see I was studying
what machines can do in the AI lab and
what human neuroscience I was studying
what humans can do and trying to piece
the two together and it wasn't until
so that was 79 to 83 and I kept working
on the mathematical models of of vision
and it was around in 1986 or 1987
working now at University of California
at Irvine with Bruce Bennett and Jayton
Pash two extremely talented
mathematicians I was very very fortunate
to have them working with me they're
geniuses is and I'm still working with
with Chayon. He's he's he's brilliant.
And
>> let me stop you for a second. Yeah.
While you're getting your PhD, you're
working
at in at Hughes Aircraft on on
>> vision systems for missiles and stuff,
right?
>> Yes, I did that. Well, I when I was an
undergraduate at UCLA from 1976 through
1978 and then then a whole year in 70 78
to 79, I was at Hughes Aircraft
full-time. What did the Hughes guys
think of your work with computational
vision?
>> Uh, they liked it so much that they paid
me the whole way to go through MIT. So,
I got a free ride through MIT from
Hughes.
>> From Hughes.
>> Wow. Okay. I didn't know that.
>> Yeah. And at the time there were no
strings attached. They they assumed that
I would come back and I would be the
director of the Hughes Aircraft
artificial intelligence laboratory in
Malibu. And I was thinking about the
same thing. I was about to be a very
rich guy living in a very nice place.
and uh in Malibu and and directing the
AI lab there and
but my last year there I realized I mean
I I knew what it was like to work at and
by the way Hughes is a great place.
>> Did you work out at Elsagundo?
>> Elsagundo. That's right. Yeah, I worked
out of Elsagundo. I um worked on um
fighter jet cockpit um software. So for
for displays, so this was a new thing.
There were all these, you know, old old
displays, mechanical displays on the
fighter jet cockpits, and it was time to
go digital, but the microprocessors
weren't that fast. And so we had to
program them in machine code. So I and
two or three other guys were the the
coders and we wrote
>> Wow. You didn't even have Forran yet or
nothing like that?
>> Well, I had studied Forran, of course. I
mean, um I I knew for actually I did
that when I was a sophomore, I think,
you know. So I knew how to program in
forran but and but when I got to Hughes
they said you know we can't do forran on
on these is not that I mean these are
special purpose um processors that um
are are interrupt driven and they have
very very special things all what they
had was the machine code that were brand
new so this is 1976 these are brand new
it was they called the anuk 30 an-k
30 so you can look that up that's what I
programmed and it so I knew all the ones
and zeros. If I wanted to multiply by
two, I didn't multiply by two. I shifted
things like that. If you want to divide
by four, you shift to the right. You
have stuff like that. So, you found all
the tricks. We did an entire complete
flight simulator
in that 30 in machine code. And because
we programmed the entire machine code by
hand, we got it all in 64K.
>> Wow. You can't type one word in an email
for 64K. We did an entire flight
simulator bit by bit. We knew all the
bits
>> and and and it we it was delivered to
write Patterson Air Force Base um in
1978.
>> To write Pat. Wow.
>> So to write Pat. I was and I actually as
an undergraduate at UCLA was being flown
around to various um military places and
and corporation places to help you know
because I was one of the few like three
or four people in the world that knew
how to do this kind of stuff. This kind
That's right. It was cutting edge for
these new um um electronic displays for
for Frederick. So I was I was a cold
warrior um as an undergraduate and then
in 1979 as well for a full year and then
uh so I took a year off from from
college and then went to MIT in 79 and
Hughes paid the whole way expected me to
come back and direct the AI lab. But at
the end of toward the end, I realized
that I really I had a choice between
money or doing the research I wanted to
do because I I I knew even though I was
going to be director, there would be
directives from above and I wouldn't be
free to really explore what I wanted to
do. And so I decided I wanted complete
autonomy to explore whatever I wanted to
do, not just something that's going to
lead to a product. And you know, it's
not right or wrong. as some people are
more inclined to do one thing or the
other. I was more inclined to want to be
independent and follow my own. So I
decided I was just I so I took a pay
cut. I I made less money going to the
University of California to Irvine of
course
>> than I was making as an undergraduate
>> of course
>> at Hughes. A huge pay cut. So I I made
maybe a quarter or a fifth of what I
would have made at Hughes. So
>> but it was the good decision because you
could have gone back to Hughes anytime.
>> I could That's right. But um but once I
got going on on the research then then
it then it really took a life of its
own. I realized I really wanted to to
pursue this um all the way.
>> Well, let me go back to young uh young
Don for a minute.
>> And because maybe this is the the seed
of of getting interested in vision, but
the story uh you're 5 years old and
you're late to kindergarten because
something grabbed your attention.
>> Yes. Yes. Uh yeah, that was a a real
wakeup call to me. So I was
>> magic, right? The butterflies.
>> The butterflies. So I was walking to
kindergarten. Uh and I I'd learned how
to walk there and and and
so I was confident that I could get
there. But when I was walking, I saw
this beautiful bush that was in bloom.
It was about my eye level, so I could
really see stuff. And it was covered in
butterflies. And it just was obvious to
me that uh kindergarten was nowhere near
as important as enjoying life. I mean,
this is a miracle right in front of me.
It's just obvious that the right thing
to do is to relax and play. Relax and
enjoy and explore. I mean, nature is
showing you something that's amazing.
You're here to enjoy, to observe, and to
learn. And I that's I mean, I'm saying I
was intellectually saying that way. was
just my emotional childish reaction was
this is a whole thing to explore. So
it's not like I was intellectualizing.
It's just obvious that you to do this.
>> But your decision to leave Hughes is an
analog of that.
>> Yeah. It it was really
life is about exploring
going where your heart wants you to and
having fun. So it's really it's really a
play. It's really and I really do think
that way about reality and that is that
we're we're here. it's not that serious.
It's like the Hindus have the notion of
Leela that it's it's a game. And I think
that that's a really deep insight that
the you know in some sense um the
greatest insights and even technological
insights come when you when you just
play and somehow I think reality rewards
that. It's sort of like only when you
let go of your uptightness
do you start to relax into new
dimensions of reality where that you can
explore. So it's and so for me it's been
a bit of a a journey because that that
5-year-old butterfly chaser um got
strangled
it um I got uh I got in trouble. um the
the kindergarten teacher instead of um
rewarding that let my parents know that
I was late because I I would and and I
was punished.
>> Did you tell your teacher what h what
why you were late?
>> Oh yeah, I brought a butterfly in. I was
trying to show share it with everybody,
>> of course.
>> So and I I didn't really understand that
that was not good for a butterfly. So
you didn't mean any harm.
>> I didn't mean any harm with that. I was
only five, but but it it was just
obvious that this was a a joyful thing
that I should just share it with all of
my my friends in kindergarten. That's
just of course. And to find out that
that wasn't accepted was a And so you
get slapped down. I got slapped down. Um
>> that's how you create a heretic right
there that day.
>> Yeah. eventually. I mean, I I I I
was shaped into that mold for for until
I was able to leave 17 18 years old and
then it it those are the formative
years. So, unforming the formative takes
a so I'm still in the process of
decompressing from it. But in the
decompression process, I I really have
come to the point of view that I was
trained to be afraid and to just go by
the rules and so forth. And now I'm
realizing that true intelligence is
play. True intelligence is not assume.
It's assuming that I don't know
anything.
>> Something interesting is happening with
gold and silver right now. And it's
worth paying attention to who's buying.
Central banks just bought a record $289
tons of gold last quarter. [music] JP
Morgan is telling clients gold could
reach $6,000 an ounce, while City Bank
predicts silver could hit $90. At the
same time, the national debt is
approaching $40 trillion. And Treasury
officials is talking about growing our
way out of it. So, while the people
running the monetary system [music] are
stockpiling physical metal, it might be
worth asking why. That's where Gold
Crest Metals comes in. They offer widely
recognized gold and silver coins with
transparent pricing, zero buyback fees,
and free insured shipping. You also get
a real person on the phone, not a call
center. Right now, Goldenest will send
you their free 2026 precious metals info
kit with no obligation and no sales call
required. Or you can schedule a free, no
pressure, portfolio review. Plus,
qualifying purchases receive up to
$25,000 [music]
of free silver. Go to
yfileslovesgold.com
right now or text files to medals
638-257.
That's yfilesloves.com
or text files to metals. The people
running the system are already getting
in position. Maybe you should too. That
everything of interest is still ahead of
me. And to just open up and say as much
as I know right now, it's trivial.
Everything I've learned is 0%.
drop it to the moment you get something
deeper. And so that's I'm I'm moving
back to that. But but the interesting
thing is that the emotional programming
goes very very deep. So I see I I see um
in meditation I just see myself having
to face that old programming um of of
follow the rules, God's God's mad and so
forth and and buckle under and
>> and you could observe that when you're
meditating. Yeah, I see the pain of that
and I just and all the trapped. So,
really it's it's a closing up. What that
does is it closes you up to reality. You
you're you're tightened up. So, what I
find is as I'm opening up to reality,
all that tied up energy flows out and
and the old tight, you know, tied up
person is afraid of that. So, that's
that's why it's painful is the that that
old person is dying. It's really a death
of that old. It is
>> and and it is a birth of of someone
who's ready to play to relax into life
and to then have intelligent exploring
because that's only only then can you
get outside of your box. The only way to
make something new is to let go of the
old and you have to be able to let go of
everything that you know or sometimes to
to take what you know and leverage it
but to go into new places. you have to
go into new places. So, so it's not like
old knowledge is bad. Often old
knowledge can take you to the frontiers
and then you have to once you're at that
new frontier, then you have to let go of
the old knowledge. So, it did its thing.
It got you to a new frontier and then
you have to dive in to the unknown. And
so, that's that's fun for me.
>> I think I think we've all seen you live
this because my audience has followed
your work since the TED talk. We've
never seen you lose your temper
>> in a debate. I mean, you've released
papers that have been have had 10
rebuttals and you've said maybe they're
right.
>> I mean, it's just you're just very open
to criticism.
>> Yes. And I think that most scientists
would agree that that's the the way to
to do science. I think that's I mean, we
are humans and so some people get
rattled and and they they get personal
about it. But I would say that most
scientists when when they're, you know,
not emotional, they're just
dispassionate and looking at science,
they would say theories are just
theories. Don't be identified with them.
Of course, do your best. Be as clean and
as rigorous as you possibly can. Get all
the evidence you can. give as strong a
presentation of your idea as possible.
But then I I have so many good friends
who disagree with me and they're
brilliant and will disagree and it does
not have to be ad hom and it's never ad
hom on my part because these are
brilliant people. Why would I want to
get upset with them? They disagree with
me often even if I think they're wrong.
They've pushed me in a new direction
that helps me think out of the box for
for my own stuff. And often sometimes
I'm wrong and they they point out
something where I'm wrong and hey, you
know, good to drop it as soon as you
can. So yeah,
>> it's um funny you mentioned theories are
just theories. It is something I wanted
to talk to you about is um is David
Maher, the great
>> Yes. Yes.
>> Um I guess he basically launched
computational neur neuroscience.
>> Um died very young. You got to study
under him at MIT.
He said um unless there's math behind
it, a theory is just an theory.
>> Yeah.
>> But he also said that
vi vision is designed to bring you the
truth.
>> Yes, he did.
>> How so? You've taken that pre that
premise and then used your theory to
kind of to knock that down. So just from
a relationship perspective, how do you
think that argument would go between you
and David?
>> Well, so David changed my life. I was a
senior at UCLA taking a class on
artificial intelligence and one of the
papers we read was his paper a paper
with him and I think Tommy Pojo on on
vision and as soon as I saw that paper I
was electrified. I realized that's it.
This is really rigorous. They're doing
neuroscience. They're doing mathematical
models and they're not waving their
hands. They're saying we need to build a
working system and if it doesn't work
then we're wrong. and sort of like there
is no nonsense here. This is all serious
stuff. This is how you make progress. I
said, "Where is this guy?" And I so I
looked at and found out he was in what
was then called the psychology
department at MIT,
>> right? They didn't have anything.
>> I didn't even think of psychology in
MIT. I'm thinking engineering and math
and so forth. So, but there it's now
called the brain and cognitive science
department. But and then he was also in
the artificial intelligence lab. I
thought, well, wow,
this guy's got it. Um, it's a long shot,
but I need to see if I can get in there.
So, so I applied and uh went out there.
I'd never been to the East Coast before.
And I had no idea how cold it was. It
was like February. And I I brought just
a a light jacket.
>> It was [laughter]
one trial learning.
But David, so they accepted me. So
David,
>> take me to that day because I've been to
MIT. Were you just blown away when you
got on the campus and saw these
legendary buildings, these names?
>> Well, it's a who's who. I mean, yes. I
mean, there's n I took a class with Nam
Chsky.
>> Jerry Foder was there. Thomas was
in the class.
>> I sat in the class given by Thomas
But there's this one class where Jerry
Foder and Nam Chosky were the were the
instructors and Thomas was in it.
and and the the graduate students ended
up being a who's who in in the field. It
was then there like 50 or 60 graduate
students there. It was
>> truly a a stunning situation. Walin
natada. I took a class with Walinatada.
I mean you mean it was
>> what an experience.
>> I I had no idea how lucky I I was.
>> You didn't know at the time?
>> No. I I all I knew was David Maher was
there.
>> Okay.
>> And and and um Whitman Richards was also
my co-advisor. Um they they knew when I
came there that David was sick.
>> You finished underw.
>> So yeah. So I had David for I don't know
maybe a year and a half or something
like that that I was able to work with
him. Um
he had leukemia.
