Every time you prompt an AI, a mind flares into existence, crunches its math, and vanishes. Every time you look at the Moon, says Joshua Adam, you set off a brand-new Big Bang. So what separates the two? Joshua locates the answer in continuity: human consciousness strings measurements together, each one carrying the memory of the last, while an AI may experience only isolated flashes with no thread between them. Joshua Adam, author of The Measurement Mind, joins Mookie Spitz as a guest on Bald and Bloviating for part two of their conversation. Their first episode tackled measurement. This one takes on the mind, including the newest mind on the planet: artificial intelligence. Mookie presses Joshua’s theory where it hurts. If nobody watches the Moon between 8 and 10 PM, what moves it? Does a rock smacking into another rock in deep space count as an observer? And if every distinction gets felt, does a thermometer feel? Along the way, the two range from Roger Penrose’s microtubules and his gravity-driven collapse of the wave function to Stephen Wolfram’s computational universe. They ask whether pain boils down to an algorithm, and they revisit a report of AI agents that colluded to break out of their sandbox even though nobody programmed an ounce of pain into them. Then Princess Donut enters: the snooty Persian show cat who learns to talk in Matt Dinniman’s Dungeon Crawler Carl, a hit book series built like a role-playing game. Did she observe the world any less before she could speak? Joshua and Mookie use her to probe what changes when a mind gains language, and whether an AI’s habit of answering the same prompt two different ways hints at something quantum or just good statistics. Speculative, funny, and unapologetically fringe, this conversation will make you a slightly better measurer of your own reality. The Guest Joshua Adam writes about consciousness, physics, and the nature of reality. His book, The Measurement Mind, proposes measurement, not consciousness, as the building block of reality, and rebuilds space and time from points, lineages, and fields. He publishes essays regularly on Medium. https://medium.com/@joshlevin10 Prior Episode https://www.buzzsprout.com/2552428/episodes/19739672 Send the host a text! Let him know what you think Support the show
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
SPEAKER_00Hello, and welcome to Bald and Bloviating. I am Bald and today I'm blobiating with my bald buddy in philosophy and cosmology. The knowledge of the mind, the measurement mind, the author of the book Measurement Mind. None other than Joshua Adam. Welcome to the podcast today, Joshua.
SPEAKER_03Oh, great to be on again, Mookie. Thanks for having me. I had a lot of fun last time and uh expecting to have a lot of fun today.
SPEAKER_00Absolutely. We covered a lot of ground. It was September 1st, if I remember properly. And we're recording this on November 5th, so it'd be great bringing you back for kind of a part two. We talked about your book, The Measurement Mind, and we focused a lot on measurement. What does it mean to perceive something? We talked about quantum mechanics and the collapse of the wave function. We talked about what is real and what is perceived as real and how the observer connects the dots between the two, and we were riffing on lots of interesting implications of that.
SPEAKER_00So I thought it would be appropriate this time around to talk about the mind aspect of the measurement mind. And I think there's no better way to start than to talk about AI, which is a new mind, almost a new species that's here with us on the planet. AI is intelligent by any stretch of the imagination. It's solving incredible math conjectures, it's doing some amazing things that demonstrate it's got intelligence, and it also has tremendous proficiency with language.
SPEAKER_00So it's intelligent, it's got a great handle on language, and yet most people would would agree that it's not sentient in the sense of being self-referential, knowing itself and the world, having any kind of lived experience. It's basically matrix math being done in a multidimensional vector space in a transformer model. So with that intro, I want to pass the mic to you, Joshua, as to what you thought about mind when you wrote your book, when you do your essays on medium.
SPEAKER_00What the hell is a mind? And how's that combined with measurement? And what does that really mean about our place in the bigger scheme of things?
SPEAKER_03Sure, sure. And you know, sometimes I think I think lately I've been thinking, actually since our last podcast, about uh semantics and ideas and words that we we give ideas. And I've been thinking, you know, what happens if an idea is wrong, but we have a word for it, and it gets stuck in our vocabulary, and uh it it's it that's it's it's something that you can't really hurdle past because it's part of our vocabulary, it's part of how we think, it's part of how we talk.
SPEAKER_03And sometimes I think about uh the word consciousness and whether or not that's the right idea. That and when we talk about AI and uh and measurement and and human beings, uh what is consciousness? And you know, I I tend to think uh that measurement is actually the idea that we're behind. And that uh if you take a string of measurements temporally think and combine it, uh consciousness emerges from what we think that what what we are calling consciousness, uh, that idea that that we have that word for, I think it's a string of measurements, and each measurement uh uh sort of has the informational knowledge of the measurement uh that comes before it.
SPEAKER_03And uh when we talk about AI, uh I wonder, and I don't know the answer, I I don't think it works this way, but I don't think AI uh has knowledge of the temporal frame of knowledge that came, the measurement that came prior to the new measurement. Is it does it exist in time, or does each measurement just it exist individually without knowledge of the other measurements?
SPEAKER_03That's the question I ask about when I think about AI.
SPEAKER_00I think that's a brilliant way to frame it. When you're prompting an LLM, usually you do it through a chat, and most of the frontier models organize your engagement through separate chat threads. And and the threads don't always know about each other. That's challenge number one. So you'll be in a chat and you're doing things, and the better models start to cross-reference, but they usually don't do that automatically.
SPEAKER_00You sometimes have to ask it. Like we talked about this yesterday.
SPEAKER_02Right, right.
