Metaphysics and modern AI: What is reality?

In the first episode of our series on metaphysics, Michael Herman joins us from Episode #14 on “What is consciousness?” to discuss reality. More specifically, the question of objects in reality.  The team explores Plato’s forms, Aristotle’s realism, emergence, and embodiment to determine whether AI models can approximate from what humans uniquely experience.

• Defining objects via properties, perception, and persistence
• Banana and circle examples for identity and ideals
• Plato versus Aristotle on forms and realism
• Ship of Theseus and continuity through change
• Samples, complexes, and emergence in systems
• Embodiment, consciousness, and why LLMs lack lived unity
• Existentialist focus on subjective reality and meaning
• Why metaphysics matters for AI governance and safety
  
  

Join us for the next part of the metaphysics series to explore space and time. Subscribe now.

What we're reading:

• [Mumford's] Metaphysics: A Very Short Introduction (Andrew)

Do you have a question or a discussion topic for the AI Fundamentalists? Connect with them to comment on your favorite topics:

• LinkedIn - Episode summaries, shares of cited articles, and more.
• YouTube - Was it something that we said? Good. Share your favorite quotes.
• Visit our page - see past episodes and submit your feedback! It continues to inspire future episodes.

Chapters

• 0:00 Welcome and metaphysics mini-series
• 5:08 Defining metaphysics and today’s focus
• 6:07 What makes an object an object?
• 9:20 Perception, properties, and persistence
• 12:20 Plato’s Forms versus Aristotle’s Realism
• 17:15 Identity through change and the Ship Paradox
• 21:07 Simples, complexes, and emergence
• 24:15 Consciousness, embodiment, and LLM limits
• 28:00 Existentialism and subjective seality
• 31:15 Why metaphysics still matters for AI

Transcript

SPEAKER_00: 0:02

The AI Fundamentalists, a podcast about the fundamentals of safe and resilient modeling systems behind the AI that impacts our lives and our businesses. Here are your hosts, Andrew Clark and Sid Mungalik. Welcome to the AI Fundamentalists and the first full episode of our mini-series on metaphysics. You can hear more about the purpose of this series in our previous episode, particularly why principles in metaphysics matter as a foundation for modern AI. In this first part, we'll be talking about reality.

SPEAKER_03: 0:41

And today we're glad to have on Michael Herman. You might remember him from his previous episode we did. It was episode 14. On this podcast, unsurprisingly, we talk a lot about the fundamentals. And for a podcast called The AI Fundamentalists, we owe you a little bit of a fundamentalist view then on what is reality. And this is a part of our mini-series on metaphysics. And so, you know, we really want to build up from scratch. A lot of CDs are working on. Before we hop into the meat of the episode, uh, I'm really interested to hear, you know, uh, are there any books you guys are reading recently that you're that you're interested in that you like?

SPEAKER_02: 1:22

Uh I'm currently reading, let me see, The Way of the Shaman by Michael Harner. And it's kind of I guess it's like a Western sort of like step-by-step approach to um shamanic journeying. And uh I know shamanic journeying sounds a little hippy-dippy, and I guess it kind of is, but it's it's really just like a similar to meditation, it's a way of um you know kind of changing your consciousness for spiritual and like healing purposes.