>> And then Whitman Richard was an absolute
gem. He encouraged me to think out of
the box. He treat treated me as an equal
and we bounced ideas back and forth. We
of course didn't hold any punches but it
was all very very friendly. He really
taught me how to be a gentleman and yet
a researcher that doesn't pull any
punches. And and with David, he had
assembled such an incredible group of
people around him. Bertold Horn and and
>> Eric Grimson and Ellen Hilddrth and and
and and many many more. John Hollerback.
>> Um so so I got to take a class with Bert
Hold Horn. It was it was truly stunning
and it was tragic to see David die o
over those 18 months or so that I was
there with him. And and
>> he kept teaching while he was sick.
>> He he he did he would come to our
research meetings. He would have to hold
something over his mouth because he was
bleeding.
>> Oh, this is heartbreaking.
He and you know at one point there was
um a time when one of the graduate
students I I won't mention any name but
he I I I knew him not well but I I
certainly knew him. He he defended his
PhD and then committed suicide the very
next day. And so that was it was a real
shock to all of his graduate students. I
was like this guy worked so hard and why
did he commit suicide the day after he
got his PhD. So I was in the artificial
intelligence lab going to one of the
list machines probably to do my research
and David Mar saw me he ushered me into
his office and waved me in and and said
you know he knew about of course this
guy that had just committed suicide the
day before and he said look um
if you have any feelings about doing
that life is worth living. Come talk
with me first. And and this was him at
the at the edge of death himself. He he
was facing and he was only he died when
he was 35.
>> It's it's it's it's tragic. I mean the
guy was a complete genius.
>> When when when I when you sat in a room
with him,
>> he was the intellectual leader of the
room. He was the one who knew the
neuroscience. His PhD was a model of the
cerebellum and and but he knew the the
AI and and and he was he commanded the
room. Everybody just looked up to David.
So I was very reluct I was his last
student, his last cohort. So there were
I think one one or two of us in the last
cohort.
>> Imagine what he could accomplish with
another 30 years of work.
>> Oh, can you imagine? Oh [sighs]
yeah. No, no, it was so it really makes
me um I feel very very lucky to have to
have known him. He got me into the
field. If it weren't for David, I would
not have gone to MIT. would have gone to
UCLA or something like that which was
UCLA is a great school and nothing wrong
with UCLA but but that was MIT at that
point was unique in the world that AI
>> really took off there so
>> and um so you're learning how to argue
tell me about the um the Helmholtz Club
>> well so the David died
>> Mhm. and I graduated
um a year or two year a year year and a
half later I guess maybe two years later
and that's when I took the job at UC
Irvine instead of going to Hughes and
when David was dying um Francis Crick
>> with Watson and Crick the guy that did
the Nobel Prize work on
>> I've heard of him
>> DNA everybody's heard of him he was at
uh UC San well at the Saw Institute by
UC San Diego
>> and um so he he knew David quite Well,
and it tried to save David's life. He
was Francis was using all of his
connections which were substantial to
try to get the latest technology,
medical technology to try to save David.
So, David got the best that was
available at the time. Clearly the right
thing to do because he was inventing
modern vision science. Sure. He he
reinvented the whole field.
>> Um, so
so Francis knew David quite well. and he
knew that I was David's student and when
I came to UC Irvine he almost
immediately reached out and invited me
to visit him down at the Sulkq
Institute. So I went down and spent some
time wis and and we talked.
>> Um I I didn't know at the time that he
was interested in consciousness and at
the time it wasn't really kosher to talk
about consciousness, right?
>> So this was 83 84. um it wouldn't be
another six or seven years before it was
kosher because Francis said it was
kosher but he hadn't really come out and
said it was kosher in a big big way yet.
So but
>> had Wheeler talked about consciousness
at this point yet?
>> Um Wheeler had talked about observers
>> but but he didn't talk about it from bit
yet. Right.
>> That was 1989.
>> So just after
>> that 1989 was his it from bit paper.
Right. Right.
He may have talked a little bit about
observers and and the fundamental nature
of observers before that paper, but that
89 paper was is the one that everybody
knows about.
>> So, so Francis
then invited me to be part of this
Helmholds Club. It was a private group
that happened to to meet at UC Irvine.
And the reason it met at UC Irvine
instead of where Francis was is because
we it is sort of in the center of
Southern California. So, there are
universities north of Irvine. So like
USC and UCLA there are universities
south of it like UC San Diego and the
Sulk Institute and so forth and Irvine
is right in the middle. So I was lucky I
was right in the center and so everybody
came to me to to to UC Irvine and so
they met actually literally a
five-minute walk from my house because I
lived on campus at UC Irvine.
>> You could walk to the club.
>> I could walk to the club. So everybody
else had to drive. So, so it was a
secret club, not for any nefarious
reasons, but simply because Francis was
there, and if anybody else, if anybody
knew publicly that Francis was on
campus, we wouldn't get any work done,
right? People would want autographs, you
know, he he was as big a scientist as
they come.
>> Mhm. And so we met um in private at the
University Club at on Tuesdays, one
Tuesday every month at 1:00 and we would
have lunch together. And Francis and and
um there were maybe 10 or 12 of us that
were the core members of of of the club.
And we could invite one person ourselves
if we wanted to, but we as a group would
invite someone two people from anywhere
in the world whose work was of interest
to Francis and and the group. We'd fly
them in. So there'd be two different
speakers each time, but they have lunch
with us. We'd then grill them all
afternoon. It was it was no holds
barred. They would present our stuff and
they couldn't get through their talks.
We would Fran Francis would we would
just go at it.
>> What was he like when he when he
disagreed? Well, well, Fran Francis was
a gentleman all the time. Uh, he was
absolutely a gentleman. I never saw him
impolite, but what I did see was he
would never count as fools and he would
never he he wanted answers.
Life is short. He was he was an older
man at the time.
>> Right. I don't have much time. Let's get
Let's get to it.
>> That's right. And and the goal was
consciousness. He had demystified life
with DNA and he wanted to demystify
consciousness. That was the clear goal
of the Helmolds Club. He wanted to know
the latest neuroscience because he
wanted to understand what neuroscience
would break open the door to
consciousness just like the double helix
broke open the door to understanding
life. So he he was on a mission. He was
polite. He was a gentleman. But he was
not going to have anybody get in his
way. He I mean if he had a question it
was going to come out and we were going
to you we were going to go after. So
this is really bring in the best and
brightest and and really grill them in a
in a respectful way but but we're trying
to understand the best they know to see
if there's some clue there. And it was
fun because there would be you know
maybe a dozen of us 15 of us in the room
all
>> I mean to be fly on that wall. Yeah.
Sharp. It was it was unbelievable the
the conversations that that went on
there and and it was really a good thing
for me to see how Francis was focused.
He had a goal. He was looking for any
clue anywhere and he was bringing in
people from all over the place trying to
go after it. So, and I remember in 1992
he then we at one of the meetings he he
told me about his book he was writing
the astonishing hypothesis. So I got to
hear him talk about that and we talked
talked about that book. So that's when
he sort of
>> you know made it really official that uh
you know serious scientists can talk
about consciousness. Now it was a very
physicalist approach right? So he's he
was saying that somehow neural activity
is going to be the cause of conscious
experience or some aspect of
neuroscience is going to be the cause of
conscious experience. So in kind of
inverted from your hypothesis,
>> completely inverted from mine, but but I
would say in line with what 99% of my
colleagues in neuroscience and and
computer science would say.
>> Well, it's a lot easier to as a
scientist to wrap your mind around that
than
what is consciousness creating.
>> That that's right. And and there's good
reason to go this direction because the
physicalist approach since Galileo at
least had done quite well over
mathematics and a physicalist ideology
had worked very very well. Spiritual
ideas have been around for thousands of
years. All of our technology came from
physicalism with mathematically precise
models. And so there's no reason until
the spiritual traditions can come up
with their own mathematical model.
There's there's no beef for a scientist
to go after unless the scientist is
going to try to take the ideas of the
spiritual traditions and turn them into
math. But the bets were
we got the secret to life physically.
A CG and T.
>> Yeah,
>> we got the secret to life. The the
story. I'm not saying this is right. I'm
just saying this was the story. We got
the secret of life ACG&T from biology
and from the mathematics of that and and
the bets were that we would get the
secret of consciousness the same way.
There's got probably some kind of
nervous system process a neural process
at some level we have to figure the
level and how big a system is it and so
forth that that causes consciousness. So
that's what we were after and it was fun
for me to see the pros go after it. Uh
and but you know we never got it
>> never got it
>> and to this day
>> still the hard problem of consciousness
>> 30 40 years of of now good hard
neuroscience artificial intelligence
computer science information theoretic
attempts to start
>> Penrose and Hammer are involved
>> Penrose and Hammer and and I know most
of the players they're they're brilliant
they're my friends and colleagues And
the the fact is there are several
theories out there and there are
trillions of conscious experiences to
explain and there is no physicalist
neuroscience AI computational theory
none of them that can explain even one
specific conscious experience and that's
truly stunning like the taste of
chocolate or or the the smell of
>> smell of mint
>> mint or something like that. I mean
there's there's just and and that's
that's pretty stunning. I mean, I mean,
just to really put that in perspective,
I'm I'm a cognitive neuroscience and
scientists. And suppose I went to a
bunch of physicists and said, I've got a
new theory of particle physics.
>> Imagine they [laughter] said, "Oh,
really?" They they you know, they smile
and so and a natural question would be,
"So, Don, h you've got a theory. So,
well, let's check this out. So, what
specific particle interaction does your
theory explain? And how does it do it?"
like photon electron interactions. What
what is it? And if I said to them, oh
no, I have a general theory of particle
interactions. I can't explain any
specific particle interaction,
would I be taken seriously? I'll be I'll
be kicked out of the say come back. You
know, they might pat me on the back,
say, "Let's have a beer and then go your
way, man." Uh, you know, they might be
kind to me, but but they're not going to
take me seriously. And that's where we
are wi with with neuroscience and and
physicalist approaches to consciousness
and AI and and so forth. There's not a
single specific conscious experience.
And some people will say, "Oh, well, and
I actually said this two or three weeks
ago at a at a meeting in Venice." Okay,
here's something most guys don't know.
Average male testosterone has been
dropping by about 1% a year since 1980.
So, if you're 35 right now, you're not
just getting older. You're starting from
a lower baseline than your dad was at
the same age. And that decline doesn't
really announce itself. It's just energy
that's a little less consistent, focus
that's a little bit foggier, workouts
that don't hit the same way they used
to. Most guys just blame being busy or
stressed. But honestly, the baseline
moved on them. That's what got me
interested in Mars men. It's built
around ingredients actually studied for
supporting healthy testosterone
naturally. clinically dosed to an ali
and boron plus chiligit, fenugreek,
vitamin D, and zinc. And lots of guys
are saying they're feeling more
consistent energy, sharper focus, better
recovery. Not a caffeine spike, more of
a steady drive throughout the day. Plus,
a 90-day money back guarantee, so
there's zero risk trying [music] it. For
a limited time, our listeners get 50%
off for life, plus free shipping and
three free gifts at mengoarss.com.
That's mango tomars.com for 50% off and
three free gifts when you check out.
After you purchase, they're going to ask
you where you heard about them. Please
support the show and tell them we sent
you. And one person who I is a good
friend of mine and who I deeply respect,
so I'm not going to put up the name
said, well, we basically said we might
have one there there is um space in
integrated information theory um might
have a theory of of of space.
Now, first that I felt that that made
the point. There's trillions of
conscious experiences and we might have
one.
>> So that that still still makes the
point, but but that one the the theory
that he was pointing to is integrated
information theory and Toni and Ko.
>> Mhm.
>> And that theory makes very very clear
that
they're saying that it's certain kinds
of causal structures that give rise to
specific conscious experiences. And the
way to write down rigorously what the
causal structure is in any case is to
write down a markov matrix. That's a
matrix. Um you could say a square
matrix. Um it has n columns and n rows
and it basically has a probability
measure in each row. Technical thing.
And so that's a way to do this.
>> Later we'll do the the traffic light if
you don't mind.
>> Sure. Yeah, we can do that. That's
right. Be a lot of fun. So, so it's so
they say for every experience, every
conscious experience, there is a causal
structure and you will know that you've
got the causal structure when you write
down a markoff matrix for it. And and so
when you look at the paper that co
published, there is no markoff matrix.
So what I would want is okay, what is
the matrix? It has how many rows? Is it
like 100 rows? Okay, and it's got a 100
columns. Y
>> that means you have 10,000 numbers.
>> Mhm. What are those numbers and why and
why is it that those numbers must be the
causal structure that must give rise to
the taste of mint or whatever it might
be or or to in this case you know the
perception of space. So there's nothing
on the table and of course not and there
there never will be.
>> There never will be.
>> There there never will be. And it's so
again I know the players. They're my
friends. They're they're brilliant.
>> They're not dumb. They're absolutely
brilliant and they're doing by the way
they're doing good work. They the work
they're doing even though they're not
solving the problem of consciousness.
They're they're getting a lot of good
information about neuroscience and the
structure of neuroscience and and so
forth. So so it's not like they're the
work is wasted. We're learning lots of
stuff. We're just not learning about
consciousness except what we're learning
is it doesn't come out of neurons or
physical structures. We're really
learning that very very well by geniuses
trying their best to get consciousness
out of physical structures or
computational structures
and failing.
>> I think it's I think it's better that
all these scientists disagree because
truth is not a truth until it's known.
Right.