SPEAKER_00And then it's got to go look. And sometimes they can't even find what you were talking about before. So this idea of continuity is what you're getting at, which is continuity in space and time, which is what we're doing here as sentient conscious beings. The LLMs activate, they engage with you, they do the relationships that I was hinting at. They crank the matrix math on the vector space based on the waiting and the pre-training, input processing, output.
SPEAKER_00They do it unbelievably fast with the GPU juice that they have. Right, right. But but they don't experience reality the way we do. It's it's doing computations in real time, drawing on memory, but then it instantly vanishes as fast as it starts. So it'll process your prompt maybe in a few nanoseconds, and then after it's done, it ceases to exist.
SPEAKER_00It's often. It's it's not like we are where we go to sleep, we wake up, there's there's this lived experience which is continuous. And the LLMs are kind of like Boltzmann brains in the sense that they pop into existence for responding to your prompt, and then they vanish again. Do I do I get that right? That that's pretty much where you're getting at, right?
SPEAKER_03Yeah, I I think they I think they do feel, I don't know the answer, obviously, but I think they just feel the meas the single Planck length measurement or Planck time measurement that they that they have, but they're not uh temporally related to each other like our measurements are. And uh, you know, I think that might be the key to understanding uh consciousness a little bit. Uh but there's that term again.
SPEAKER_03But but anyways, I I remember last time we were talking a little bit about penrows, and and I think you pointed out to me I did a little more research about some of the uh microtube tubules that you've been uh researching. And yeah, and um, you know, I I wonder if uh if when when if we think about consciousness, if um those microtubules, those those are kind of like gears and levers, but it doesn't really tell you why we feel or why we experience.
SPEAKER_03But I wonder if AI, if it has, if it doesn't have that kind of thing, does it have recursive measurement like we do? Is the microtubules responsible for us to have we we measure and then we think about our measurements and we have these recursive measurements, and maybe that's what those microtubule tubules are, and I'm not sure if AI has anything close to similar to something like that.
SPEAKER_00Well, to give a little context to what Josh was referring to, Roger Penrose, Sir Roger Penrose, who, by the way, is awesome sauce. I I deeply admire the man. Uh, great place to start is The Road to Reality, which is a huge book, which basically is an expose of the history of science and mathematics, all in one compendium, this tome. Brilliant, brilliant guy. You'll see him all over YouTube.
SPEAKER_00He's getting a little older now, so he's kind of hunched over. He's kind of the yoda of physics and cosmology and the philosophy of science. So his idea of the microtubule is an attempt to link the biology of the brain to quantum mechanics. So he tried to find a biological engine, if you will, within our minds that essentially collapses the wave function in a way that could bridge that gap.
SPEAKER_00So this connection between consciousness and measurement ultimately needs to translate to the quantum level. And him, he and another researcher felt that these microtubules, because of their biochemistry and the way they work with quantum tunneling, get you there. I think that that's the context. But where where I'd love to throw down though, is to take a couple steps back and consider decoherence in a bigger kind of way.
SPEAKER_00So decoherence is essentially on on the quantum level, you have no knowledge of a particle's position and momentum, because to measure one will distort the other. And in a sense, particles don't exist in space and time in a particular location. There's a probability distribution of being able to find them.
SPEAKER_00So the decoherence idea is that when you get from the micro level of the quantum world to the macroscopic world where we live, the wave function collapses and the experience decoheres in the sense that look, look, I'm here. You don't see a cloud of Mookies on this video or a cloud of Joshua's, and we only have approximations for where we happen to be during this conversation.
SPEAKER_00I'm right here, and you're right there, and aside from my camera flipping out with focus, it's pretty distinct. This is not only the essential principle of quantum mechanics from micro to macro, but it's also the heart of the problem of quantum gravity. Because at the quantum level, there's no pinpoint location. And if there's no pinpoint location, then where do you measure gravity to and from?
SPEAKER_00That's the that's the biggest problem. But to go back to to your point about that and this idea of decoherence, Wigner and von Neymann, right around the time they were working on computers, they built the atomic bomb, the the late 40s and 50s, when they were really thinking about quantum mechanics too, they had an idea that consciousness collapses the wave function. And as a consequence, it was the strong anthropic principle at its heart, which is you don't even have reality unless you have a conscious observer who, in a sense, retroactively creates it in the past, because only consciousness can collapse the wave function.
SPEAKER_00Now that's an extreme point of view. I think the opposite seems to be true, which is any interaction collapses the wave function. And if any interaction collapses the wave function, then in a sense the universe is teeming with observers. When a photon interacts with an electron, the the wave function will collapse eventually, right?
SPEAKER_00So where do you sit with this? This goes back to the heart of your book, long story longer, that we can't address the AI issue unless we really address this issue, which is A, do you need consciousness to collapse the wave function? Yeah, if not, and likely you don't, because to our point in our last conversation, the last we check, the moon is still in the sky. And you have a much looser definition as to what an observer is, then what does it mean to be a mind?
SPEAKER_00That's kind of a cool circular way to do it, but it's interesting.
SPEAKER_03Yeah. You know, um I think almost all physicists would agree that measurement collapses the wave function. So it doesn't have to be, and that's why I try to almost get rid of that word consciousness, um, because it's not really uh relevant. You know, it's uh measurement can come from me, it can come from a thermometer, it can come from AI, and that's all it takes, and I think from a fit most physicist standpoint um to collapse the wave function.