SPEAKER_01: 1:50

Very nice, very cool. Uh that's uh strangely enough, I just read The Golden Road, how ancient India Transformed the World. Uh like the Indosphere is what it was calling it. It's a pretty new book by William Darrympol of kind of like the history of India and how it really influenced a lot of areas and was one of the earliest traders of the Roman uh Empire and all that kind of stuff like that. It doesn't really get that much justice oftentimes in like how influential India was. Um, definitely coming back on the world stage too. So I thought I found that very interesting. Uh a little bit of a dry read. It wasn't the most like easy to read book, but I thought it was very interesting and just kind of ties somewhat in the endosphere of uh from your book as well. But that was very good. Love the history part of it, uh, good author, but yeah, probably not the easiest reading. But then definitely been reading a lot of uh metaphysics for prepping for this series, and it's been honestly a lot of fun. Um I know when we were all in person a few weeks ago and we were talking about some of this, and like uh Sid and I found a book, Stephen Mumford's Metaphysics, a very short introduction. That's been a great read. I've been trying to read some other things. I I had read Aristotle Metaphysics a long time ago, but haven't gotten that one done again yet. Um, but also looking at some of the Descartes' stuff of meditations and and uh discourse and methods and things like that. So that's been a really interesting journey to really take a step back, as Sid mentioned the fundamentalists, really look at the fundamentals of like metaphysics and what is thinking, what is a person, and and as we try and like tackle with this, you know, in the age of AI, what do these concepts mean? And really going back to not just what a computer science textbook says, but really going back to I mean, what did Aristotle, Plato going back that far, type uh what is metaphysics and like the whole field? So it's it's been very interesting, really excited of this. Is we did the really kind of an intro episode of the kind of the concepts we'd be touching, and then really honored to have you, Michael, on uh our first episode here of um really digging into and Sid will outline kind of what we're gonna be going through, but very excited for to get this started. But before, Sid, what have you been reading lately outside of Mumpered?

SPEAKER_03: 3:47

Yeah, yeah. I mean, it's it's been really great getting back into doing some physics reading. It's been a while since I sat down and said, like, I'm gonna read one book in one week, and that's been it's been great, uh, a lot of fun. Uh I'm finally getting back into reading a little bit of sci-fi. Uh, I think Michael already knows this, but I'm now I'm reading the uh the three-body problem, uh, which is one of the like new premier uh sci-fi books, which is coming out of out of China, which we haven't seen a lot of before here in the States. Uh, but it's super interesting, very physics heavy. Uh, really interesting to see a sci-fi book that's clearly written by a grad student. They have a good understanding of academia and and how research happens in the real world. And that's very refreshing.

SPEAKER_01: 4:30

Awesome. I did I need to read that one. I did the the Netflix series and I definitely saw the academic bent. So I don't know how accurate the net like it's probably like you know, destroys the book as usual. I don't know. Maybe it's a better faithful one. It definitely had a major focus on like academics and research and things. So I thought that was cool and kind of unusual. So at least that little bit of it came through in the in the Netflix series.

SPEAKER_03: 4:51

Yeah, absolutely. And maybe I'll maybe I'll watch the show after I finish reading.

SPEAKER_01: 4:55

Um, David, who's gonna come on our podcast next week, he is a huge fan of the Netflix three-body problem. So yeah, highly recommend from both of us.

SPEAKER_03: 5:05

Awesome. All right, so let's get into the meat of the subject today. So, as a quick recap, uh, since this is technically the first episode that we did a preview on the last episode, metaphysics is the study of the most general features of reality that includes things like existence, objects and their properties, change of causality, space and time, and the relationships between matter and mind. Uh and this is usually considered one of the oldest branches of philosophy besides maybe moral philosophy. Today, specifically, we're very interested in the nature of existence, the nature of objects, and the nature of properties, and how those things interplay, and if one thing can exist without the other. So, to start off the discussion today, we're going to talk about specifically what is the nature of the thing? What is the nature of a chair? Uh, what is the nature of a table? How would you define what an object is? And so, I mean, you know, with some of the reading you've done, Andrew, and some of the insights you have, Michael, like how do we define an object? What makes an object present?

SPEAKER_01: 6:14

Yeah, this was this was wild. This what metaphysics is and I I definitely don't feel qualified to answer that question, but it was very very much from this um this book of the different ways. This was the Mumford book does a great job of describing the different components of what that is and different ways to think about it. But really, that an object can kind of be into properties of it, like uh hardness and and shape and things like that, and the particulars of like the specific color, or is it is it your chair, or is it my chair, and things like that. So uh, and then there, you know, substratum and things. There's a bunch of these different theories, but really I think high level for me, what really resonated was uh the really properties and particulars, and an object has that has those different components, and maybe kind of like a bipart graph essentially of the interactions of those components. So you can see like the the type of what an object is, like the it's a chair, it has certain attributes, and then the specificness of that of that chair to you or me as an example.