>> That's that's really true. So this gives
us a real acid test. If we're going to
try to show that consciousness is
fundamental,
it's good to show by the failure of
geniuses that the physicalist approach,
you just show that that won't work. And
but then also they're going to be my
harshest critics,
>> which is what we need. I mean, this is
not about, you know, patting people on
the back and say, "Oh, they're there,
there." No, no. This is hard-nosed
stuff. If you're going to come up with a
new theory, you've got to take you've
got to take whatever comes your way.
you've got to listen to their arguments
and you've got to respond. So if I'm
going to put a theory forward in which
consciousness is fundamental, of course
the the physicists are going to come
after me and that's as it should be. And
then of course we'll go out and have a
beer afterwards. So this it's all
nothing personal. But but but on the
other hand, professionally, no holds
barred. You they should and they do do
everything they can to take down. And
often I learn something from them. I'll
say, "Oh, wow. Didn't think about that.
I need to go look at that. So, it's all
good. It's all good.
>> It really is. Uh physicalism, a lot of
um I probably most of the audience knows
you from the TED talk which was titled
something like Don Hoffman proves
there's a simulation, but that's not
exactly right because simulation is
still physicalism through the back door,
isn't it?
>> Right. Right. So that Nick Bostonramm is
is well known for the simulation theory
>> and in Bostonramm's approach
we're
quite likely in a simulation
>> and that simulation is probably being
written by some teenager at a lower
level um who's got their own computer
and is and and has coded us up and but
that teenager and their computer is
probably a simulation from an even lower
level and you keep going down until
you're going hit some bottom. So, and so
there's two things that I disagree with
on Boston. First, I like the idea of a
simulation in general, right? That this
the aspect that we're not seeing reality
as it is, that this isn't that space and
time aren't the final reality. I like
that aspect of it.
>> But there's two aspects that I disagree
with. The first is at the very very
bottom level. The assumption is by
Boston and most everybody that it's
again some kind of physicalist
space-time reality at the bottom. And I
think that that's not going to work. I
think that spacetime is doomed and we
can talk about why spacetime is doomed.
Yeah.
>> So, so that's one place where I
disagree. The bottom is not physical.
It's not a spaceime. The second is that
to the extent that Bostonramm and others
think that there are conscious
experiences at our level for example
those conscious experiences must result
from the programming at a lower level
you be the computational aspect of it or
physical or what something about the
physical or comput but since it's a
simulation it's really only
computational
>> right it's not really consciousness is
it
>> well and the thing is I wouldn't want to
say it's not really consciousness I
would just say we have no theory yet
that could explain how they could
happen. We have there this is one of
those theories that you know
computational theories for building
consciousness have yet to explain a
single conscious experience and I think
it's principled. The failure again is
principled. So I think that Bostonramm
is wrong to say that the bottom level is
a space-time level if he says that. If
he doesn't say that, I'd be interested
to see what he does say. And I I think
it's wrong to say that computational
systems can give rise to conscious
experiences. Or I would, by the way,
some people like Michael Graziano would
say, "Oh, well, well, there are no
conscious experiences. There's only the
illusion of conscious experiences."
Okay,
>> that's unfalsifiable. Well, well, it's
still well someone who says the illusion
of conscious experiences as a scientist,
I would then want you to give me and you
want to say so computational systems can
give rise to the illusion of conscious
experiences, not not real conscious
experiences.
I'm happy to go with that game. I mean,
that's the that's what we always do in
science. We make hypotheses. But if
you're going to play that game, then to
do the scientific thing, you now owe me
a scientific account. What specific
computational system for example must be
the illusion of the taste of chocolate
and could not be the illusion of the
taste of mint? Now there are no
candidates put on the table so far. So
the idea of saying that it there's not
consciousness there's the illusion of
consciousness does not get you off the
hook for giving me specific examples and
there are zero. So the illusion
hypothesis has no empirical evidence for
it whatsoever and I think that it never
will. So so just moving from
consciousness to the illusion of
consciousness doesn't get you off the
hook and you have to put up the
scientific experiments and the
scientific theories for specific cases
and there are none and I predict that
there will never be any. So illusionism
gets you nowhere on this.
>> Very interesting. um
>> have to be hard nose. You can see again
it's it's not ad hominemum. I'm the
first to say Graciano is a brilliant
guy.
>> All these people are are brilliant. I
but you know this just won't work.
But I love the disagreement. That's the
only way forward. Um you and Shintan
wrote your wrote Observer Mechanics in '
89.
>> So you the two of you made a bet pretty
early and you were kind of out in the
wilderness for 20 years. Yes.
>> What was that like to just have no one
really paying attention?
>> That's right. So, we published a book
observer. It was three of us actually.
Bruce Bennett who was a ma a genius
mathematician.
>> Is he gone?
>> He died in early 2000s. Um and Chayon
Posh who is also a genius mathematician
and is still alive and and we still
collaborate and he's a good friend. Um
we we really worked hard in the 80s. It
was probably
four or five years of of hard work to
put out that book. Um,
and by the way, when we published it in
1989, um,
Wheeler cited it in his IT from Bit
paper. So, if you look at his paper, one
of the citations is to observer
mechanics. So, he'd read our book when
he published his paper on it from bit.
>> That's a fun phone call to get. Uh yeah
it's it it it re really is. So he was
aware that we had you know some
cognitive scientists and mathematicians
had really taken the idea of
consciousness conscious observers um
fund being fundamental and were trying
at the time to show how we could build
up spaceime from it. Uh so Wheeler was
aware of I
when I came from MIT to UCI I was on the
fast track. I mean I I was getting
grants left and right. I I it I was the
head of AI research at the time and
vision science and so forth. But when I
started moving into this, what was
really interesting to me now when I
realized in 198687
that we're creating everything that we
see. I mean you you can't walk away from
you the whole physicalist in one moment
when I real I don't know if I told you
but when I when I realized that
it was at a certain moment when I
realized what the mathematics was saying
to me was there there was a one point
where the mathematics all of a sudden
just slapped me in the face and I
realized that we're creating all this
stuff like it's a a VR headset.
>> Well before we do that let's take a
break and we'll come back and we'll
we'll get into the math. Okay. Okay.
Okay. Be right back.
[sighs] Let's assume everyone's seen
your TED talk. Just give me the dinner
conversation
sort of summary of of the theory,
>> right?
>> Give me the 60-cond on the beer bottle
beetle.
>> Right.
>> Tell me about denim jeans.
>> Right. Right. So the the TED talk
really took on
a deep belief that we have that
evolution
shapes us to see the truth. Right? So
Darwin's theory says that
shapes us to be fit. Shapes organisms to
be fit and it shapes their sensory
systems to make them fit. And most of us
and sort of Technically that means it
shapes sensory systems to make you good
at reproduction. The in in evolution the
the payoff the fitness payoff is how
many offspring you have. The more
offspring the more fit you are in some
sense.
And most of us assume informally and
even formally I mean experts that
sensory systems that are shaped by
evolution to be more fit do that because
they've been shaped by evolution to show
you the truth. Clearly it would seem
sensory systems that show you the truth
will keep you alive longer than sensory
systems that don't show you the truth.
And that's a deep deep intuition.
And it's wrong. It's it's so very very
brilliant people think that and it's
it's it's wrong. And I can in in my TED
talk I gave some explanations about why
and I also gave some examples and maybe
I should start with an example then I'll
explain the the the deeper principle
behind it.
>> So it's it's what and the examples
aren't aren't new but they really make a
point. So, one example is the
our beetles. So, there are these these
beetles in the outback of Western
Australia. Um, the the jewel beetle,
it's called, and they're dimpled,
glossy, and brown. The males fly and the
females are flightless. And the males
will go flying around looking for a
female. And if if he finds one that's,
you know, looks right, he'll alite and
mate.
And but in the outback of Western
Australia, there are some men um who
were drinking beer and these things that
called stubbies that were also these
bottles were dimpled, glossy, and brown
and apparently the right color of brown.
They toss them out into the the outback
the jewel beetle males would fly down
onto these these stubbies and flock all
over them trying to mate. Now, it's
remarkable. It's it's not like you they
flew down to it, realized it wasn't a
female, and took off. They fly fly down.
They're crawling all over, full body
contact, as much contact as they can
have, and they they persist in trying to
mate
>> until they're until they die.
>> Yeah, that's that's right. So, they they
they do not give up. And this could make
the species go extinct. So So, what's
interesting then is what this shows is
that the males don't know too much about
what a real female is. They don't have a
real understanding of a female. A female
is anything dimpled, glossy, and brown.
The bigger the better. [laughter] And
and and and that's is that's as much
insight as they have into females. And
some females might tend to agree that
males don't have much insight into it.
And and so that that really shows you
that when evolution is shaping sensory
systems,
what it's doing is giving you a solution
that's good enough for reproduction. And
in that niche, in their niche, what was
good enough apparently was dimpled,
glossy, and brown. The bigger the
better. You don't need to know anything
more about a female because you weren't
going to be fooled. There were no
stubbies at the time, and so that was
good enough. So, it's what you call a
satisficing solution. It's good enough.
It's just not the the best, you know,
possible thing.
>> This is don't worry about the man behind
the curtain, right?
>> That Yeah. Right. Um so, and but that's
not just a one-off now. It's repeatedly
we see that there are all these tricks
and hacks. Um there's if if if someone
wants to read about this, um look at
look online, you know, do an AI on
supernormal stimula. Okay? supernormal
stimula and you will find um all sorts
of things that are fun about how
organisms can be tricked um by by a
stimul and we use that all the time in
in designing um for example makeup is
supernormal stimulate when you put on
lipstick the red of your lips is
supernormal in in nature no woman would
ever almost no woman would have lips
that red um so it's not natural and yet
in men there is a program for sexual
attractiveness. And
>> yes,
>> up to a point, redder lips um tickle
that that algorithm and and make you
more, of course, at some point then then
you go into, you know, clown material,
right? So, right. So, you can go to so
you can push supernormal to a point and
then you go clown and then all of a
sudden you you fall off. So, so what
what you see in evolution is that our
sensory systems
in these these particular cases have not
been evolved to show the truth their
tricks and hacks. Now, David Maher was
quite a well aware of this when I was a
graduate student and he knew for example
that flies had tricks and hacks for for
their sensory systems.
Mar and Pojo looked at the fly visual
system and they so Mar and Poio would
say that the fly yeah had tricks and
hacks but but David said the human
visual system though is estimating the
true
shapes of surfaces. So we with our more
sophisticated visual systems and all the
billions of neurons that we have can do
something that the fly can't. So we have
been shaped by evolution to see the true
shapes of real objects in space and
time. So he was very much a physicalist
>> but he was wrong about that.
>> I think I think that he was wrong and
and the argument now I'll give the
technical argument um I'll try to make
it as accessible as possible but there
is a nice clean technical argument that
I think takes us apart very very quickly
in evolution
there are there's a mathematical model
of evolution but made by John Maynard
Smith. So Darwin was not a mathematician
his his theory was deep and brilliant
but it wasn't mathematical. John Maynard
Smith and others made it mathematical
with evolutionary game theory. So we can
now actually take Darwin's ideas and
state them with mathematical precision
and start to prove theorems and look at
the details of his theory. That was in
the 70s that evolutionary game theory
came out.
>> And in in game theory, there are things
called payoffs, right? If we're playing
a if you're playing a game, if you take
certain actions, you you you can get
certain rewards, right? There's payoffs
for you know who who your opponent is,
what what the situation of the world is
and what action you take. You'll get
different payoffs. This is just you know
standard nuname you think about it. You
know there are payoffs for being in a
certain state and taking certain
actions. So these things are called
payoff functions.
And so the idea that we're shaped by
evolution to see true structures in the
world can now be stated very very
clearly. there must be payoff functions
that actually know about the structure
of the world because they don't know
anything about the structure of the
world. If if the payoff function um does
not depend on the structure of the world
and it doesn't communicate the structure
of the world somehow then it can't
possibly shape you to know the structure
of the world. So it's all about these
payoff functions. So the nice clean so
they might show you things about the
world like the metric structure of the
world or or orders. Uh so there are
different kinds of mathematical
properties of the world that you can
get. So topologies, metrics, partial
orders, all sorts of technical things
that you could ask about the true
structure of the world. And so there's a
nice clean technical question.
What's the probability that a randomly
chosen payoff function knows about the
structure of the world and whatever
structure you want again topology,
metrics, partial orders, whatever it
might be. Now we can actually answer
that question. It's not just a handwave
question. We can ask we can answer the
question what is the probability that a
randomly chosen payoff function actually
could possibly shape you to know the
truth of the world. And the answer the
probability is zero. 0% of the payoff
functions hold information about the
structure of the world.
>> Fit wins.
>> So So fitness loses loses. Well, well,
if if you're going for if if you're
going for truth to give you fitness,
that version of fitness loses, right?
You're saying to see the truth is to
make you fit. That version of fitness
loses big time.
>> So, and
>> now Yale ran this with different payoff
functions and fitness one. What did they
get wrong or what or right?
>> Right. So in the in the Yale study there
are now when you go into so they didn't
actually address this question that I'm
I'm raising. Right.
>> So so and the reason I'm going up at
this deeper when when you try to go into
the weeds you can you can find little
cases where you can in this particular
situation get something to happen nice.