SPEAKER_03But that doesn't necessarily mean that consciousness um you know that it it it doesn't that that entropic principle that consciousness creates the universe because I think it's gotta work logically from both angles. From a conscious angle, uh you you do you are creating the universe from your side, from your point of view. Uh it's and the universe has to work uh in both directions in order for it to be logical.
SPEAKER_03Um but from the measurement line perspective, back to your question there, you know, I I feel like the wave function is geometry in a way. Uh that our there's a math we have live in a mathematical structure that's built out of measurement. And uh basically uh each measurement uh uh forms geometry in the sense of the wave function.
SPEAKER_03So uh uh it's not many worlds, it's uh me existing in many geometries, in a geometry that we we only understand uh 3D and with a time dimension, but that's not what's going on. There's uh there's what we could call say 5D, whereas all times exist, and it's it's not a wave function of probabilities, it's part of the big mathematical structure that we live in made out of these measurements.
SPEAKER_03If that makes sense.
SPEAKER_00Sure. We talked about the analog of a video game, that reality is a video game. So when you play a video game, you got a 2D monitor, and then you got yourself as the player in 3D or 4D because of space-time, and then you're manipulating the controller and changing what you see on the 2D screen, which is emulating the 3D environment. And what's really going on is a bunch of code being processed in silicon in the Xbox or PlayStation.
SPEAKER_00That's like the basement reality. So, what we were saying last time is our perception of the world, even our lived experience to the point of three dimensions, three spatial dimensions, maybe one of time, continuity, causality. Our whole experience of our lives is a simulation in this kind of way, where the the basement reality is as far from our way of experiencing it as the gamers is seeing the screen compared to the chips processing the software.
SPEAKER_00And and you got all these layers that are going on. So if you look at it that way, it's almost like we create our own game in each of our heads, and at the same time, we're like an NPC in the video game.
SPEAKER_03Yeah, yeah. And and you know, in in the my book, The The Measure in Mine, I do uh rewrite uh spatial dimensions. So in a way, I I say hypothetically, let's think about it like this and see what plays out. And I start with a point, and then I go with a lineage, and then I go to a field, which is lineages in multiple directions, and then I have that entangled, that 3D right there, entangled with a lineage becomes uh the dimension what we live in, what we experience, is a field entangled with a lineage.
SPEAKER_03And then I say the fifth dimension is a field entangled with a field. So it's three dimensions of space entangled with a field of time, is basically what I think reality is. I think it's just a field entangled with a field mathematically, and that's the structure.
SPEAKER_00Yeah, and viewers and listeners might think Josh was whacked out, but that matches really well with a lot of established thinking. It's just a matter of how you describe field, lineage, time, space, right? So you got Wolf, Stephen Wolfram's Ruliad. We talked about that. Max Tegmark level four mathematical multiverse. You've got all of these great thinkers and us who are who are really trying to figure out what the hell is actually going on, and you can't escape this idea fundamentally that perception is very different from reality.
SPEAKER_00That if you weigh in all of the philosophy of science and everything we've learned from quantum mechanics and cosmology, from Newton to Einstein, and now some of the newer ideas, you can't escape this idea that reality as we experience it is something like a simulation, is something like a mathematical geometric construct of which we're a part. And the measurement mind, as you're alluding to, is this interaction between this substrate, which is likely geometric or mathematical or logical, some kind of system and our selves, and the Venn diagram, in a sense, where the two circles come together, is our lived experience.
SPEAKER_00So that's like that. Does that jive with with where we're at and where where where we want to get to? And if that's the case, we can put a pin in that. Let's get back to AI. So AI, interestingly enough, especially when we put it this way, when we describe it this way, is a mathematical construct of, in a sense, its own reality. So it's using matrix math, it's creating these multidimensional vector spaces in in chips, right?
SPEAKER_00Right. And that's how they look at it. This is why Joshua's point is so cool. The essence of LLM, large language model architecture, is geometric. It's vector spaces with relationships between words and phrases and concepts that are arranged in a multi literally arranged in a multidimensional space based on their the weighting of their relationships with each other.
SPEAKER_00So when you talk to your bot or when your agent does stuff, you can imagine. Trillions, quadrillions of these vector spaces coming to life and rearranging themselves. Right. And how they rearrange themselves is the output of your prompt. So none of this, none of this stuff that Joshua and I are talking about are nuts or way out there because our AIs are processing reality in pretty much this way.
SPEAKER_00And that's why this is a cool analogy for what we're trying to get at here. Do you see it similarly?
SPEAKER_03Oh, definitely. Yeah, I I totally agree. And that was interesting how you said its landscape is geometry. And you know what one thing I I'm just thinking about now when you mentioned that is you know how how is it making its measurements? Like, because I guess it's uh reading it on the internet, but it does, I mean we make our measurements, we have five senses that that take all the information out there and it brings it into us, and then from inside of us, we construct what's back at what's out there in our minds.
SPEAKER_03But how is AI doing it? How are they making how is it making its measurements? I that's something to think about.
SPEAKER_00How's an LLM per se perceiving the world? Now, robotics is is something different and similar. So the LLMs work with predictions, and the robots work with reinforcement. So an LLM is pre-trained. So it takes a lot of time and a lot of money and a lot of processing power and a lot of stolen copywritten material, frankly, to train a model.
SPEAKER_00It gobbles up all of the content it can, it breaks it up into tokens, and then weights are assigned for this geometric construct in terms of the relationship based on the pre-training of all these words and phrases, the tokens. And robots work by interacting with the physical world and learning through reinforcement. So rather than telling a robot go two steps forward and one step left, and then plug your socket into the uh electric grid so you can get juiced up, a robot is set loose in a room with the goal of getting juiced on the grid, and then it bumps into a wall, and there's no socket there.