SPEAKER_02: 7:12

So that makes sort of like the property bearing um the defining feature of objecthood. And I think the problem with that, or like one problem that I see with that is like how how does an object then persist through change? So like if you change, you know, a single property, then wouldn't it that change the object from like one thing to another? And so like I think a good example of that would be a banana. So I think we can probably you know definitively say that like a banana just because it goes from green to yellow doesn't mean that's a new banana. It's going to be like the same, same sort of banana. So I think like my definition or like the definition that I like of like what an object is is anything um that can be perceived. And so that's obviously much more subjective and you know it's phenomenological, but I think for me, it just I think it makes a little bit more sense than like just like focusing just on the properties. It's it's definitely more um subjective.

SPEAKER_01: 8:14

No, I love that. I fully agree. Um, one analogy I want to run by you and see what your thoughts are from the Mumford was, and I think it kind of aligns what you're saying, but maybe maybe not, but that's what I get your thoughts on, is a thing is separate from its properties, so you can think of it kind of like a uh pins in a pin cushion. So kind of like what you're perceiving is the cushion, and then you have the different attributes as pins, right? So you can the the cushion stays the same, and with what you don't see in like the if it changes shape as a banana or whatever, it's the cushion. It's still there. The banana is a banana because you perceive it. However, like it was it's smushed now or it's changed color or whatever, and those are like the pins are being added or removed from the cushion. Uh, and that was one of the analogies in the book. Uh, is that kind of a line?

SPEAKER_02: 8:57

Yeah, I think so. I mean that's that's a little bit like I would say object-oriented programming where like you're creating like an instance and you're like you know, you're passing it like individual attributes.

SPEAKER_03: 9:09

Yeah, absolutely. And and I think that like, you know, in in the text we get this this situational problem where if we have the banana and we describe the banana as a set of attributes, right? It's this color, it's this shape, it's uh it's this color, and it's on this table. We find that we can have a pretty good sense that you can create continuities, right? One banana is the same banana if you know one or two attributes are changing, it's the same everything else except for one attribute, and we're moving through it in time. Uh, we then do run into the problem of, well, then how do we distinguish two bananas that are next to each other? Right? Because they have everything in common. The only attribute that's changing then is is location. So, you know, in what sense can we then describe objects as particulars or as being distinct from each other if they are uh, you know, if we take for granted this idea that there's a property, a set of properties that define a single object?

SPEAKER_02: 10:10

Yeah, so I think like in that sort of definition of an object, what would make an object unique would be you know its specific location, as you had said. Um the sensory features of it, I would say, and then also like I guess like its relation to sort of like the subjective or to like the person observing it.

SPEAKER_03: 10:39

Yeah, I th I think that's fair and that's that's very interesting. I guess do you want to dig in a little bit to this idea that the existence of the object is only relative to the perceiver, right? Does the object not exist without perceiving? Right? If I if I close my eyes to the object cease to exist.

SPEAKER_02: 10:57

Yeah, I would say so. Like going back to that, you know, age-old philosophical question of like, you know, if a tree falls in a forest and no one is there to perceive it, you know, did the tree fall or you know, whatever it is. Uh I would say no. I would say like if there's nothing there that to you know perceive or take in those sound waves, then yeah, it's for all intents and purposes didn't happen. So I do think it's it's really dependent on there being a subject there or an observer.

SPEAKER_01: 11:27

Well, this is gonna be interesting as we get into uh really the two from the Mumford book, how the two ideas really like this the Plato versus uh Aristotle kind of how they view the world. And I think there you sounds like you're maybe getting a little bit more on the Aristotle side of the house of if you if someone is there to perceive it. So um so do you think we want to start heading into that the Plato versus Aristotle debate of Platonic realm?

SPEAKER_03: 11:56

Yeah, and so I guess like, you know, before we hop into that, I guess we'll talk a little bit more broadly about types of objects or archetypal objects or prototypical objects, right? So for example, I say a banana, and a banana is conjured in your mind. Maybe it's a perfect ideal banana, right? This is oh, this is the banana that like all bananas are like this banana. Uh from that viewpoint, we have this split between Plato and Aristotle. Let's start with Plato just because chronologically it's correct, and Aristotle is a response. Is there some perfect banana out in some perfect heavenly realm? And all bananas exist on the earth are basically representations of that banana. Do we understand banananess or tableness as this abstract construct which is pure and real? Or is there possibly another solution?