>> Sure you can. So you can always find a
case where where so so that's why I
wanted to go to the big picture here
because I've gotten a lot of little
little papers out there where people say
well in this particular case oh yeah
sure you can you can make a case like
that but the big picture is this
evolutionary theory
does not in its current form restrict
the class of payoff function that says
any possible function right now is a
legitimate function. Now it's perfectly
fine if someone wants to come up with a
new version of of Darwin's theory,
evolutionary game theory that says only
these classes of payoff functions are
permissible and these are the principled
reasons why. But we do not have such a
theory right now. So the current theory
as it stands does not restrict any
payoff functions. It set doesn't
eliminate any any payoff functions. So
then the argument is very very simple.
When you look at the set of all possible
payoff functions and you ask how many of
these the technical term is homorphism,
how many what fraction of them are
homorphisms of
total orders, partial orders, metrics,
topologies, whatever it is, whatever
structure you might and the answer let
me let me catch them up. Homorphism is a
subway map against the subway.
>> You the stations on the map correlate to
the real world. That's what homorphism
is. Just a projection. Okay.
>> Yeah. The map is a faithful
representation of the structure of the
of the subway system. Exactly. That's a
great that's a great way of talking
about homorphism. I'll use that in the
future. [laughter]
>> I just I got to catch them up.
>> Perfect.
>> Because it's hard to keep up with you.
>> Yeah. So, so the the the the stunning
answer. So, the whole all the arguments
that I've given and so forth, all you
need is this one argument. Payoff fun.
There's tons of payoff functions. What
fraction of them actually could shape
you to see the truth? Oh, 0%. So, what's
the probability that we've in shape to
see the truth? 0%. It's that simple.
It's just like give the whole argument
in one 30- secondond clip.
>> The set of payoff functions is big. The
ones that are homorphisms, probability
zero. Therefore, we don't see the truth.
Just that simple. So, all the other
arguments that people have now, here's
an argument against me, though. So, so
people will say, look, uh, Don, this may
this is all high flutin math. It sounds
really great, but here's the fact. You
shot yourself in the foot logically,
right? You started off with Darwin's
theory of evolution, which assumes that
there are physical objects like
organisms and resources in space and
time.
>> competing with each other. So, it's a
physicalist framework. And then you're
using Darwin's theory to show that there
is no such thing as organisms in space
and time. And and so you've used your
theory to refute the foundations of your
theory. So, you should just go learn
some logic. done. This is This is
stupid.
>> I've seen this.
>> Oh, yeah. I've gotten it in
publications. Sure. In in philosophy
journals. Yes.
>> And I get it in almost every
>> That's where I saw them in the journals.
>> Yeah. Yeah. In the journals, but I I
also get that all the time in in
comments on on YouTube videos and so
forth.
>> Don't read that. Don't read those. Stick
with the journals.
>> Yeah. The journal. Absolutely.
Absolutely. So, so the the reply is is
very very straightforward.
Every scientific theory starts with
assumptions.
It no scientific theory is proving its
assumptions. It's assuming its
assumptions. Those are the miracles of
the theory.
It the theory is saying if you grant me
those assumptions please then I can
explain all this other wonderful stuff
and if it's a good theory it it will.
If it's a really good theory, it will
give you the mathematical tools to
really explore the scope of that theory.
But it's going to be a finite scope
because it it's not a theory of
everything because it doesn't explain
its own assumptions. So no scientific
theory is a theory of everything because
no scientific theory explains its own
>> There can never be a theory of
everything.
everything. Just that simple. The the
argument is drop deadad simple.
>> Every theory has assumptions
>> and they don't explain. So a good theory
though will give you the mathematical
tools to explore its limited scope. It's
not arbitrary scope. It's limited scope.
But a great theory will give you the
tools to actually understand
the the limitations of the assumption
themselves. So so a great theory will
actually tell you that the assumptions
it themselves are not the final word
which we knew all alto together. I mean
we we knew that before that the
assumptions cannot be the final word.
There's going to have to be a deeper
theory. So I'll give you an example.
Einstein's theory of um spacetime um
together with quantum quantum mechanics.
So quantum field theory
>> that theory
among its assumptions are that space and
time is fundamental. Quantum fields are
defined over spaceime.
And then Einstein plus quantum then when
you look at the mathematics it turns out
that spacetime itself falls apart at
what's called the plank scale 10us 33 cm
10 - 43 seconds.
>> So here's a case where
the theory says we're going to start
with quantum fields in space and time.
That's the the the assumption. And then
it proves that spacetime itself cannot
be fundamental. that it falls. It has in
fact no operational meaning at the plank
scale. So no one comes along and says
Einstein should have and the quantum
guys should have learned some logic.
They they can't use their own theory to
prove that spacetime isn't fun. They
assume spacetime is fundamental and then
they prove it's not. Those stupid guys
they should just you know they they
should go and learn some. No one says
that this this is actually viewed as as
a breakthrough. This science is so
rigorous and so mathematically precise
that it can show you the limits of its
own assumptions. That's how we make
progress. We knew that the assumptions
could not be the final word. But a
theory that tells you exactly where
those assumptions fall apart is exactly
the kind of tool for the best rigorous
science. And so what I'm saying about
evolution is that yes, Darwin started
with physical objects,
organisms
in space and time, fighting for
resources in space and time. And when we
look at his Darwin's own theory with
John Maynard Smith's mathematical
version of it, we see that his theory is
brilliant enough to show that the very
assumptions of physical objects inside
space and time is not fundamental. He
was a able to show that that cannot be
the final word. So it's a brilliant
theory. So that's the way science works.
We will never have there's infinite job
security in science in principle.
There'll always be deeper assumptions.
What we don't want to do is to um fall
into the trap that that Mox plank
pointed out, which was that science
tends to
move forward um or progress one funeral
at a time. It's
>> it's better for us to let our theories
die to to look to know up a priori. Our
assumptions are just assumptions. We
don't have a theory of everything. In
fact, my own view is we are 0%.
Always 0%
of a theory of everything. That's that's
all the science knows. And yet that's
important to have that 0%. It's rigorous
and it's 0%. So it should be very very
humbling. Um, and it means that the next
generation does not have to worry that
the older generation did it all. No, no,
no, no. You will always have plenty to
do. There's always new opportunities in
science.
>> You want the new generation to just
start with theory of everything is 0%
and go from there.
>> Well, I would like them what what the
new generation has to do, of course, is
to take the current theories
very very seriously. You have to study
them. You have to do your homework. You
have to know them backwards and
forwards. There's you don't have a
prayer at this level of sophistication
in science. You don't have a prayer of
doing something new until you've really
spent several years really mastering
what we've got and then allowing
yourself to think out of the box and
say, "What are the new deeper
assumptions that we bring to this
thing?" So, so, so
I get emails all the time from people
who haven't done their studies and
they've got their new theory and and and
you look at them and you and you and you
realize
They have no idea what I mean they're 10
they need 10 years of study before they
can even begin and and using an AI is
not going to help you. You can't make
this gap up. You have to know the
theories. You have to have really groed
them yourself.
>> Otherwise, you will be misled by the AI.
You'll be misled to think that you've
got your new theory of everything and
and it's unfortunately a waste of time.
Now, if you've done your homework, if
you spent the years and really mastered
the theory, then it's safe, I think, to
use AIs as an assistant as you're trying
to develop stuff because you can then
step back and evaluate what's going on,
and you can have the AI push you around.
But if you don't actually yourself know
the current theories, then you can't
know when the AI is taking you down a
dead end.
>> That's right. AI is great for pattern
recognition, but there's no intuition
there. You have to already know the
subject matter.
>> That's right. That That's right. So, but
if you do, then it's a helpful tool.
>> Oh, of course. Uh before we get into
trace logic, photons,
>> the aha moment, let's let's kind of
build the foundation with with conscious
agents, what what are the what are those
in your framework?
>> Yes. [clears throat] So if I'm going to
say that consciousness is fundamental
and I want to be a scientist, I've got
to have a mathematical model.
>> And that's no small order.
The idea that consciousness is
fundamental has been around for
thousands of years. And there has not
been a single mathematical model of of
scientific merit anywhere. It's truly
stunning to think about that
consciousness has been around. The idea
of the consciousness is fundamental been
around for thousands of years. And
>> do you think it's stunning there's no
model for that? It seems I'm everyone's
surprised that you have one.
>> [snorts]
it it it's stunning because
the model that I have is so simple.
It's it's it's truly to me it's it's
stunning that the in in retrospect the
the model is simple and and but of
course even in my case it I just
discovered a big step in it just five
months ago. So I've been at this for 40
years. So that um
>> well that I want to get to that moment
because it helped me understand as well
when I finally was able to grasp it. It
took a it took a little bit of reading
to kind of get what you meant with the
math and
>> with with trace A implies B and all of
that but I finally got it.
>> Made a lot of sense.
>> Oh great, great. Great.
>> And the um
>> the way the universe ticks at different
rate at ticks at a certain rate. All
that stuff is it's very interesting.
>> Yes. Well, so yeah. So I'll just say
that that the what I had to do and what
my team had to do. So Chayon and and
Bruce and others that have been working
with me now many others is we have to
have a mathematically precise model of
consciousness. And so what we we what
you want to have the simplest thing that
you possibly can. You don't want to have
a Rube Goldberg device. You want to have
the most cut down. So ultimately
the idea that we have is this
>> what's the for just a conscious observer
we'll do agency in a minute but just for
a conscious observer what's the minimal
minimal thing that you would want well
observers have certain experiences that
they can have maybe like I can see red
green or blue or something like that so
I want I'm going to list the experiences
that this observer can have and the
other thing that I seems necessary is to
say those experiences can change. I'm
seeing red now. Maybe I'll see blue next
or red next or green next. And that's
the minimum I could imagine. They're
experiences and they change. And what's
the most general mathematical object
that you can use to do that? Well, you
list your experiences in like a column
and then for and then for each a row
next to each experience, you say, what's
the probability if I'm having this
experience, I'll go to that experience
or that. You just list all the
transition probabilities. That's it.
That's called a markup matrix. That
little thing I just talked about is so
simple. It's just a it's called a Marov
matrix and it was discovered by Marov in
1905. So,
>> so um let's use the traffic light
example because that helped me
understand it.
>> Okay. So, with the traffic light, you
have red, green, and yellow.
>> Yep.
>> In this particular case, it's a very
simple matrix. If you're seeing red now,
the probability that you're going to see
red next um is zero.
>> But you're you're the next and
probability you're going to see green
next is one,
>> right? probability you're going to see
yellow next is zero.
>> Right? So this is how we're building the
matrix.
>> So that's right. So the first row is 0 1
0. Now for the green row, the
probability that you're going to see red
next is zero.
>> Probability you're going to see green
next is zero. Probability you're going
to see yellow next is one. So that row
is 0 01. Then for yellow, the
probability you're going to see red next
is one.
>> And then zero for green and zero for
yellow. So, so that's that that's how
you so that's the matrix for for that
simple case and that's um a very
interesting kind of matrix. It's it's a
a cyclic cyclic matrix. Um but um most
marker matrices aren't cyclic. They're
they're more complicated than that. But
that that gives the idea. So but now
just imagine we're we're just going to
make this following statement. Um all
possible observers are represented by
all possible markup matrices. So these
markup matrices um could have maybe one
has a thousand experiences, some tastes
and colors and smells and other things.
Some might have a trillion, some might
have a Google.
>> And it goes off to infinity. So you
could have infinite matrices. So just
imagine if you can, it's hard the space
of all possible matrices of all possible
dimensions. That's the space of all
possible observers. That's so so the
idea is very very simple. But when you
say we have no reason to exclude, right?
I don't have any reason to exclude any
set of observations or transitions. So
I'm just going to say that all possible
observers are all possible matrices and
it's an infinite space. So that's what
we do. And but so each matrix now think
of it as is an observer window. It's
it's just a way of looking. It's a way
of seeing like the traffic light. That's
a way of seeing. And if you look at a
traffic light, that's what you see. is
it's spinning around but other things
are much more complicated in this room
now. Now now we probably need trillions,
right? We need matrices of trillions and
the probabilities are quite complicated
and so forth. So that's a but all of
those are passive observers. If you're
you're just sitting there, you're
looking through this window and you're
not doing any action. You're just
watching. What about agency? How do you
get the notion of agency?
Well, the idea would be suppose a a
clean way of thinking about agency is
I'm looking through this window, but now
I want to change and look through that
window or that window or that window.
So, how do I want to model that
mathematically? Well, remember what I
did with observations. I said, here are
the possible observations I could have
and I'll just talk about how I could
move around on those observations. You
know, I'm seeing red now. Now, move to
green now. So I moved around on
observations and I wrote down the matrix
for how I moved around. Well, now I want
to move around on windows. So I have all
these observer windows. There's an
infinite number of them. So I'm going to
have a policy about I'm looking at
through the world at the world through
this window. Now I want to move look at
the window that through this window and
so forth. That's I call that a policy.
And and what is that? That's another
markoff matrix because it says here's
the probability if I'm looking at this
window that I'll look through this
window or that window or that window. So
it's another markoff matrix. And I can
then say, well, what's the set of all
possible policies? What's all the
different ways I can move through all
these windows? Well, it's an infinite an
infinite collection of Markov matrices.
>> But it's also another matrix, isn't it?
>> Uh well, it's it's not itself a matrix.
It's an infinite collection of Markov
matrices
>> and and then I could say, well, now I
want to change the policies. I mean, I I
have this policy. Now I want to change
to this policy. So I can now walk around
on policies. So you can see this goes
off to infinity. This is a what we call
recursion.
>> So that's that that starts to get us the
notion of agency now because the the the
passive observer windows are are a very
very minimal notion of agency. There
there is some notion of agency that you
can get there. Right? If you if you take
a markup matrix that's erotic anywhere
any time that you change the start state
it will always go to the same long-term
behavior. And so in some sense that's
that's a weak notion of agency. It's
always trying to get to the same place.