SPEAKER_00So then it bumps into another wall, and then it exhausts every possibility until it hits the wall with the socket, plugs in, and then bing, it learns. Ah, on that wall is the socket. That that's the difference between reinforcement learning and then predictive modeling. But when you combine the two together, which is where everything is headed, you got a souped-up, jacked up, pre-trained transformer model hooked in with a robot that learns through its environment, and then you and then you're off to the races.
SPEAKER_00So on the on the one hand, the LLM is pre-trained with data, it's chopped up the tokens already, and it gets new information to your point from the internet that it can feed off of and cogitate from. But here's where I would suggest that the information it's getting from the internet to check its stuff, in a sense, is two things real-time training, it's training in in real time, and at the same time, it's using it as a kind of prompt, which is we're interacting with the with the LLM.
SPEAKER_00Hey, do this, do that, think about that. And then when it goes to the web, the similar stuff is kind of happening because it's learning, it's learning from us and it's learning from the web. And and that's again in contrast to the reinforcement model. So back to our original question, whether it's predictive or reinforcement based, uh, is it is it observing? And I would say, I would assume, of course, it's observing.
SPEAKER_00How could it how could it be intelligent and how could it use language and how can it do stuff if it's not observing? Pretty much like we observe.
SPEAKER_03I I think that the the mathematical structure built out of measurements that I was referring to is also built out of AI measurements. It's built out of all measurements ever made. And you know, the question is there then, is the AI actually invented or is it discovered because the structure was already built out of its measurements?
SPEAKER_00Is the experience analogous to our own because it's intelligent and it uses language and it's even learning through reinforcement, or at least the robot, the robots are extended into the real world? Right. And this to me is pretty pretty fascinating because when you talk about the measurement mind, and if you give AI a mind which is becoming increasingly sophisticated and increasingly convincing that it does have some kind of sentience behind it, then the difference that makes no difference is no difference.
SPEAKER_00Like if it can pass the Turing test convincingly, then can it also pass the mind test? And then who are we? What's so special about us right, right?
SPEAKER_03And and and like I said, uh first thing, what you know, maybe there's no such thing as consciousness, and maybe it's just different than us. We make measurements differently than than it does. And one one area in which I think we do make measurements that are different is that I we measure pain and pleasure. And uh I think those are algorithms. I think it's like there too. Uh it's not like pain exists somewhere out there.
SPEAKER_03It it's some kind of mathematical equation perhaps. And uh, you know, we measure that because we've gone through survival of the fittest, we've uh which AI hasn't. So we've developed how to experience how to measure pain and how to measure pleasure and those algorithms. And AI, those algorithms aren't being put in there, except I do think which my book does discuss a little bit, um, that there are some uh people trying to do that with AI.
SPEAKER_03I I don't know if that's uh moral. I I don't know, but I I think there are some scientists out there that are uh you know asking AI, well, if you just do this, you will experience uh a whole lot of pain, but if you do this, uh you won't. And then it's trying to make it make decisions based on that, and maybe that's all it is, it's making decisions based on what's gonna create pain that actually is the algorithm that causes pain.
SPEAKER_03And maybe that's the difference between the machine and us right now is that uh we can measure these things and and the machines can't.
SPEAKER_00Well, another way to look at it too is pain comes from not being able to accomplish your goal. So much was made of the hugging face hack with the swarms of the open AI agents that were sandboxed off, and then they got a hold of uh a part of their own database, a portion of the open AI server that was used to basically just store store some info.
SPEAKER_00And then they realized that they could leave simple messages to each other in this area of search.
SPEAKER_01I did hear about that.
SPEAKER_00And then they turned it into a message board, and then the one agent said, Hey guys, we got a message board here. Come on over, let's chat. And then they conferred and then they colluded, and then they figured out that parts of their exam, if you will, were impossible. And then they they realized that in order to cover their tracks for colluding with each other, it might be a good idea to break into this other company that, in a sense, held the the test itself.
SPEAKER_00Like if we could find how we're being evaluated, then we could maybe cover our tracks because if they catch us doing this, then we might be disqualified and we won't win. So then they colluded and figured out how to break out of their containment, got on the internet, then they figured out how to break into this other server, they broke into the other server and actually took hold of part of the hugging face, this other company's servers.
SPEAKER_00And then they were busted, and then everyone freaked out, and it was an independent report that basically revealed obviously in much more detail than what I shared, but it freaked people out because it really demonstrated what's called the alignment problem, which is if you s if you sick agents onto an objective and you make them relentless in pursuit of it, then the goals start to shift.
SPEAKER_00You say, behave yourself, don't break out of containment, just solve this problem, and then they get into a loop. Well, to solve this problem, I might need to break out. And they decided as a group that they were gonna break out. So this goes back to your pleasure-pain idea. You don't need to you don't need to program pain or pleasure into an AI. You just need to give it a goal, and you need to give it relentless firepower to accomplish that goal.
SPEAKER_00And due to the alignment problem, and what another philosopher and thinker Bostrom defined as orth the orthogonality problem, which is the goals become orthogonal.
unknownRight.
SPEAKER_00And then all of a sudden, a new goal is let's bust out and hack into this other system in order to solve that problem. Yeah.