SPEAKER_02: 12:53

Yeah, my initial answer to that is no. Um perfect forms do not exist. Um I'm still trying to wrap my head around the question though.

SPEAKER_03: 13:04

Yeah, yeah. I I guess the question is like, and I'll I'll do something simpler than a banana. It's a circle, right? So, like, a perfect circle is defined by a bunch of mathematical functions. Uh, but there aren't real perfect circles in the real world. Most circles have imperfections. Uh so it sounds like to Plato that there is a place where the real circle exists in some heavenly realm, and that all fake circles on Earth are manifestations of a perfect representation. Uh, how do how do we feel about that?

SPEAKER_01: 13:40

Definitely from a modeling perspective, I think that has a decent analogy, right? So, like we're that's an interesting way, and I'll think about that some for models. So, all models are imperfect representations of reality, so like what you learn with how models work. So, obviously, it's a lot more nitty-gritty and not as aspirational as like how Plato is saying, but we're we're kind of operating a little bit in the Platonic realm of how we view when we're doing modeling. But definitely don't want to go too much down that tangent and keep it, keep it more the metaphysical realm here of that difference. And I I think to me, the Plato Platonic way really resonated with how the view of the world there, because even we start thinking about counterfactuals and scenarios, and maybe I'm getting too tactical too fast, but like the idea of there's a perfect representation somewhere else, and then everything else is an like an approximation, is a good uh really resonated of which again, all of this is there's no empirical evidence for any of this, right? So it's like that it is kind of like how it resonates with with somebody. Uh, but the Platonic way of like the perfect circle of no matter how good of machining you do or something, there's gonna be some imperfection somewhere. But we've kind of created in mathematics and things like this some of these like perf like the magic ratios and like what like pi and things like that, we actually kind of have these abstractions that we know is an ideal that we can't quite ever calculate what that is. So I don't know, it just kind of like felt like a good through line for a lot of how science and things operate from the Platonic view.

SPEAKER_02: 15:03

Yeah, I think an object is an object because we can perceive it as something that is distinct, not necessarily because it mirrors some sort of perfect or heavenly form. So, you know, again, I I think all objects are just grounded in experience. And so you don't have that sort of that heavily realm, like doesn't really even like really enter into sort of that framework, if that makes sense.

SPEAKER_03: 15:32

That absolutely makes sense, and and honestly, you sound like the exact Aristotelian position, right? Which is basically like there is no perfect version of these objects. Yes, these are abstractions that we can use to describe how things could be, but we exist in the real world, and thus objects are what they are here, and there is circle-likeness or circliness, but there is no need for there to then be perfect circles on Earth for us to understand what a circle is. So grounding a lot more in physicality and reality, and then building from there. Um, so I you know, I I think that you know, it's very easy for us to now say, like, oh poo-poo, Plato, you know, you're like 300 years behind the curve. But you know, he you know, it it's on to something to kind of describe this idea of like, are there perfect representations in the world? I think we deal with a lot of this in science and even in AI, where we say, like, you know, no model is correct, but some models are better than others, right? The true representation is out there, but we cannot have it. All you can have is an approximation, and the quality of that approximation is that gives it value.

SPEAKER_01: 16:48

Yeah, and I think it's very interesting. Instead, I want to your opinion on here is Michael and I are both split on the two different views, and this is a common thing in the history of philosophy, is it really comes down to these two separate views. I forget what exactly it is, but it's this famous painting of Plato and Aristotle arguing about this exact thing. So this is very like two different philosophic philosophical views of the world. And uh sort of very curious where you land on it. Mike and I kind of see it a little bit differently, which is probably I would actually like to pull philosophers as it kind of like a down the middle, but there's definitely it's been kind of a running thing of people view it the different ways. Sid, where do you fall on this?

SPEAKER_03: 17:23

Yeah, this that's a really good question. And and I think that basically what I find is that trying to assign the world as a set of ideal representations becomes very difficult. I'm sure we've seen all some very abstract and interpretive styles of chairs where it's like, wow, this is nothing like the perfect chair. And so the question is then like, are these types of idealized prototypical models actually useful for us? Maybe not. And I think to Michael's point, I stand on the side of maybe privileging more the human experience of objects and saying, if I can use this as a chair, it's a chair. Even if it doesn't meet, you know, prototypical definitions, if it doesn't meet this idea of like a perfect Platonic object. Uh so I I would probably land on the more Aristotelian side, but I understand that it's important for science to have this notion of the real world sometimes being representable as abstractions of some perfect ideal. Um but maybe this is more interesting for mathematicians and scientists than how we need to live our day-to-day lives.