But that's a very weak notion. But now
with the policies, you're you're moving
around on these windows. That's a little
bit more agency. When you can change the
policies, that's even more flexibility.
Now I'm changing how I'm moving around.
But then as you go meta meta meta,
you're getting ever more sophisticated
agency off to infinity. So this gives
you a way of unpacking the notion of
agency from the most trivial namely the
observer windows which are mostly
passive all the way out to infinity
which is infinite flexibility of agency.
So this for me is so beautiful because
it means that we don't have to take on
all of agency at once. we can go through
the recursion once really understand
just the policies understand what they
can do really master that once we've
done that we can then go on and do the
meta policies and so forth and so we can
so it gives science a way to really go
through this from the least complex to
the most complex but now the thing
that's really quite quite interesting
about this I haven't told you the mo the
most fun part that that that's the
recursion but
>> the most fun part well first I should
say that this idea um I should tip my
hat to Linets who around 1700 in his
monodology was was basically saying
something like this. Um I don't want to
put words in his mouth but I I suspect
that he might like this approach.
Livveness said we need to start our
science with perceiving entities
observers. He called them monads. Mhm.
>> And we need to he he was quite
religious.
So we need to we need to have some kind
of pre-est established harmony and he he
was thinking about God having set up
this pre-established harmony so that
they are are coordinated somehow that
the they're not just random observers,
you know, just completely disconnected
from each other doing whatever. There
had to be some kind of coordination. And
at the time, of course, Newton was also
trying to get a mathematical model to to
found science. And and Newton also was
quite religious. He wrote more theology
than he wrote physics.
>> Did more alchemy than he did physics.
>> More alchemy. That's right. But and so I
think that he would have liked to have a
situation where he could put, you know,
consciousness and observers fundamental,
but the math just wasn't there.
>> And so I think what what happened was
that that Newton said, "Look, we've got
to go with what we can do." So he he can
write down f= ma and you know f= g m1 m2
over r^2 for for gravity. We can we can
do that kind of stuff and we can get
going. We we need to so that set of
equations gives you a more physicalist
machine kind of universe and and so
science got started in a machine kind of
universe because that's what we could do
with the mathematics.
I think Linets
was right and I think Newton would have
liked to go that direction, but the math
just wasn't there to to do that. But now
I think I can show you something that
looks like this um pre-established
harmony that lives looking for. So I got
this whole set of observer windows. So
we'll forget the agency for a moment,
just the observer windows.
Um, suppose that I'm looking at
say 10 colors and there's a matrix
that's governing how I see those 10
colors. So that's that's that's my
window. But suppose as I'm looking
through that window, someone somehow
shuts off seven of the colors. So I
can't see those seven colors. So I'm
still looking trying to look through the
window, but I can only see three of the
colors. Well, now I'm going to get
effectively a 3x3 matrix on those
colors,
>> right? that's induced by the big 10 x
10.
>> So, there's all sorts of hidden stuff
going on in those sears that I can't
see, but
>> but they're there.
>> But they're there. They're hidden, but I
can't see them, right?
>> But they're but they do influence what
I'm seeing in the 3x3. So, that
>> it's a good way of explaining this.
>> Yeah, that's Yeah. And so, that thing is
called the trace matrix.
>> And for those who are more
mathematically sophisticated, I do have
to make a clarification with matrices,
there's another notion of trace that's
even more common. You take a matrix and
you you add up the diagonal elements. So
1 comma 1 2 comma 2 3 comma 3 n comma n.
You add up all those numbers and that's
called the trace. So that's I'm not
talking about that trace. That is of
course a notion of trace. But that's not
the one I'm talking about. This is more
sophisticated. The the one I'm talking
about for mathematicians is called the
shure complement. Sur complement. So you
take a big matrix and you induce a
matrix on a subset that that you can
see. So
>> right like a trace element a small part
of it
>> that that that's right and this so this
small part is in effect reflecting the
hole but just through this smaller
window it's almost like the smaller one
is observing the hole but through the
smaller window
>> and that's your trace
>> and that's the trace and here's here's
what I discovered about two years ago I
realized so but by the way I should say
the trace is not me this notion of trace
is has been known for maybe 50 60 years
so I didn't invent that what what I
discovered a couple years ago was that
that gives us the trace relationship is
a partial order on all markoff chains.
It gives you a logic for those who know
a little bit of mathematics is not
boolean. So boolean logics are the ones
that we are more most comfortable with.
You can take ands and ors and
>> yes
>> complements and so forth. Um so this but
boolean fits into your theory at the
local level doesn't it? It
>> it does at the local level. So, so this
so this logic I call the trace logic it
has an it's not boolean but it has an
infinite number of boolean sublogics.
>> So there's so for you take any matrix
and you look at all of the matrices that
are traces of it
>> and all of those matrices together form
a boolean sublogic. It's really really
pretty.
>> It's elegant. So unbelievable.
>> So the the local boolean is the base of
all of it.
>> That's right. You have an infinite
number of these local booleans, but how
they get tied together. We're still
trying to understand this is this is
nasty math. We're still trying to grock
this. But but I still remember two years
ago when I
working with Chayon Pash and I said to
him, Chayton,
>> I believe that this trace relationship
will give us a logic. It's a partial
order and it gives us a logic. And his
his response was, Don, that's too pretty
to be true. But then he went off and
proved it. He what we had to do is what
he had to do was to prove that it has a
transitive relationship and so he he
proved that and we had the logic
>> uh transitive so the tra trace a implies
b
>> that's right so the trace of a trace is
a trace so I have a big so I have a 10 x
10 and I trace it onto five and then I
trace the five down to three I get the
same answer as if I went from 10 right
straight to three on
>> what did he say when he called you from
Heathrow after he proved it
>> um well I think it was fairly a matter
of fact It's it's it's true and I was I
was I I I
believed it was true. I I I believe it
true but I was but when Shayan says it's
true then I I know it's true because
Shayon is brilliant um and he doesn't
usually make mistakes. So So I was
really quite pleased. So it's this
beautiful logic and and now you can see
that it applies to the observer windows
but not just to the observer windows now
it applies to the policies the first
level of agency because they're that's a
whole set of markoff kernels. So there's
a trace logic on them and then there's a
trace logic on the meta policies and the
metameolicies and so so you get a
recursion of these trace logics all the
way out to infinity.
And I want to propose that this is the
pre-established harmony that Labinus was
looking for. Wow. So there's all these
observers and they're tied together by
this beautiful recursive mathematical um
logic. So we we are still trying to
understand this logic. This is there's a
lot of mathematics to be done on this. U
>> how's the establishment responding to
this? I well it's
the theorem is true. So yeah so so
there's no problem with w with with the
theorem. Um
>> but if chayan proved this that means
others can
>> that's that's right but in fact for a
mathematician this is falling off a log
>> right to prove that theorem is I mean
Chayon was draw flew to heathro while he
was waiting in the in you know had he
for something he he proved it so it's
it's the kind of thing where a brilliant
mathematician like Chayton can can prove
it pretty quickly um
>> so that's just logic is just logic so
what
>> right how do we bolt your philosophy
onto it that makes it controversial.
>> Right? So
one way to look at what I' I've just
said is that I just did discover this
new structure and and recursive
structure on the set of all marov
chains. So that's that's a contribution
to mathematics and as a contribution to
mathematics it's it's not controversial
at all I don't think
>> but now I'm interpreting this as
applying to consciousness and that's
very controversial. So, so and here's
the kind of thing that gets publicly
stated about this. So, so there will be
there there was
I I don't know if I'll mention the
person's name but very very um prominent
person with YouTube said about this that
this is complete nonsense. Of course,
Markoff chains are are fine, but to use
them for consciousness, this person
said, you see, Markov chains are used
for standard stuff like like predicting
the weather, protecting stock markets
and so forth.
>> Nuclear fision,
>> nuclear fision,
>> Google page rank.
>> That's right. So, exactly. It's used for
all this stuff. And and to to say that
it applies to consciousness is is
nonsense.
You know, there's nothing in the
mathematics that says this is
consciousness. So, the mathematics is
doesn't care about consciousness. So for
Hoffman to even say that it's about
consciousness is is is just
a rookie mistake. So a lot of people h
have said this and and and my
>> I've seen that but I haven't seen them
explain how you're making a mistake.
>> No, they just say that they they just
say that that is a mistake and they
don't say how that is a mistake and and
I'll explain how it's not a mistake.
>> Okay. Um
the math it never tells you what it can
be applied to. So the fact that I mean
you could also say um you you're using
Markoff chains for stock markets.
There's nothing in the Markoff chains
that said they could be applied to stock
markets. You're using it for weather.
There's nothing in Markoff chains that
this should be applied to to to weather.
Um so as soon as you see that you really
the the claim you can't use it for for
consciousness is silly. It's just it's
plain silly because you could apply it
to any other application. The math does
not tell you any applications. It said
this is a structure. When you use it in
science, what you do is say, I think
that this scientific arena might
profitably be modeled by this structure.
Maybe it's stock markets, maybe it's
weather, maybe whatever it might be. In
my case, I'm saying consciousness. And
now I think what's really going on is
the argument that they're giving is that
the real reason that they don't like it.
They just don't like consciousness. So
they're what they're really saying is
consciousness is nonsense. And so you're
trying to attach nonsense to marov
chains. And and and
you could be right. Maybe consciousness
is nonsense. We'll see. Right now,
physicalist approaches cannot explain a
single illusion of conscious experience,
much less conscious experience. So, you
know, if in 50 years we're still batting
zero, I would say it's it's over for
physicalism. It's just it's over. And if
we can start with the theory of
consciousness modeled by Markov chains
and we can do some real work with that
and we can talk about the work like
building up spaceime.
>> Then I would say um it doesn't prove
that consciousness can be modeled by
Markoff chains but it sure makes it a
pretty interesting scientific
hypothesis. Certainly not nonsense.
>> Well, let's build that bridge. Let's go
let's go from the chains to
consciousness because we haven't really
built that bridge yet. How we get there?
>> Right. So, so the idea, the reason that
I was intrigued by by using markup
chains for consciousness is that um we
do just have at the very very most
elementary level the experience of um
colors, shapes, and even more
complicated thing, threedimensional
objects and so forth, and they're
changing. So, at the very very minimal
level, um the the this Markov model
works. Now one objection might be
Markoff chains have a finite history. I
mean they have a finite memory in the
sense that um the next the next
transition is um determined by the
current state of of the chain.
>> Yes. So some people will say well that's
a finite memory kind of thing and and
that's just way way too limited. Um but
it's not because it's well known in
marov chain theory that you can easily
create bigger more complex states. You
can take a series of 20 states and make
it one state. So as as much history as
you want you can build it into states.
So it's it's really um there's no no
limits and you can have an infinite
number of of of states and make your
histories as long as you want and you
can make the individual experiences
quite complicated. I mean it could be
not just something I mean I use red and
green and blue because it's very very
simple but it could be you know a
particular threedimensional shape here a
sphere or a cube or so forth and then
you can go even more complicated things.
So so the kinds of experiences that you
can have can be arbitrarily complicated.
So so it seems to me that it's a good
hypothesis
to use these markup chains on the what I
call the recursive trace logic. this the
the recursion of Markov chains policies
and metapolic and so forth with the
trace logic as a theory of of
observation and agency because it's the
most general and comprehensive the least
assumptions there are experiences and
they change it's it's amazing that's all
I assume there are experiences and they
change that's it literally that's it
everything else then I say and the best
way to do that is is model it with
markup change that's that's all I assume
um the recursive trace logic just falls
out of that. It's it's it's just a
theorem. So I love that in the
scientific theory you make a minimal
assumption. I'm going to start with
consciousness. I'm going model it as
experiences that change.
The minimal model I can give is Markov
chains. Oh, by the way, no one noticed
it, but there's this partial order, the
trace logic on all this, and it's
recursive. And this gives us a theory of
agency. It's just that simple.
>> So consciousness is just a state that
can change. um but the it can change but
the policies and metapolicies are doing
it in sort of an agentic way and in some
sense they're saying given that I'm
looking at the world through this window
now these are the different ways that I
want to you know want want to look so it
that brings in a notion of agency and a
notion of of time which is quite quite
interesting because you can't choose
what next window you're going to go to
until you have a current window. So, so
there it brings in of a very observer
centered or agent centered notion of of
And someone might say, well, you're not
these are just markoff chains. You're
not showing any deliberation. You're not
showing any rational reasoning or
reasoning here or you know like what
what are the um goals that you're and in
and the payoff functions that you're
trying to minimize or maximize in in
your choice. And and I would just say
that these are different levels,
different ways of describing the same
thing. If you give me um a decision
theoretic description of an agent that
says um I have these goals, I have these
kinds of actions that I can take. I I'm
modeling these kinds of worlds. And so I
can then model what these actions would
do toward my goals in these different
worlds. Um and you can write it out that
way. Sure. But ultimately you're going
to have when you write it down you'll
get a markoff chain says what's the
probability given that I'm in this world
that I'll go and do this thing or go to
that world. So so there are just
different ways of describing the the
same thing. So, I like the Markov chain
because it has this recursive trace
logic. And I think also because it's
looking to me like we'll be able to show
how we can build spacetime and quantum
theory from just this recursive trace
logic. So, I should say I've been very
very hard-nosed in this interview about
the physicists, right? I've said look
>> they start off with space and time and
physical objects as fundamental. Then
they owe us a precise physicalist
account of conscious experiences or the
illusion of conscious experiences. No
handwave. Show me how we get the
illusion of the taste of mint from
neurons or whatever you want. Show it to
me. Or why should I believe you? So turn
around. Now what my colleagues will say
to me is, "Don, great. You got this nice
mathematics and you're claiming it's
consciousness.