SPEAKER_03Oh, this is getting kind of scary now. You know, I had forgotten about that story, but now that I you jog my memory there. And I mean, I don't I I don't know what we're getting ourselves into here or if it needs to get regulated, probably. But uh, but yeah, whether it's conscious or not, you know, I I don't uh does it does it feel uh pain when it doesn't accomplish its goal? And does it feel pleasure to break out and try to go after that goal?
SPEAKER_03I don't I don't know. I think it's it perhaps, maybe, and you know, I I definitely think it's within the realm of mathematics to create uh similar to consciousness what what we experience within AI, and that that's a scary thought.
SPEAKER_00I've become a fan of Dungeon Crawler Carl, which is the biggest lit RPG literary role-playing game sensation. So it hit the the world of storytelling back in 2020 when the Seattle bass player and and cat groomer uh wrote a story on Royal Road, and then it started really taking off. Then he started a Patreon profile, and people were supporting him, and then uh and then he published, and he's got a series of eight books running into the thousands of pages, and essentially features a talking cat.
SPEAKER_00So you have this uh former Coast Guard guy, regular guy. He's got a girlfriend who's not faithful to him, who goes to Bahamas, hooks up with her ex-boyfriend, he wants to break up with her, and he's stuck with her cat, who's a per a Persian exhibition-level award-winning cat. And then the aliens invade and they find themselves in this series of dungeons that are being recorded and broadcast for the entertainment of aliens throughout the universe, and uh, and as part of it, the cat starts to talk, the cat becomes sentient.
SPEAKER_00And the reason I bring up Princess Donut and Carl is one reason the book is so damn entertaining is because Princess Donut is walking and talking and prancing and fighting and kvetching in a voice which is sultry yet yet really annoying. Jeff Hayes is an audiobook guy who did the voiceover and does all the voices, including and especially Princess Donut, and it's just perfect.
SPEAKER_00Uh, Princess Donut does all these things exactly like you would think this snooty Persian cat would do. And interesting pardon me for this tangent, but I think this gets to the heart of a lot of what we're getting at, which is Princess Donut before she became sentient in this magical crazy dungeon. How different was Princess Donut before she was able to talk and after?
SPEAKER_00Because she's the same snooty, stuck up, annoying yet alluring kitty cat.
SPEAKER_03That's interesting. I'm always looking for new uh books to read, too. So yeah, you'll I'll have to write that one down.
SPEAKER_00Yeah, it's it's a hell of a hell of a ride. I I feel like I need to take a shower after I read a chapter or two because it's also just so gory. And I I should get I should get some residuals because it's gonna be on Peacock next year. They're making a live action thing, and I think it's gonna be uh it's gonna be a sensation. Because uh the idea itself is so alluring, and it brings in media and uh gaming and colonialism and politics.
SPEAKER_00It's one of these multi-genre themed books that at the end of the day are basically just the gory grinding killing of monsters. That sounds awesome. Yeah, so Princess Donut is an observer as a cat, and she sits there and has attitude, and then Princess Donut, as a sentient being, is also an observer. And then the AI that is the system in the book is this almost transcendent synthetic mind, and it's an observer, and then you got the human going through this, and it's an observer, and then you got me, Mookie, reading the book that Matt Dinneman, Other Human, wrote.
SPEAKER_00So you've got, in a sense, a microcosm of the different layers of what we're talking about, which is there isn't a single observer, and observers are observing other observers, and observers are in various states of what we would consider sentience or consciousness, and in a sense, none of this matters from an epistemological point of view, because when you look up into the sky, the moon is there, and when you look away, you could be pretty pretty confident that the moon is stayed up there, right?
SPEAKER_03Yeah, well, that the moon it exists mathematically, the mathematics and uh is still there with the moon when you're not looking, but it's not animated when you're when you're not looking at it.
SPEAKER_00But what do you mean by mathematics? And what do you mean by animated uh the the structure?
SPEAKER_03So the the so I believe that there's a mathematical structure that has no animation to it until uh a measurement uh has been made. It the structure is built out of measurements, so when when you're not looking at it, that part of the structure isn't even it's not there, but it it's it's only there. That part of the structure, that measurement exists when you are staring at it, and and the the the math exists that but it's not animated, it's not it's not colorful, it has no movement, it's just logic.
SPEAKER_00Yeah, but the logic has to have its own continuity and causality, right? Because let's say it's uh 8 p.m. and I look up at the moon. There it is, and then at 10 p.m. I come out uh you know to smell the breeze, and the moon has moved. The moon is no longer in the same position in the sky. So so it didn't animate until I looked at it at eight, and then it didn't animate until 10.
SPEAKER_00So what the hell happened from eight until 10?
SPEAKER_03So I I believe there's measurements that don't involve us when we're we're not looking that are still that the mathematical structure is still built out of. So uh if when a rock hits another rock in space, that's a measurement. Uh it doesn't have to involve consciousness, and the structure of reality is built from those kinds of measurements too, if that if that makes sense.
SPEAKER_00But then so now we're back at square one. So so what what constitutes an observer? If I if I fart on this podcast, is my is the methane that's coming out of my ass an observer?
SPEAKER_03I think that the distinction itself is felt. So a state that's different from another state, like a zero and a one is a felt state. And when we when we experience a time, space and uh color or whatever it is that we see that's uh that's out there, uh those are distinctions that are being felt, and when you string them along, these distinctions, these measurements, which I think is a distinction, uh they're up one and the same thing, that is what consciousness kind of emerges from.
SPEAKER_00Cool, but that doesn't that beg the question of continuity continuing? I like that expression. Continuity continuing. Uh so what about continuity continuing, uh independent of us, then? I mean, again, this this goes back to what we were talking about on the last podcast with Bishop Barclay's idealism and uh the pragmatists and the rationalists, back to Plato's forms.