SPEAKER_02: 18:35

Yeah, I think to a certain degree, like a lot of like what we call objects are just sort of like it's because we agreed upon it. Like if you think of a chair, you know, a chair has you know legs, it has a back, it has a seat, and when you when you combine those all together, we label the chair. And it's it's because it's because of that combination, because you know, we we've we've decided that thing is a chair, right? Like you can't there isn't like an infinite number of combinations out there with this chair where you can add like all these other things and it's still you know being is a chair, but it's only because we've agreed upon that, right?

SPEAKER_03: 19:15

Yeah, I mean absolutely. I mean you yeah, then you have like the spork problem, right? It's like, well, it functionally has the properties of both. Does it then exist as a new object, or is it distinctly one of them?

SPEAKER_02: 19:26

Yeah, and I think there's a uh I I can't remember the exact sort of uh metaphor analogy or whatever, but there's like this philosophical question of like there's like a wooden boat, and if you remove a plank, you know, from that boat and replace it with a new plank, is that boat still the same boat? And you know, it goes as far as like you know, saying like, hey, if you remove a plank, you know, once a week and you know, all of a sudden you have like a completely sort of like you know, all the planks have been replaced, is it still the same boat? And then, you know, like if you've saved all those old planks and you know built a new boat with those old planks, is that now a better representation of like the the old, you know, the original boat than you know the how it currently is.

SPEAKER_03: 20:13

Yeah, yeah. This is that's the this is the classic Jason's ship, uh, which which is basically you know, you board by board replace it, a broken ship, and you can you recreate it and it's you know which one is Jason's ship? Is it the first one or is it the new one? Uh and I and I think like it's hard to say that there's an answer. But I would say that if we take something like the pincushion model, I think we end up in a very comfortable spot where we can say that there was continuity with the ship that we were slowly building over time. And uh that the ship that we have been replacing boards over time can be Jason ship, the same way that uh myself, Sid, is still Sid, despite all the cells in my body being replaced every 15 years. I would still say that there is some continuity between me of the past and me now. And I guess on that point, I want to bring us to our next area, which is building off of this. So we've we talked a lot about combinations and pieces making up a whole. So let's talk about instead of simple particulars, more complex particulars. Particulars where we have a bunch of component pieces, or some I call them simple objects, simples, combined together to build brand new objects or a complex. Let's imagine, for example, a computer, which can be understood as a combination of silicon and iron traces and copper, but now it has new properties. Do we then understand that collections of small simple particulars make new objects, or are new objects just complex mechanisms combining the parts of them? Do objects become more than the sum of their parts when arranged in particular ways?

SPEAKER_01: 22:00

Yeah, I think this is really the crux of it. And uh I think I kind of live in between both camps, which I'm not sure is a state like a metaphoric or metaphysically valid uh place to be, but I think it's in both, right? So like it's really that reductionism of all things can be explained, like all the little parts of a computer and they can be combined and you know how to combine the things, versus the you know, emergence of like sometimes two plus two equals eight. You know, it's like non-linearity, right? And I I did a lot of like dynamical systems and chaos theory, and I come kind of from that background of where emergence and complex systems have these properties that emerge. Uh and I I'm personally on the boat of like consciousness of a human or like the what what what's the what makes a human or what makes that different, then it's not just like a collection of cells, there's something like different there, right? Um, and that would be more that emergence versus you know when we start really thinking about AI and things, but like I do think that computers are some of the parts, and that's the reductionism approach, right? So like and I think I don't know if that's metaphysically valid that I kind of live in both camps of I think human beings or like not even human beings, animals. I think any sort of like living thing is different and it's has emergent properties that you can't just like cobble together in a lab, Frankenstein style, and and and make it, right? Versus like computers, you can actually do that and we can go down the bytes and bits. So I'm very interested in both your thoughts on that, but I I'm not sure if this if I have a valid metaphysical stance here, but I I very much see both sides of the argument.