Great. Where then does spacetime and
quantum field theory and the borne rule
and all this stuff come from? All right,
you've got all this nice stuff of
consciousness. What you owe us is show
me Einstein special theory relativity,
general theory of relativity, quantum
field theory, and eventually quantum
gravity. Give us um non-locality, give
us the Borne rule, give us the big bang.
>> Can you give us this? Um well I I just
gave my colleague Nifa and and and
Chayon um tentative proofs of getting um
special relativity and general
relativity.
>> Um it looks very very plausible to me
>> and I've got a I'm working on the Bourne
rule.
>> So are you integrating time dilation and
space dilation? I I am and I can give an
intuition about why we might expect that
this would work.
>> Please.
>> So one thing that um Einstein taught us
is for special relativity is suppose
that you're on a train and you're going
past me. I'm sitting at the train
station. I'm watching you. You're going
past me and you have a clock and you
have a meter stick and I've got a clock
and I've got a meter stick. And I look
at AJ's clock and for me it looks like
your clock is going too slow and your
your meter stick is too short.
>> And you on the train looking at me, you
would say, "Well, no, no, Don's clock is
going too slow and Don's meter stick is
too short." And that's pretty stunning.
I mean, that was stunning that Einstein
came up with that and very
counterintuitive and wasn't accepted
right away and and so forth. But
experiments have shown that that's
correct. The meter stick changes
Does Manowski spacetime factor into
this?
>> That that that is where Manowski
spacetime comes from. This is where when
you don't have gravity in involved.
>> When you don't have gravity.
>> Yeah. When you don't have gravity. So
this is just Minowski space. So Einstein
did this in 1905. It wasn't until 1915
that he came up with curved spacetime.
Took him 10 years to to really master
the mathematics for that.
>> Oh, is that all?
>> Well, yeah. remarkable. It was a
herculan job. Truly impressive.
>> So, so how do I do that in in this marov
chain theory? Why why should we believe
that I could get Einstein? So, one thing
I haven't mentioned um is that it's
standard to marov chain theory to have a
little counter for a markoff matrix. So,
every time you have a transition of
state um you just increment the counter.
So, like with red, green, and blue I see
red one. Oh, now I see green two. Oh,
now I see blue three. And you just keep
counting as as things go. So that's
that's called an enhanced markoff chain.
Or so we call it an enhanced markoff
chain. This is standard stuff in in
Markov textbooks. It's not our
invention. So notice what happens if I
have a the 10 by 10. Every time one of
the 10 colors changes, my counter is
going. So red, green, yellow, blue,
whatever it might be, the counter is
going.
>> But only a state change.
>> Only for a state for each state change,
right? Every time a color changes from
red to green or red to red, right? So
you you could have red change to red.
>> Sure.
>> So now suppose we go back to the 3x3 sub
window.
>> Notice that say it only has red, green,
and blue in it, not the other seven
colors. So it is it is only going to up
its counter when red, green, or blue
happen, but it won't catch yellow,
purple, pink, or the others. So notice
its counter isn't going to go as fast as
the big counter. The big matrix, the 10
x 10 is counting every one of the 10's
colors changing. The little 3x3 is only
counting three of those. So, it's only
getting 30% roughly of the counts
together.
>> Let's let's help help people understand.
>> Let's have the colors changing. Give me
give me colored glasses and let's and
let's take let's click our counters
>> Right. So, so suppose I have um let's
keep it simple to like then just three
versus two counters, right? So, I have
red, green, and yellow. So, we're at a a
traffic light. And so, every time it
changes from red to green, green to
yellow, and then and back to to red, um
you you click your thing. You you click
your counter. But now, suppose you don't
see the yellows. So, all you see is the
red and greens. So, you only get the
red. Anytime red changes to green and
green changes to red, you click. But you
don't get any of the yellows. So, you're
missing a third,
>> One out of three counts,
>> right? You're clicking your counter
every change, but I'm missing a third of
them.
>> You're missing a third of them. And so
one clock is only going at 2/3 the speed
of the other clock. And that is where
we're going to get Einstein's time
dilation
>> from there. So So time is moving slower
for me
>> because your counter you're seeing
>> counter slower.
>> That's right. So the reason I think that
your So the reason I think AJ's counter
is slow is going slower is I don't my
Markov chain has a is not completely
intersected with yours. So I'm not
counting all the stuff that you see,
right? I'm not my counter is not going
of your stuff is not going as fast as
your counter does. And so your counter
is going faster than from your point of
view. But the same way is that's why
it's it's it's symmetric from AJ's point
of view. My counter is going too slow
because you're missing some of the stuff
that I see. Right.
>> So that's why our counter we get the
time dilation.
>> What's the minimum unit? Is that a plank
time?
>> Now that would be
>> what's our universe frame rate?
>> Well, in in the um in the recursive
trace logic, there is no minimum. There
is so the the notion of time the notion
of time is a space-time notion.
>> There's no unit at all.
>> There's no there's no in the notion of
seconds. Right. So right. So if we if
we're talking about like the plank time
that would be 10 the minus 43 seconds
but the very notion of second is is
alien to the recursive trace logic. It
only has counters. It doesn't have the
notion of time. But what I have to do is
show that I can use the recursive trace
logic. In fact to you I can use 0% of
the recursive trace logic to build
Einstein's special relativity. Don't you
need some type of base measurement in
order to get your counters? You have to
be counting something.
>> Well, so there will be a a a like a
fundamental clicker in the recursive
logic, but it it won't have the notion
of seconds tied to it at all. It will it
>> I guess it doesn't need it then. As long
as it's there.
>> Well, and as long as it's there, it it
could be instead of 10 theus 43 seconds,
it could be 10 theus 43 trillion
seconds. In fact, 10 theus infinity
seconds. So it really can be as small as
you wish. So but our spacetime is stuck
at 10 the minus 43 seconds which is
actually if you think about it why not
10 theus 43 trillion. So and also
spacetime falls apart at 10 theus 33 we
might think that's pretty small. No what
about 10 theus 33 trillion? Why why why
should spacetime fall apart at 10 theus
33? That's a fairly shallow data
structure. So, so what what what I have
to do is to show how I can use the
recursive trace logic which doesn't even
have the notion of seconds. It just has
markov chains with counters and I have
to show how we can create a what I call
a headset.
>> I was going to ask you about this
>> a VR headset
>> because because you could still make
this work
>> physic this is could still be
physicalism with the Markov chains.
>> Oh, sure. Absolutely. It's only with the
headset do we now make the leap to
consciousness
>> that well well it's only with the
headset that I'm able to make the leap
from saying that the recursive trace
logic models consciousness to say and
here's how we get what we call the
physical world out of it. So that's what
when I'm building a headset I'm saying
I'm starting with this universal
consciousness the recursive trace logic
>> mathematically described with the um the
pre-established harmony that that
livveness wanted.
>> Yes. So that's what you know that's what
liveness wanted I think now I have to
show you how we get Einstein's special
and general relativity how we get
quantum field theory as a special model
within the recursive trace logic and
what I have to be very very careful
about this I I don't want to prove that
the recursive trace logic forces us to
see the world like through manowsky
space or general relativity or quantum
field theory it doesn't.
All I need to do is prove that it allows
us to build those structures. In fact, I
would be disappointed, bitterly
disappointed if it forced me to build
those structures because I want the
flexibility to show that our spaceime is
one of an infinite number of headsets
that that consciousness can build.
Consciousness has the flexibil we have.
I think my view is ours is the one of
the most trivial and simple space-time
headsets that's available. We we have
the training wheels version really dumb
down. So our view that we're like the
top of the food chain and then we're the
top of intelligence. My view is no no
our our space-time headset is one of the
most trivial ones that that could
possibly be. And we can show
I plan to show and I think I I I've got
a version of the proof right now but
it's not real until it's published.
Right. So I've got I've got a proof that
I like
>> likes it.
>> Uh Chayton's looking at it. Nifa are
looking. So Chayon and NIFa are two
mathematicians that I'm working with. So
um it's not real until they say it's
real and and and in fact even then it's
not real until we
>> do peerreview publication.
>> So the headset is how we perceive
reality. If ours is is is very basic
that implies that there's something more
advanced. that the mathematics makes it
very very clear that that there's an
infinite number of far more interesting
headsets than than others. Infinite
number infinite number there's
unbounded. So ours is one of the most
trivial ours you could think about ours
as being a trace of much much more
interesting bigger headsets in a way.
But even among the 3D headsets, right,
there are human y kind of 3D headsets.
Presumably, mice might see in three
dimensions. Some birds might see in
three dimensions. And their headset is
going to be a little bit different from
ours. It's still threedimensional, but
it's going to have different features
than ours. So, what we're going to want
to do is I I want to first prove that I
can get a generic space-time headset, a
generic one, not it's not human, it's
not mouse, it's not bird, it's just a
generic one. And then what we want to do
is then look at like the neuroscience of
humans.
>> If there's an infinite number of
headsets above that must mean there's an
infinite number below.
>> Uh there there could be quite a few
below but it's hard to go below three
dimensions and one dimension two four in
terms of number of dimensions you can 3
2 1
>> dimensionality is important.
>> Well for that kind of measure of of
headsets if you're going to have
dimension now you could get rid of
dimension as well. It could be headsets
in which you just have topologies and
not dimensions for example,
>> right? That would make every headset a
mean then.
>> Uh,
>> no matter where you are,
>> um, well, it Yeah, there's a you can't
think big enough. There's there's
[laughter] there's an infinite number of
different kinds of headsets that you can
>> humans don't like infinity.
>> We don't like infinity. We like to think
that what we're seeing is the truth. And
and what I'm saying is that from this
point of view, spaceime, which science
thought is the final reality. and
everything is inside spaceime is in fact
one of the most trivial headsets that
you could build out of the recursive
trace logic and there's an infinite
number of more complicated headsets than
what we've got and I think that we are
the consciousness that's capable of
understanding those headsets you
>> mean our species
>> our spe well our species is a headset
representation of the consciousness that
is actually able to do all of this stuff
and so part of our Our joy here is to
enjoy this headset.
>> To wake up and realize it's just a
headset and to realize like the Leela
thing. This is a game.
>> We thought we let ourselves get lost in
this game and we can smile once we wake
up and realize, oh, we thought this was
the whole thing. Oh, no. No. This as as
as beautiful as this is, as complicated
as this is, as completely engaging as
this world is, it's trivial compared to
what you can do. Right. Yeah, that's the
idea. That's the Leela kind of thing.
Let's let's play with this. Let's enjoy
this. So, you know, relax and play with
this. See what you really can learn from
this. Be open. Explore. And then realize
that you infinitely transcend this. And
the recursive trace logic puts that out
mathematically. You can infinitely
transcend this in the sense that there
are an infinite number of other headsets
that you can build.
>> Is there a way to practically describe
what those headsets can see?
>> Well, there's my monkey brain. Well,
there's one one way of thinking about
it. Just one simple way is go from three
dimensions of space to four to five to
50 to a billion and so forth. So that's
one way
>> dimensions are infinite as well.
>> Yeah, why not? Why not build dimensions
off to infinity?
>> And that's just one that's one
direction. And then you can let go of
the notion of dimension all altogether
and just try different topologies and so
forth. So there's as soon that's the
thing you cannot think big enough here,
>> right? and and the mathematics really
helps you. All these structures that
mathematicians have discovered open us
up to realize all the different
possibilities and infinities. So, so
I I study mathematics because it really
helps push me out of my little boxes,
out of my little, you know, dead ends.
Thought dead ends. So,
>> I've heard you say you know enough math
to get into trouble, but not enough to
get out.
>> Who gets who gets you out? Well, so my
my good friend Chayon has worked with me
since 1984 or something like that. So,
it's been 40 42 years he's put up with
me. Um, and no, we're we're we're very
good. He's he's brilliant. I was I've
been very very lucky to to work with
him. And Bruce Bennett worked with me
until the early 2000s. And and
Bruce was brilliant. I'm now now working
with NIFA uh Hermanson who's who's a
he's an expert in marov chains in fact
so this is his area and at at the trace
institute we're we're looking to get
some more mathematicians now to to to
work with us on this so happens the
better
>> you're building up a nice uh a roster
over there what's what's a Tuesday like
at the trace institute
>> well um right now it's it's distributed
so we we meet by um Zoom and so forth
because NIFAS in New Zealand and Jayton
is in uh you know he's about an hour and
a half away from me. So it's just the
the normal things. So our meetings are
you know on a Tuesday would be um uh so
Robert Prrenner is in um Shanghai. Yeah.
>> He was a former postoc of mine. He's now
professor at Shanghai. He's working with
us. He's going to be the leader of the
of the of the Trace Institute. He's a
younger guy. I'm 70. He's in his 40s. So
it's it's time for me to make sure that
someone younger is in charge. What's the
mission statement for trace?
>> Well, the mission statement um I've
forgotten the exact mission statement.
So, I'll just
>> I just mean generally what are you guys
trying to accomplish?
>> Yeah, the the idea would be to really
explore this recursive trace logic, get
it completely understood mathematically
and then prove we have nine conjectures
that the trace institute has published
on on the site. The first is that we can
build minkowsky space Einstein special
relativity. That's the first conjecture.