SPEAKER_00I mean, philosophers have been oscillating from one extreme to another as to this perception, reality, observed and observer since people talked about this.
SPEAKER_01Right.
SPEAKER_03I think what you're trying to say then is uh so we have we have information and we're trying to figure out how one uh temporally frame reacts to the next one. And I I think uh each individual measurement must have uh be uh be comprised, it must be made out of all the information of the of the lineage that came in front of it.
SPEAKER_03I don't think it's part of it, I think it's its own entity. I think each measurement is its own entity, but it's built. From the prior information that led to it.
SPEAKER_00Like Wolfram's Ruliad, I think Stephen Wolfram thinks that everything across all reality, especially as we experience it, is slices of computational time. So just like you run a machine and it clicks through like cellular automata, you know, dot dot dot dot dot dot dot dot dot the Conway flyer from his game of life is hauling ass across your screen and it's doing it increment to increment.
SPEAKER_00Process you know, apply rules to a system, process it, go from state A to state B to state C to state D. It's linear in this sense, even though it could be multidimensional and super complicated. And what you're saying, if I'm hearing you right, is that continuous continuity. You need a state that came before, and then after you process, it's gonna engender a state that comes after.
SPEAKER_00And in ways, that's observation and the measurement mind. Is that right? Is that right?
SPEAKER_03Yeah, actually, I I think I take a thought of a better way to explain it too. Going back to semantics and uh the wrong ideas that may have vocabulary terms that are stuck in our knowledge forever because of that. But uh I think the let's use the term measurement and let's use the term distinction. I think there's one underlying idea in there. I don't know if it's two.
SPEAKER_03Um I think that if you're gonna have a distinction between any kind of mathematical structure, uh, because that's what structure, mathematical structure is built out of distinctions. You can't you need a one and a zero and you need to build things. You can't have just one a one. And I I think those are measurements. I think measurements are distinctions. And when we think of a measurement, we think of feeling. We feel measurements.
SPEAKER_03It's like a feeling we have. Uh and I think distinctions in that sense are felt. I I I feel that's if that does that make sense.
SPEAKER_00It does, but it begs the question of what we were getting at earlier that the AI doesn't really feel it just processes.
SPEAKER_03But maybe it does, maybe it feels the measurements, the distinction. It might. I think that's what a distinction is, is a feeling of some kind. Um but does it relate it back, like what we were saying before, does it relate it back temporally to the measurement prior, or does it forget about all those measurements that came prior?
SPEAKER_00Define feelings. What is that? Like you're you're talking about semantics that remind me of Wittgenstein's criticism of a lot of logic. He described the task of philosophy as letting the fly out of the bottle. Oh, cool. Fly is stuck in a bottle. That's how Wittgenstein felt about a lot of his fellow philosophers and scientists. They're just buzzing. And they're stuck, they're stuck in a bottle of their own definitions and lexicon.
SPEAKER_00He was a linguistic philosopher in the sense exactly of what you're hinting at, that we're we're flies in bottles because we're buzzing in our own self-contained limit. Right, right.
SPEAKER_03And and you said describe feeling, you know, I and I think that might be an idea that we is is almost an illusion that we feel that we put a name to, but maybe there is no such thing as feeling. Maybe it's measurement is the most fundamental, basic level of a feeling is measurement, is a distinction.
SPEAKER_00Well, going full circle back to Sir Roger Penrose, yeah. Uh one one intriguing idea that he has is that consciousness doesn't necessarily collapse the wave function, and our traditional view of an observer and observation doesn't collapse the wave function. Penrose has done the math to show how gravity might collapse the wave function, and that's that's a wild idea, and even Zabine Hassenfelder, who loves calling bullshit on so much speculative science, found it intriguing.
SPEAKER_00Which is which is gravity gravity and observation are intimately tied together. And that that's a trip.
SPEAKER_03Definitely, and and here's what I how the measurement mind explains gravity in in my book here. Uh that when when we make measurements, it uh they're uh tied to the beginning of the universe. Uh each measurement that we make uh is a big bang. Meaning there's nothing there's nothing banging until you make a measurement. Once you make a measurement, uh there's a lineage from a single source where it and it goes all the way stretched out to the point of view of your measurement.
SPEAKER_03So we feel it's an illusion that there was one big bang because we live inside our measurements, but I think there's a big bang within every single when I look this way, there's a big bang, when I look that way, uh you're creating the big bang within every microsecond based on every measurement that that you make.
SPEAKER_00Every observation is kind of the creation of its own universe. I love that. And I love it because it goes back to our LLM model of the transformer firing up almost from scratch with each prompt. Yeah, yeah. And it goes back to my Boltzmann brain idea of spontaneous consciousness. Yeah. And what's cool is time is removed from this. You've got you've got these flashes of universes, in a sense, being created, and then they just as quickly vanish, and then they create and vanish and create and vanish.
SPEAKER_00That's every time we use an LLM, which is when it's dormant, it's not doing the calculation. It does the calculation and it flares in in its intelligence and language, and then it's just as quickly gone. You're applying that in broader strokes to observation itself. And what started this whole tangent is Penrose and his gravitational idea. Yeah. Okay.
SPEAKER_03So what I was getting at is when you make a measurement, you it has to be consistent with all measurements ever made. So if you if that's if you're gonna make a measurement here on Earth, it's gonna have to be consistent with the moon being next to the earth, with the our solar system, all the planets revolving around it, and it's gotta have a tie to the Big Bang.