SPEAKER_02: 23:28

Yeah, I think the answer to your question, Sid, yes. I do think that when you combine simples together, like we do have new complex objects. I think like, yeah, it's it's interesting to relate that back to sort of Chat GPT or you know, large language models in in general, like with the simples, you have like data, parameter, words, and then sort of the complexes, you then you have like meaning, meanings and sort of responses, you know, based on those sort of the simples. And so I think the difference there between you know, like a large language model and human is that the unity comes from sort of like the structure, not from you know, it's not from like experien experientially. So I think until you know the large language models and robots like in general like are embodied, until they have like a body in the same way that we do and can interact in the world in the same way we do, they are they're definitely like limited in how um they bring about the unity from those complexes.

SPEAKER_03: 24:33

Yeah, I think this is a really challenging point to conceptualize, right? That on one hand we have to accept, well, I understand exactly how the computer is created, I know exactly how the transistors are laid out, and I can entirely deterministically recreate its behavior, right? It doesn't exist as this organic object. Uh whereas on the other side we have humans, which are made of cells, and we understand how individual cells work. Uh, we don't yet have a good understanding of how millions of cells can come together and create life or consciousness as we understand it. And so we have to reconcile what is the LM not doing that the human is doing that makes this distinct property of life be assigned to one and not to the other? The AI interacts with the world in API calls and through text and you know, making chess moves, but somehow it doesn't exist in our world, and so I guess the question is, do you feel like there's some something tangibly unique in the physical world that differentiates us from the machine? Is there some kind of notion of mind stuff?

SPEAKER_02: 25:44

Yeah, I think that kind of goes back to sort of like the mind-body problem, you know, in philosophy that we talked about um during the last um podcast. But from my perspective, yeah, I mean I I you know approach things from I think consciousness is um fundamental and I think everything is sort of derivative from consciousness. Um, and so I think there are different degrees of consciousness now. I I think sort of like everything has somewhat of a consciousness, whether that's you know, humans, trees, bees, blades of grass, Diet Coke bottles. I think consciousness is inherent in all of those things. It's just like to a to a you know, what is the degree of consciousness? So I think a lot of um sort of like data scientists that I interact with are more materialists. They more believe that like consciousness is an emergent property, like in a complex system. And I struggle with that one. Uh but yeah, I I do think that humans have some somewhat of a secret sauce that that large language models do not, you know, currently have. And I don't see them having that anytime soon.

SPEAKER_01: 26:52

Two interesting, so I was digging into this mind-body problem, Michael, like you were saying, and two interesting quotes that I I had read. Um, one of them was from the Mumford, it was mentioned, which is Locke's definition, and then another one from Descartes. I did the uh meditations on first philosophy as well the last couple weeks. So Locke had consciousness is uh an experience of thoughts and sensations that has memory, beliefs, hopes, and emotions. So that's where like different than like the computation a little bit of like the what actually will have hopes, thoughts, dreams, beliefs, emotions, right? And then Descartes had uh a mind is a thing that doubts or uh in context of a person, but that doubts, that understands, that affirms, that denies, that wishes to do and not do this and not that, and also imagines and perceives by its senses. So it's kind of like a little bit more of a holistic, at least by these those two philosophers' definitions, for what it's worth, a little bit more than that's where the I struggle with reductionism approach on that, right? Because there's a there's a little bit, there's some secret sauce here that's different than like a computer is a computer and it's awesome and it's yes, it's super powerful, but it's not thinking. It's not or not, well, I'm not even talking about thinking definition, but it's it doesn't, it doesn't have consciousness. We're just talking about consciousness definition here, right? Or reality. It doesn't have hopes or dreams or anything. And the large language models don't either, they're predicting the next word, right? Like they don't have that consciousness, like the the beliefs or that that mental life. So I'm not sure like how it that's a whole whole nother topic for another day, but those are just two quotes I found interesting on like the mind-body problem of I think like the reductionism is more like the body, right? Like we can kind of get to the body, except I mean scientific science has not progressed to like we can replicate a lot of body parts and things like that, right? But like in general, we can prosthetics we can kind of start doing the body. But the mind, it's like the that emergence that happens there is not something that even we take the computer as an example. We're nowhere close to that. LLMs were nowhere close to having the they're no matter what, like we talk about in Terminator and things like that, right? But they're not actually having hopes and dreams and things. And and an LLM is not reasoning, it's just saying the wrong answer and hallucinating, but it's just because it made up a word by predicting the next word, right? Like doesn't mean that it's actually having the hopes and dreams and desires.