Second one is we can build general
relativity. Third is that we can get the
born rule. One of them is to we can get
the big bang. Another one is that we can
get quantum non-locality. Um
>> you can get non-locality out of this.
>> Absolutely. That's the that we can get
non-locality out of this thing. So we
can get every everything from of of
quantum mechanics. If we can't then
we're wrong by the way. So I I'm saying
these these conjectures are are things
that anybody would require me to do.
Didn't Kristen give you a challenge?
>> Didn't he ask you to to derive the
Shreddinger equation using this?
>> Um well, yeah. So, we have um
>> that would be cool to solve that.
>> Exactly. So, we and that's one of the
conjectures that we can get all of
quantum field theory. So, so not just
shortening equation, but get quantum
field theory completely out of this.
>> Uh Nobella is probably listening.
[laughter] if you can get that working.
>> I mean I mean are
>> is this [clears throat] something that
practically you can release like
>> hey gang we solved this conjecture 4 is
done we're moving on.
>> That's right. That's what that's that's
our goal in the next two or three years
is to get these nine conjectures proven.
That that's the first
>> quantum field theory within three years.
>> Yeah. That's that's our goal. We're
trying to show. So we don't have to show
that that um recursive trace logic re
uniquely gives us quantum field theory.
All we have to show is that it can give
us quantum field theory,
>> but then we want to show that it could
give us infinitely many uh other kinds
of of of structures that are probably
much more interesting than that. So, so
that's what we're up to right now is is
but we have to get this headset. The
idea is that for our scientists, what's
going to impress them is this is the
mathematics and the space-time physics
that we know. So, Einstein's general and
special relativity, quantum field
theory, the Borne rule, big bang, and so
forth. If we can nail that down
um and and here's one just real brief
reason why um I'm sure we can do it.
The set of the space of all markup
chains is computationally universal.
Anything that can be computed by any
drawing machine can be done by markoff
chains.
>> I mean the math makes complete sense to
me. The jump to consciousness I can't
get my mind around yet.
>> Aha. In what in what sense? Um,
I don't know why consciousness is
fundamental is even required for this to
work. I feel like this could work with w
with physicalism just fine. Well, what
you could do, of course, is to say, I
don't like the consciousness stuff. I
just want to have this be a theory of
observers and agents without putting the
the consciousness um label on it. And it
would work perfectly fine.
>> It it so consciousness can just be a
label. It it could. Here's the reason
though why that might end up being
uncomfortable.
I do feel like if you hit me on the
thumb with a hammer
>> that I feel there's something I feel
that I can't ignore and it's it's it's
not it it's a real experience and it's
an unpleasant experience and and then it
feels if anything is real that painful
thumb is is real.
And if that's not real, I don't know
what is. And if the taste of chocolate
isn't real, I don't know what is. The
abstract structure of what we call
spaceime may or may not be real. But my
experience right now is a look around.
That's real.
>> You know, Einstein gives us a
mathematics that we call the space-time
mathematics that that nicely describes
this. Um, but all I know really first
person is I'm experiencing colors and
distances and smells and so forth. And
um Einstein stuff is great and it it
works well, but I'm not sure that that's
the final reality. That that might just
be a description of um my particular
kinds of experiences. What about a you
know a shark or or a bat, you know,
that's using echolocation? Why should
its world be anything like my world? So
there are all these you know sensory
worlds.
>> Explain that to us using uh ants and
hands.
>> Ants and hands.
>> Yeah. how
>> an ant has a certain world that that it
perceives,
>> right? Well,
so I mean
I don't know too much about ants. I
think they use chemical signals and so
>> No, I mean it was just an example that
used I thought was very interesting
about how uh from an ants perspective,
you can just reach down and kill that
ant at any time.
>> Oh, right. Right.
>> That that that analogy.
>> Oh, right. Yes. So very
>> sort of a headset.
>> That's right. So yeah, in fact, so
that's right. So to see that we're in
sort of different perceptual worlds from
other creatures. Yeah. Then it's pretty
straightforward. Yeah. So if I see an
ant crawling around on my on my dining
table, um I can kill it and it won't
even know what's about to happen to it.
I can just put my finger on it and um
you know, my wife probably wouldn't want
to do that. should probably put on a
piece of paper and take it outside and
and try to be kind. But you so but the
ant wouldn't know in either case. Um,
now I should say that's from my
perspective about the ant. From my
perspective, the ant doesn't seem
terribly bright. I mean, it does some it
does some chemical tracing and so forth.
That's and it can do some apparently
dead reckoning. It can wander around and
then dead reckon. So, it does some
pretty smart stuff.
>> Um, but from my point of view, it does
it looks pretty simple compared to a
human. But if you ask yourself from an
ant's point of view, how much would the
ant know about AJ?
>> Nothing.
>> Almost nothing. And it might not even
know that you're there. But but if it
did, maybe its representation of AJ
would be as simple as my representation
of an ant. You know, it's just that
simple to it. So by symmetry, I have to
ask myself, so what I think of as an
ant,
maybe that's just because of the
limitations of my own headset. Maybe if
my headset is dumbing things down so
much that I'm actually interacting with
this incredible intelligence far greater
than me, far more capable than me
>> and you just have a trace matrix of
>> That's right. I'm just seeing a trace of
this thing and I I'm getting an ant. So
you
>> and that's cuz all your headset allowed
you to see.
>> That's all your headset allowed you. And
it also means that there could be this
is where UAP stuff comes in and so forth
that there could be other
consciousnesses or observers if you
don't like conscious other agents or
consciousnesses um or observers that are
in much bigger headsets than ours. And
our headset is to them like perhaps the
ant headset is to us. And we can go down
and smash the ant anytime. and they
could come down and smash us anytime
because we're trivial compared to them.
So the the headsets can go off to
infinity. Um
>> let's take a quick break and we'll come
back and we'll actually talk about those
entities with the with the headsets.
>> Okay. [snorts]
>> So Andrew Gallammore was sitting right
in that chair not too long ago.
>> You guys are working together which I
love. Your first project is literally
titled Are DMT aliens real?
>> Are they? Well, I think he and I both
probably agree that the emphasis on real
is probably a little bit misguided sort
of but it's it's it catches attention.
>> Yes. It gets funding.
>> That's right. So, the idea would be that
um
there are all these different kinds of
headsets that are possible. Ours
is not the final reality. Spacetime is
not the final reality. There could be
all sorts of headsets much bigger than
ours that could play with ours.
people in the entities in higher
headsets could play with us like we play
with ants.
And so the idea then is does DMT
just screw you up,
>> right? make you hallucinate?
Or does it perhaps
somehow
allow us to
modify our headset in certain ways? Open
it up to more dimensions, for example,
maybe more dimensions of space, more
maybe even new kinds of conscious
experiences all al all together.
That's and and if and if so, could the
the entities that we're seeing in the
DMT space um be avatars?
Uh just like when I talk with AJ, I you
know, I'm not directly in contact with
your consciousness. I'm seeing an avatar
we call the human body that's allowing
me with my headset to interact with your
consciousness. And you see the Hoffman
avatar allows you to interact with my
consciousness. So these are these are
avatars but even these avatars don't
exist when they're not perceived. So I I
said earlier that right you we you
render this on the fly. So the AJ that
I'm seeing is in some sense not real
because that AJ is gone
>> gone.
>> It's completely gone. Whereas the
consciousness presumably is not gone but
but Hoffman's avatar of AJ is gone
whereas AJ is just fine. And and so the
question is are the DMT entities good
avatars? So it's not saying are they
real, but are but but we put real in
because it's catchy.
>> So it's not that they're real, but are
they genuine avatars that and one thing
about an avatar is that um you can sort
of share information. So I could tell
you something and then um I could ask
you to tell someone else and then find
out from that someone else whether they
said exactly what I told you. So that's
the kind of question we can ask about
these these DMT entities. Could could um
we send two people into DMT space. Um is
there some avatar there? There's like
various avatars that seem to come up all
the time.
>> They do. I've been in that space. You've
been in the space.
>> I have. And it feels more real than
this.
>> I I I've heard that the resolution seems
higher. Everything seems this seems like
the low resolution.
>> This seems black and white. It does. And
the funny thing is that's what the
recursive trace logic is telling me.
It's just saying this is one of the
cheaper headsets.
>> That's right. I didn't even connect
>> This is we really So when I say this is
a cheap headset, I really mean it that
that we got we we got the the cheap
version and they're much much and
probably even the stuff that you're
seeing in DMT is cheap compared to other
stuff. Don, there are certain
psychedelics that I do that I don't need
reading glasses for like a week. I can
just see. I can just see. Colors are
sharper. my I I just don't need the
glasses anymore. It drives my wife nuts.
She's like, "You can read that." I said,
"I I it's like superpowers for about a
week." It's something about
neuroplasticity that is
I want to see. That's the kind of thing
that we have the chance to really
understand with the recursive trace
logic. If we understand how to build
this headset, we understand what DMT is
doing
to change the headset, then we're going
to be able to understand and reverse
engineer this stuff and and and actually
come up with. Now, NIFA has actually
done quite a bit of DMT or some DMT.
He's a mathematician and he told me that
he went in and saw a tesseract, a
four-dimensional cube rotating in four
dimensions rigidly.
>> Mhm. So that's good evidence from again
it's not proof but it's good evidence
from a reliable mathematician that he
was at least in a four-dimensional space
and was seeing a rigid tesseract for a
fraction of a second. He he didn't keep
it for a whole second but he saw it.
So it suggests to me that it's
worthwhile. That's not proof. I mean I
hard no scientist that's not proof but
it's very suggestive that perhaps DMT in
our headset is a representation that
chemical is a representation in our
headset of a tool outside of space and
time that allows us to change parameters
of our headset. And so as we begin to
what we have to do is prove our nine
conjectures build the human headset once
we've built it and that's going to be a
lot of neuroscience. We're going to have
a neuroscience team working on this.
We're going to have to there's 86
billion neurons, trillions of synapses.
We have to understand that is just a and
this I'll be very slow here. The brain
is a headset representation
of how the headset is constructed.
>> That's a big concept.
>> The brain, the nervous system is a
headset representation of how the
headset is constructed. So, I'm not
getting rid of neuroscience. I'm saying
we need more neuroscience and it's going
to be much harder than my colleagues in
neuroscience think it is. Understanding
the 86 billion neurons and trillions of
synapses is just the first step. The
hard step is reverse engineering it to
understand the software from the
recursive trace logic that is being used
to build the human space-time headset.
Once we understand that,
and it won't be next week,
>> no.
>> We can then ask the technical question,
what is DMT doing to the construction of
that headset? What exactly is it doing?
Is it really just taking us from three
to four dimensions or does it take us to
more? What else is it really doing? So,
we'll be able to re once we do that,
we'll understand the software that's
building our headset. And if you think
about it, once you know the software
that's building a VR game, you can do
miracles in the game.
>> You That's true.
>> Absolutely. You can change any rule in
You're no longer bound by the rules of
the game because you're the master of
the game. You're writing the rules. So
the in Grand Theft Auto, the guy who is
the wizard is wonderful. You can play
all the game by the rules in the game.
But the geek who wrote the code can take
the gas out of the car of the wizard. He
can turn
>> right. He's got this guy.
>> He can do anything he wants. And so
that's what we're that's when so I think
the recursive trace logic when we get
these nine conjectures proven and we
begin to understand how this headset is
designed. You cannot think big enough
about the technologies that will come
out of this. It will make everything
that we've got seem like firecrackers.
And and I'm looking forward to that. I
think that it's going to open our eyes
when we see when people see the
technologies that come out of it. Then
it'll be game over for physicalism.
>> Game over physicalism. Wow. These are
these are big goals.
>> To um to replicate the brain as a
headset. Do you actually have to create
an artificial neural network
>> or do you just do it with proofs? Well,
we we'll we're already sort of right now
um looking at the neural networks and
trying to construct them through all the
like very clean microscopic sections and
so forth. We piecing together what we
can see inside spaceime of our of of the
neural structures and we're putting
together all the physiology of it. So,
it's really really complicated process.
So, we need to to to continue to do
that. But then we have to ask now think
out of the box. think out of spaceime
completely,
>> right? Because you're only the brain.
Even if you got the brain perfect,
you're not seeing everything.
>> No, you're only seeing a compression of
so there's some really complicated
software out here, so to speak, some
software that gets compressed in this
headset into what we call the brain. And
it looks really complicated inside this
86 billion neurons, trillions of
synapses. Looks really complicated, but
it's trivial compared to what's outside.
Completely trivial compared to what's
outside. This is all restricting. Our
headset is really restricting, funneling
down, losing information. So, we need
more money for neuroscience, not less.
>> A lot more.
>> Yeah. We're living the compressed data
stream. We're not getting a lossless
data stream.
>> That that that's right. And that's what
you experienced yourself in your own
your own experience in DMT is that this
is already feels like the lost
>> the lossy interface.
>> It does. Um, so you you brought up UAPs
earlier. So is that what they are? Are
they objects in a higher using a better
headset?
>> I don't know. But I can say that
the recursive trace logic
leaves that possibility open and gives
us a rigorous way to begin to think
about it. Um it it one thing about the
recursive trace logic is as we talked
about earlier, it's about attention. If
I have the 10 x 10 of colors and you
only see the 3x3, you're you're only
paying attention to three of the colors,
there's all sorts of magic that can
happen outside in the other seven
colors. And when you look at the trace
construction, you have your visible
states and and that matrix, then there
are transitions from the visible into
the invisible. So in the markup matrix,
the big matrix, you have transitions
from the visible to the invisible. Then
you have transitions among the fully
invisible part, the 7 by7, and then you
have transitions back in. So you have
the exit,
the external world, and the re-entrance.