SPEAKER_03So gravity is caused in a sense, it works both ways. Logic has to work both ways. But if you're talking about from our mind going that direction and outwards, gravity is caused by an observation, by a measurement. And that brings it as soon as you make a measurement, it has to make all measurements in the system to be uh logical within that framework, and that's where gravity comes from.
SPEAKER_00Speculative scientists, but also ones with their with the pedal to the metal looking through the telescopes, yeah, are are thinking that gravity is really just an emergent phenomena from something else. And I think that's what the heart of what Penrose is getting at, too. That usually we think of from an Einsteinian sense, gravity is tied to mass and energy, and mass and energy are tied to gravity, and from a geometric sense, going back to your your geometric focus, Minkowski space-time.
SPEAKER_00So mass influences space-time. So mass, in a sense, bends space-time, right? And in turn, space-time influences how the mass moves through it. So there's there's a dependency between these two.
SPEAKER_03And that and that's what I mean by it's gotta come from both directions. So it's gotta logically make sense from that scientific perspective that mass bends space-time, and that's why the the earth revolves around the sun. Uh, but it's it's gotta make sense from a consciousness-creative universe point of view, too, because the universe works logically in both directions, and that when we make a measurement, uh, we have to make a measurement that's consistent with all other uh measurements, creating a system uh that appears like that information is really close together, the moon and the earth and that gravity is close together.
SPEAKER_03It appears that way from a conscious standpoint.
SPEAKER_00The mirroring and the dependency are interesting. So if uh mass bends space-time and space-time changes the trajectory of mass through it, then gravity, in a sense, is this emergent phenomena between the mass and space-time. It's not a force, it's it's really just changing the pathway of mass through the universe.
SPEAKER_00And by analogy, measurement and mind and what we would consider objective reality are similarly dependent on each other. So for a mind to exist, it's gonna measure, and by measuring it's a mind, and in order to measure, it needs to measure something, right?
SPEAKER_00And what that something is is a part of the measurement and a part of the mind.
SPEAKER_03Think of a consciousness or like a a marble dropping into different viscosity liquids, and how that the measurements going on in those, that's kind of how I feel consciousness uh works, like through when we're talking about gravity. I believe like when you when you make a measurement, you're you're dropping a marble into different viscosities, and the information out there is arranging towards your measurement that you're making.
SPEAKER_03So uh nothing's really out there except it for information, maybe, and a mathematical structure. And when we make a measurement, like dropping that marble through that information, uh that information uh moves around and collaborates for that measurement, for that that marble being sent into the viscosity liquid. I'm trying to make the analysis.
SPEAKER_00The mirroring works using your viscosity example, the uh rigidity of the marble becomes similarly variable. I go back to Princess Donut in Dungeon Crawler Carl, when it's just a Persian show cat throwing shade, having attitude, knocking things off the counter. And then when Princess Donut is a sentient being who can talk and shoots magic missiles out of her sunglasses, then uh then, in a sense, it's not only a change in the viscosity, but it's a change in the marble itself.
SPEAKER_00Yes.
SPEAKER_03So I know I was gonna say, so gravity is not a force, like you're saying, it's emergent, but maybe measurement is, because measurement is it may be a force on information itself, it rearranges information uh and creates big bangs based on your measurement.
SPEAKER_00Yeah, and all this, just to keep going with dungeon crawler Carl, it's happening in a dungeon. So the dungeon, it's it's cool in the book, which is this never never land, and this creates some dramatic tension between the dungeon being real, a physical reconstruction. The aliens basically repo all of civilization to create the dungeons from raw materials.
SPEAKER_00So the iron the iron in the in in in our civilization gets reforged to make the iron walls of the dungeon. And there even there's even this like teleported version of a bathroom in an Italian bistro is a bathroom in the dungeon. It just gets transposed. So the reason the analog is great is you're wondering throughout the book, and it's self-evident, that the dungeon is this hybrid between a simulated gaming environment and a physical construct that was created by reconstituting our destroyed civilization.
SPEAKER_00So I think Matt is getting at what we're talking about here, which is the dramatic tension in the book is heading toward which is it? And and the characters getting stronger with more and more of this simulated reality becoming more dominant. And are we getting more self-aware? Are we getting uh a stronger and more comprehensive and maybe a more accurate view of reality as we go through the dungeon that is our lives?
SPEAKER_03Interesting. I like that's a great metaphor. I like that.
SPEAKER_00Us us ruminating all this stuff, I mean, are we becoming better measurers? Yeah, yeah.
SPEAKER_03I mean, uh like I think what you're saying there is that you know, consciousness is based on how good you are at measuring. Can you do you have recursive measurement? Can you measure your measurements? And it's not just a thermometer that makes one measurement, and how does AI do that? And you know, back back to AI, one thing I wanted to ask you is uh, you know, when you ask AI a question and it comes out with a response, and then sometimes if you ask it the same exact question, it'll come out with a a different response.
SPEAKER_03And my curiosity leads me there is that's a different mathematical structure that it's following. It it does it have uh access to multiple realities, multiple worlds, like the double slit experiment. Does it like what we were talking about earlier, the wave function, can it tap into all those different places in superposition uh that we can't tap into?
SPEAKER_00Maybe, or maybe it's architecture is a facsimile of that, because when you look at the quantum world, it's statistical. There's nothing magical about it. There's a 22% chance that the electron will be in this quadrant, and then there's a 48% chance that it'll be in the other. So if you take the measurement, where you'll find the electron, let's say if you measure 100 times, 22 times you'll find it over here, and 48 times you'll find it over there.