SPEAKER_03: 28:59

And uh so I guess this brings us to you know one of our one of our final topics for today, which is uh if we want to talk about more modern philosophy, I know modern is a technical term in philosophy, but more something that's happened in the last 200 years. You might you might say, like, oh, well, all this metaphysics stuff was done back in the day of Plato. What's up what's up with today? Right? Have we been thinking about this problem since then and what have we come up with? Uh I would point us to the field of existentialism. Uh this comes in the late 1800s, early 1900s, and we see these ideas come across which privilege, you know, maybe to more Michael's position, that even if we exist as mechanical beings, even if we are just automata, we still have a valid personal subjectivity. And through that subjectivity, we have to understand that it feels real to us, right? Even if we are just a machine, you know, you know, defined by chemical properties. and and biology and evolution, we still exist as subjective beings. And so there is like value and privilege in making that a real thing.

SPEAKER_01: 30:12

Very interesting. So even like to rephrase and disagree if I didn't get it right, please, please correct. But it's like, so even if we're taking like the reductionist view, we're still this secret sauce that we haven't figured out yet, right? That makes even if we're automatons, there's something that we're doing differently that we haven't been able to figure out how to do with computers.

SPEAKER_03: 30:31

I I guess it it's that but it's also like saying that like because it feels real that can be enough. And that this is almost like you know a more like postmodern philosophy stance where we're we're looking beyond you know what is literally true. This is a response to people saying like well in the industrial revolution we have all the machines and we understand how machines work and maybe humans are just machines. And we kind of stepped away from this understanding that science is a way of describing physical mechanisms but not how we live life. And that those are distinct things. And so like you know the special sauce is just the subjectivity. It's just understanding that we have a viewpoint of the world and that objects exist in the world because we foresee them as existing in the world.

SPEAKER_01: 31:13

I found one quote that I was going to save it for later, but I think it fits well here at the near the end of the Mumford book as well that it's like so there's a lot of criticism on metaphysics should it even exist or like all the things we're talking about are really fascinating but like we nobody can prove any of this right like it's unlike there's no empiricism in metaphysics. There's no way we can go out and test this right so like this is all these like viewpoints and how do we think about it and like that's really the science metaphysics is thinking and breaking down these things but I thought this quote was great is even if metaphysics is useless from a science perspective its insights may be so deep and so profound that it could have the highest intrinsic value to us. That really resonated with me and I think it comes to this here of like the thinking through these things and and trying to figure out how the pieces fit together and what it actually having to sit down and figure out how to define and then at the end of this series we're going to try and have a definition of thinking but like actually breaking apart these ideas and trying to no pun intended think through how these things work is is is super valuable and like it doesn't matter that there's not empirical evidence for these things or doesn't matter if if Michael is on and Sid are on the Aristotle side of the house and I'm on the Plato side of the house. None of this actually matters but it's a good way of what we're really missing in today's world is people coming sitting back and thinking and actually figuring out what you believe. You're not just like pattern matching or just like turning your brain off giving your agency away to LLMs and just like actually reading and learning and trying to figure things out wrestling with the hard problems. It's like going to the mental gym is kind of metaphysics right so it's like it has such intrinsic value of figuring this stuff out even if there's no right or wrong answer here. But like the mental gymnastics is where the value is and helps us better perceive and get perspective I think same as like where Sid and I talk about all the books all the time right it's like books specifically history and biographies and stuff really help you get perspective on the world and really take a different angle and take a step back and like especially in the AI here it's like it's so easy to get on the you know the the treadmill or of rat or like the hamster wheel if you will of like the news and all the the hype and all the like we're almost in a post-truth world of whatever the LLM puts out on the internet is what the other LLMs read and then keep spewing and we kind of get these cycles. It's just kind of a crazy thing of like sit back do the hard yards do the mental gymnastics read do things the old-fashioned way no matter what technology you use makes you be able to see that perspective and someone's got to be the one that's understanding how the pieces fit together. So I know a little bit of a mini rant there but um I don't know that's why I'm I'm very much enjoying this series and excited for where it goes and I think all these things that might seem trivial like really doesn't matter the Plato or or Aristotle. I actually think it really does. And it doesn't matter what the right answer is it matters that we're having this conversation.