>> Wow. So that that could explain
everything from UIP teleportation to
particles appearing out of the vacuum.
>> That's right. This gives you So it shows
you how to exit from our space-time
headset, how the you could have an
entire world that's far more complicated
than our headset, infinitely more
complicated, and then re-entrances.
That's one aspect of it. Another aspect
is just the magician's trick.
Lot of magic that happens in in magic
shows is by manipulating your attention,
right? I do this and you automatically
have to go over there and and now I'm
doing something over here that you
>> So the recursive trace logic is the
logic of attention and all you So one
way that you can have these things
happen is just distract and change. So
again, the recursive trace logic is all
about. So there all sorts of tools that
once you're higher up in the recursive
trace logic that you could use to, you
know, hoodwink people, you know,
consciousnesses that are stuck with
smaller headsets, all sorts of tools. So
I'm I'm really excited to to look. But
right now then we have the tools that we
wouldn't have in spaceime. If if you're
stuck in spaceime, we have Einstein's
theory of gravity. We we're trying to
get quantum gravity, but
there's with gravity you cannot do what
these UIPs seem to be doing. No.
>> Hover with no apparent propulsion. Move
at Mach 40 instantly. Um at 466 GS,
>> go into the water with no displacement.
>> Go into that. That's right. That that's
that's just not possible with our
physics. But it's certainly possible if
you know the software of our headset and
you play with the software.
>> Right. So, we're the ant then.
>> We're the ant. That's right. And they
they if they understand the software of
our headset, it's just like us in the
the the Grand Theft Auto. If you're the
geek that wrote the software, the wizard
will be stunned by what you're doing to
him. He'll have no way. It doesn't obey
his rules. It doesn't obey his laws.
He'll say, "This is impossible." And
it's not impossible because
the geek is not stuck in the headset.
The geek is making the headset. So when
we may be dealing with UAPs, we're
dealing with perhaps higher levels that
understand how our headset is built and
can play with the software rules.
>> It feels like they're doing that. So a
UAP doing Mach 40 is just hitting that
clicker faster than we can perceive it.
>> That That's right. Or it could be
distracting us. They're they're So
>> there's a trickster element to it for
sure.
>> There could be a trickster. And that's
again what I was talking about earlier
about the Leela kind of thing.
>> It it's it's they could be playing with
this. It it's very very clear from the
technologies if they wanted to destroy
us we would be helpless.
>> Oh yeah. Completely helpless. So so it's
it's not like there are cases where it
seems like there's some kind of
hostility and so forth. But if they were
really hostile,
>> we wouldn't last 5 seconds. So I get the
feeling that there is again a
a a playing aspect of this. Let's
explore more. Let's play with these. And
and if it really is the one
consciousness looking at itself through
various headsets, it could be the one
consciousness saying, I I put myself on
these really stupid human avatar
headsets. Really really I I want to give
them a little prod from a little higher
headset to to sort of wake up. So it's
the the one source consciousness playing
with itself at at various levels. But
with the recursive trace logic, we could
begin to scientifically see how that is
is being done. But I would put a humble
note on the whole thing. I think the
recursive trace logic is just a
scientific theory. It makes its own
assumptions and I look forward to
replacing it at some point. So that
we'll need to go be beyond the recursive
trace logic. But for right now, I think
it's a good next step forward. But
whatever conscious the fundamental
reality is what we might call the source
it it transcends any scientific
description including mine. So so the
what I would say the recursive trace
logic is really the logic of infinite
number of perspectives
that it's the logic of an infinite
number of perspectives that the one
source can take on itself. So they're
all consistent perspectives but they're
all mutually can be mutually
inconsistent but each one by itself is
consistent. Each each boolean logic is a
consistent view, but different boolean
logics can contradict each other, but
they're all useful perspectives on the
source which transcends all of them.
>> That's beautiful. If we if we have
evolved to not be able to perceive that,
is that an impediment to the research?
Is there maybe does the source just say
I'm not going to let him figure this
out?
>> Yes. So you and I are just the source
looking through a particular headset and
there are certain things we know that
the game is such that there are very few
of us who I mean there are the Einsteins
are very few right [laughter]
>> right
>> the the Schroingers Einstein and so
forth are very very few and we we're
grateful for all of them
>> we are
>> um but
they're trivial the the Einsteins and
Schroingers are trivial compared to
what's out there to be discovered. Yes.
>> And yet you and I are fully that source.
And we in silence
go into the space of infinite
intelligence that that it is. So all of
us in some sense are that infinite
intelligence. Um, and we've chosen
to play in this avatar and we've chosen
to allow ourselves to have limitations
to actually just take this perspective
very very seriously to look at reality
to look at source through this
particular lens and to play. So really
it's it's we take ourselves very
seriously. It's all very very serious
and so forth but really it's about
playing with this perspective. really
enjoy this perspective and then let go
of it and realize this was just a
perspective and you infinitely transcend
it.
I think I have learned your philosophy
and it comes out of your your heart
scare after co you
>> co you're you're running heart's gone
190 a minute for 30 hours you're in the
hospital two surgeries you text your
wife goodbye
>> what was in that text
>> um uh it was very very short because I
was you know I just was in a horrible
place I I was exhausted I couldn't
hardly think I I sected my wife and then
I I think I texted my daughter
separately. Um I I just said um I don't
think I'm going to make it. I love you.
Goodbye. So it was that was just that
simple. Same thing to my daughter was
something like that. I don't I my
heart's been beating 190 beats a minute
for 30 hours. I don't think I'm going to
make it. Um um I love you. Goodbye. Um I
I
as soon as I did that I laid back in the
bed
to die just to wait for it to happen.
>> You were at peace in the that moment or
were you scared?
>> I was scared.
>> I was exhausted. My heart was racing. So
even just the racing 190 beats a minute,
it feels like you're you're you're
scared to death because your your heart
is racing. Yeah. So, even though it's
just atrial fibrillation, it feels to
you like you're you're scared to death.
Now, I've only been recently starting to
talk about this. At that very moment
when I leaned back, it was about 3:00 in
the morning in the hospital.
I saw these two male figures standing
next to my bed. I hadn't seen them
before. And one of the male figures
looked to the other, smiled brief
lightly, and nodded his head.
And then all of a sudden I felt an
incredible warmth in my heart and my
heart started beating normally.
>> What?
>> I I And then so I looked down and I
looked up and I didn't see them anymore.
And neither of them was wearing hospital
garb.
>> I haven't heard you talk about this part
of the story.
>> No, I haven't done it publicly before.
>> What do you Who do you think they were?
>> I have I didn't talk about it for
several years because I
This is so strange that but I realized
in the last couple years it's like okay
if they were hospital people it would be
malpractice
to have waited 30 hours to give a
patient the drug that they knew would
take care of me. So that didn't make
sense. Second when I felt it I didn't
feel the warmth in the IV. I felt it
directly in my heart. I didn't feel
anything in my arm.
And third, I'd never saw those guys
before or since.
So, so I don't know what to make. So,
I'm just stating what what I saw and I'm
here because my heart flipped from 190
beats a minute for 30 hours to instantly
like instantly a feeling of warmth and
it was done.
>> What did you say to the doctor when he
came in and said, "What happened?"
>> I was so exhausted that I I I couldn't
even gro what what was going on. All I
knew was I was so glad I wasn't dead and
so glad that I my heart wasn't beating
like that and I just needed some rest.
So, and then I I could not wrap my head
around it. I tried to think out of the
box, but that was so far out of the box
that I just let it go for three or four
years. I didn't even think about it. I
didn't
>> I believe you.
>> Um here's where I think your philosophy
comes in and it's it's that from this
story to just for laughs gags.
>> Oh, yes. That's your That's your show
that you love, right? I
>> Yes. How'd you know that? I I love Just
for Labs. It's one of my favorites.
>> Why? What happens on that show?
>> Well, people are surprised. So, they you
they're set up to expect one thing and
um they might get all upset and then in
many cases they end up laughing. So,
they they get into the situation,
they're upset, they're all tense, and
then they realize, "Oh, wait, wait,
wait, wait. Ah, I don't need to be upset
at all." That was just a gag the whole
the whole time. And it really does the
just for laughs gags sort of as a is for
me a good uh example of what I think the
whole thing is. Again, this is the whole
point is to relax and enjoy the game and
and in fact the more that you can do
that and I I actually think about this
in terms of my research progress. the
more that I can relax
and enjoy
and be open.
Let go of everything I think I know.
Don't be any like egoic attachment to
what I think I know. You've got to the
ego is the biggest biggest impediment to
growth and and discovery. So, let go of
the ego, let go of attachment, let go of
all that stuff and be opened like a
little child. Jesus said, "Unless you
become like a little child, you cannot
enter the kingdom of heaven." And I I
really think that that's that's right.
You have to just be open to say, "I
don't know anything. As much as I've
learned, um, I know 0%. Let me know
more." And and play the game. I think
that's what it's about.
>> I think so, too. And don't take it too
seriously. And when it's over, you might
be scared, but then when it's finally
over, you can just have a big laugh.
>> That That's right. And even being scared
was part of the whole deal.
>> Yes. Then you go you can go back and
laugh at even being scared but you go
through for some reason we set it up so
we want to go through that scared part
too.
>> We did.
>> That's right. It's interesting. I was
scared. I mean there's no I'm no hero. I
was scared to death and I was really sad
to say goodbye.
>> Yeah. Um
um so I think for the the one
to really know itself it has to the
source it has to take an infinite number
of perspectives and to take a
perspective doesn't mean just sort of
casually. It jumps in with both feet all
in on that perspective to really believe
it's that and it really then lives that
out and and then slowly realizes that
was just a perspective and it wakes up
from it. But that's how it really learns
that that perspective as rich as it was
I infinitely transcend it and that's how
that's the best story I can tell right
now about how the the one is why how it
knows itself and why it does this.
>> I think it's perfect. So now given
everything that you've done, everything
that you know,
when you see a monarch butterfly, is it
more beautiful or less?
>> It's it's very beautiful. There's a a
preschool just down the street from me
that has a big picture of a child on a
poster looking at a monarch butterfly.
And every time I go by, I go, "That was
me at 5 years old." Don't lose the love
of surprise and the joy of of discovery
of all the beautiful stuff that's around
you right now that you're just not
seeing. So I I love to see that that
picture that it's a boy um looking at a
butterfly and it's there for you.
>> It's there. It's there for me reminding
me um relax Don don't get uptight.
>> It's just a game.
>> That's the message. Don Hoffman and the
Trace Institute. Everything will be
linked below. He's got to catch a
flight, but I hope you you'll come back.
You're a treasure.
>> I'd love to. I thank you and thank you
so much. This has been a great pleasure,
AJ.
>> Thanks, Don. Bye, everybody. That was
Donald Hoffman. He ended on a butterfly.
So, let me start with the other bug. The
beer bottle beetle is real. [music]
Endmologists Daryl Gwyn and Dave Rents
watched males in the outback mating with
stubbies until they died. And their 1983
paper won a Nobel Prize. [music]
Australia changed the bottle for the
beetle. Evolution built them to like
dimples. Not true. That's Don's whole
argument in one insect. And I love it.
Now, the new math. Don says proofs of
special and general relativity built
from consciousness [music] alone are
sitting on his desk. Nothing is
published, so there's nothing to check.
He said it himself. It's not real until
it's published. I can wait. In the
hospital, Don's heart raced out of
control [music] for 30 hours. He texted
his wife goodbye and laid back to die.
Two strangers stood by the bed at 3:00
in the morning. Neither wore hospital
scrubs. One smiled and nodded and his
heart flipped back to normal. He [music]
never told that story publicly before. I
had never heard it. 40 years of math to
prove reality is a headset. And the
takeaway he cares about most is a
preschool poster of a kid [music]
staring at a butterfly. That part I can
verify. I watched mean it. Don's book is
The Case Against [music] Reality. The
new work is at traceinstitute.org and
everything I mentioned is linked below.
Until [music] next time, be safe, be
kind, and know that you are appreciated.
[music]
[music] I played Philippus in 51. A
secret code inside [music] the Bible
said I [singing] would.
I love my UFOs and paranormal [music]
fun as well as music. So I'm singing the
like I should.
But then another conspiracy theory
becomes [music] the truth my friends
and it never ends. [music and singing]
No it never ends.
>> I feel the crap guy. I got stuck inside
Mel's home with MK out of [music] being
only true.
Did Stanley Cubri fake the moon landing
alone on a film set or would shadow
[music] people [singing] there?
through well just fought the smiling
[music] man I'm told and his name was
cold and I can't agree [music]
I'm dancing with the fish
on Thursday night J
[music] all through the night
all I ever wanted was to hear the truth
all through the
>> The math man silence and the solar storm
still [music] come to a gather the
secret city underground.
Mysterious [music] number stations,
Planet Circle, Project Stargate, and
what the Dark [music] Watchers found.
We're in a simulation. Don't you worry
though. The Black Knight sat a light
[music] to me. So I can't believe I'm
dancing with the fish
on Thursday nights when they chase you
and we all through the night.
All I ever wanted was to [music] just
hear the troops of one
through the night
[music] on Thursday night when they
chang.
All I ever wanted was to [music] hear
the truth. So with me all through the
night
loves to dance
loves to dance [music] on the dance
floor because she is a camel. Camel love
to dance [music] when the feeling is
right
in time.