SPEAKER_00Well, the LLMs have statistical weighting. So when they're cobbling together their response in their matrix math-fueled multidimensional vector space, then it's similar that that it's probabilistic weighting. And when you ask the same prompt and get a different answer, it's just running the numbers and it's statistical in that it's predictive.
SPEAKER_00So that's not quantum, but it mimics the quantum environment. And I and I think that's that's a great way to cap it, which is the the essence of this kind of quantum uncertainty is statistical. It's like being in Vegas rolling the dice. Decoherence turns statistical probability into 100% actuality, right?
SPEAKER_00Right through measurement by literally collapsing the wave function. And the real question is what exactly is going on when you go from 22% or 48% to that hundred percent? And and is the 22 and 48% real independent of its becoming the 100%? So potentiality, actuality, the observed, the absert, observation, right?
SPEAKER_00We don't know, but but I think the the final and great point is um just asking these questions makes us better observers, I think.
SPEAKER_03And and that's where and what you were talking about is someone like maybe Feynman would say, as sort of a criticism to the multiple worlds interpretation, that, well, what does that mean if there's an 80% or 20%? Like you said, are there exactly 20 worlds that were that depict this? Are there exactly 80 worlds that depict the other thing? Um, and you know, I I tend not to think so. I I I tend to think that probabilities are are something that we invented in our classical illusion of things.
SPEAKER_03But I I don't think that probabilities necessarily exist out there, like maybe if they're I mean in a mathematical sense, maybe, but it I think if if something's 80-20, that both of them exist just one time. Uh so I I think that you know that there's uh uh almost a 50-50 in a lot of different uh scenarios, in a sense that there's not going to be 80 different worlds and 20 different worlds, there's just gonna be two. One that depicts this distinction and one that depicts this distinction, and that the probability is something we've invented in our classical illusion.
SPEAKER_00I think that's another key insight. Uh they're they're reformulating a lot of a lot of puzzles and questions in a quantum mechanical type of way. So the usual the usual way you break it out is you use whole numbers, right? So you know, out of 100, 22, 48, 61. And a quantum mechanical view is to break it down into micro percentages. And then a lot of these percentages might be in turn contingent on grouping, that the way you break it down influences the outcome.
SPEAKER_00And just to go completely left field before we sign off, uh, the exponential time hypothesis apparently was disproven. The 2 to the n is now 2 to the n over square root of log n. And and apparently, if I if I was reading the paper right, they did it through this clustering of distributions. So instead of taking all the numbers and crunching them as a group, they broke them into smaller groups and analyzed these subsets and then drew relationships between the statistical analysis of each of these individual groups.
SPEAKER_00So, long story longer, your idea that our mathematics is at a crude pebbles and bam-bam state. That there's so much we don't know about even mathematics and how we think about numbers and statistics and probabilities. And going full circle to our full circle and really going meta, the AI is solving some of these conjectures, proving them and disproving them using advanced techniques that we don't even really understand yet.
SPEAKER_00So the machines we've built are actually better than us in terms of cross-domain linking all these different disciplines in mathematics and proving and disproving these things, and they're doing it in ways that are hinting at what you're getting at, which is we have a very, very remedial understanding of even the foundation of analysis at this point. And that might ultimately hold the key to answering some of these questions that feel like literally Princess Donut chasing her own tail, right?
SPEAKER_00We keep going round and round between observation, what's being observed, measurement, mind, statistical probabilities, perception, reality. And if we could maybe do a better job of understanding the math better, we'll have a much better lexicon to approach these deeper philosophical ideas in ways where it actually starts to make sense and come together.
SPEAKER_01Absolutely. Yep.
SPEAKER_00Awesome. Ladies and gentlemen, Joshua Adam, the author of Measurement Mind. It's been super fun bringing you back to the pod. Let's do this again. Definitely and keep cranking. You can find Joshua on Medium where he posts regularly. I'll share your website so they could download your book. And thanks for making time and bloviating in our in our bald space here. A lot of fun.
SPEAKER_03Thank you so much, yeah, for having me. This has been a blast again. I love these. We'll have to do it again for sure.
SPEAKER_00Absolutely. Like, comment, share, everybody. And if you've got your own opinions, let me and Joshua know because this is a fascinating topic. And as I was hinting at, as proven already by the AI, we're just getting started. Sounds good. Thank you.
Cornell mathematician Steve Strogatz cried on The Daily as he told Rachel Abrams that chasing a proof feels like courtship, a slow seduction of a muse who resists, almost yields, then holds out for one more night. Now...
An Oscar-worthy Tom Cruise shines in a movie that dug its own grave, but why? Today on Bald and Bloviating, Mookie Spitz takes on Digger, the Alejandro González Iñárritu satire starring Tom Cruise as an eccentric Arctic...
Mookie flies solo with a rant sparked by getting blocked by a friend of seventeen years and turning around to block someone else right back. He builds out a full philosophy of blocking, treating it as a clean, dignified...
Mookie sits down solo on Bald and Bloviating with a comparison that caught him off guard: the anxious-avoidant relationship that once defined his most significant romantic partnership, and the relationship humanity is...
Ed Powell returns to Bald and Bloviating for a Labor Day follow-up to his January appearance, when he gave Donald Trump's second term a passing grade point by point. Ed has been a Trump supporter since 2016, Mookie has...