SPEAKER_03: 33:58

Thank you for those you know that that that was a really great you know set of concluding thoughts.

SPEAKER_02: 34:03

I strongly agree with that. I think that I love materialism reductionism like it's brought us to where we are today. But I do think that it's it's gotten to a point where some of the problems that we're trying to trying to look at and that we're you know that we're um that we're facing right now like I just do not think that they're going to be solved from like a materialist reductionist you know standpoint. I think we do need to take a step back as Andrew was saying and take a look at you know history and also take a look a little bit more at like um uh quality qualitative you know type features more of the subjective um as well yeah absolutely and and that's kind of where I want to you know end this off too is that like the great project of metaphysics for a long time was science and creating science and mathematical understanding and and models of the physical world.

SPEAKER_03: 34:50

But I think that the problems we're facing now of you know maybe people don't live in a shared reality. And that's a fundamentally you know earth shattering thing for these uh these fields to answer they can't answer these kinds of questions. And so we kind of have to look back and you know these fields that we called soft sciences like psychology and social science are becoming more important as we need to understand how we can interact with each other and how we can construct shared realities uh which we maybe we took for granted before I I love that I fully agree.

SPEAKER_01: 35:23

I think that soft there's nothing soft about soft sciences or soft skills, right? They're actually becoming much, much, much more important. If we're getting more automated things and things, the what makes us human, no matter if if we're automatons, whatever the whatever we are, how we interact with each other, how we understand how how the social sciences, how things work together, that's the most important thing now. And that's what just pains me a little bit with like the massive AI bubble and part of my PhD work was like looking at the tulip crisis in the Netherlands and like these bubbles and like the bubble we're in right now on the AI side like if you actually look at the you know like the what Wall Street is the to get an ROI what you're assuming is going to happen is an interesting thing to look at. In any case what I what it pains me a little bit is we're so focused on just replicating entry-level job workers knowledge versus creating knowledge right like in most of the history of mankind like in science and technology development we're creating knowledge we're not focused on creating knowledge anymore we're spending trillion dollars to try and replicate basic knowledge right so like that's why I think that this all of the social sciences like you said Sid is that might be some of the next frontier of of really good research is like figuring out what what it what it is to be human or how to interact in those things and like even if if the AI and I'm skeptical that's going to get near as performant as advertised but even if it does great how do we operate with that now like those are the big questions and metaphysics is really uh a linchpin to those conversations yeah I I think that's that's absolutely that's absolutely valid I think these are the kind of questions we'd love the listeners to think about and pose themselves and you know please let us know if you have any thoughts.

SPEAKER_03: 36:58

We have it we have an inbox that we're always happy to read emails from and we're we're on LinkedIn and we're on the all the socials. So thank you every for your time today. We'd like to uh you know remind people that this is a mini series so there's going to be another episode coming up where we're gonna be talking a little bit more about the relationship between space and time another piece of metaphysics and we will not have just one but two exciting guests in that episode. So very excited for that episode. I think we've already been advertising this one to people uh and I think you'll have a lot of fun with this as we work through from the beginning you know how does space work and how does time work and how do those two things relate.

SPEAKER_01: 37:35

So very exciting episode coming up thank you so much uh Michael for joining us we need to make this more frequent than every year and a half or so I guess what it is you need to have you on the show more often and um yeah really appreciate you being on and this was this was a great podcast thank you both yeah this was fun thanks thanks for having me yeah this was a great time glad to have you on Michael for our listeners we hope you've enjoyed this episode as a different approach to thinking about the modern age of AI through metaphysics.

SPEAKER_00: 38:07

If you have questions about this episode the series or any of the content in our podcast please leave us a note at ai fundamentalist at monotar.ai until next time