Authorship, Creativity, and the Future of Work in the Age of AI

In this episode we, Thomas and Mike explore how generative AI is reshaping the way we think, create, and understand ourselves. Starting with a quote from Ted Chiang (“ChatGPT is a blurry JPEG of the internet”), we unpack what it means to be an author when machines can generate stories, music, and ideas that surprise even their human collaborators.

We discuss the philosophical and practical questions surrounding AI-generated content, the shifting value we place on craft and effort, and the risks of outsourcing too much of our thinking. The conversation touches on authorship, productivity, memory, and the idea that we may already be extending our minds through technology, often in subtle and unintended ways.

From Mozart to Moravec’s Paradox, from Google Maps to Neuralink, this episode is a deep dive into how AI is challenging what it means to be human, and what we might become next.

Find out more about the work we’re doing:
🌐 Odysi Studio — https://odysi.studio
🔗 Mike’s LinkedIn — /michaelvtucci
🔗 Thomas’s LinkedIn — /thomastrincado”

Transcript

 00:00

Thomas: Hello, everyone. Um, before we start, there is a line that I want to share with you that I keep coming back to. It's Ted Chiang's. He says, don't think of ChatGPT as a mind. Think of it as a blurry JPEG of the internet. And I think that's quite clever. But it's also a bit unsettling.

00:19

Thomas: When like, what happens to culture when it starts being a blurry picture of itself? And so these are some of the questions we want to explore today. We want to talk about what it means to be productive, what it means to be creative, and there's questions around, you know, what value is, what even originality means in the in the wake of of AI, you know, when when machines themselves, they can do all that work. And so, you know, I'm very excited to to be talking to my friend, Mike. Um, you know, you've if you've followed this series before, um, you know, we've had conversations in the past. But what I haven't mentioned and and what I haven't been very explicit, um, is Mike has a very interesting combination of deep systems thinking. And he also brings, I guess a set of very special and unique um, like personal experiences. And so I think um, you know, these experiences come from him having worked at some of the largest, um, like tech companies in the world, and most powerful as well. But also like having lived in many different countries. And, um, so yeah, I'm very excited to be having this is a conversation between the two of us. We'll be exploring things around cognition, about embodiment and just, you know, how we in the, like, how we relate to machines. And, uh, so let's dive straight into it.

01:56

Mike: And uh, Mike, welcome. Thank you, again. And I mean, I can also say the same about you, I think you're one of my favorite people to talk to about these topics to think through these topics with, right, like, we say all the time that like, I think nobody really knows exactly where this is going. And everybody's just trying to figure it out. And I think you also bring a really interesting perspective to the table, which is why we founded Odyssey together.

02:22

Thomas: Very nice. Yeah.

02:24

Thomas: And, um, I mean, of course, how are you? Like, you've been quite busy with Odyssey lately. Um, you've been exposing yourself to many, like to a lot of these questions that we we we we're we're going to be exploring. And I do have a a list of questions that I, I mean, I I've said this in the past. Um, part of the reason why it excites me so much to be doing the this podcast with you is because I have your undivided attention for like two hours and a half.

02:50

Mike: You always have my undivided attention.

02:51

Thomas: But like, you know, when we're out like having a beer or whatever, it's just not the same level of, you know, where you're like the granularity to which you you can explore ideas, right?

03:01

Mike: I I agree with that. And I think I've said this to you privately, but these conversations, these podcasts are actually some of my favorite thing to do because it's a chance to, um, zoom out and kind of think of the the big picture implications of the techno technological change, um, all of the new capabilities that AI brings to the table, that to think about like what is the the broader, what may the broader impact of all of this be on humanity, on our future, both the near future and the the long-term future.

03:34

Thomas: Right. And, um, okay, so because I have limited time with you today, I want to like get straight into my the main question, which is

03:45

Mike: Let's go. Yeah.

03:46

Thomas: we, you and I have talked about, um, what it means to be an author in the, in the age of of AI when generative tools can create a lot of these things for you, for you, right? Is it are we talking about shared authorship with with um, AI? Are we talking about us being the sole authors and then having tools that help us? How do you see this?

04:09

Mike: Um, so it's a really good question. I think first and foremost, I think we have to be really careful. I so there was something that you said in that question, which I wasn't necessarily expecting, which was shared authorship. And I think we have to be very careful not to anthropomorphize AI. And I'm not saying that AI can't in some ways be an author, but um, it is not in my conception a a person. I think it is more for the moment, it is more, all the evidence we have is that it is more of a tool. It it has grown out of like the broader technological capabilities, quote unquote technosphere that humans have created. Um, and like other tools before it, it is um, a tool that we use, a capability that we've created that helps us create things, create art, create um, uh, you know, text and writing, um, create culture and um, so I wouldn't personify it in that way. Um, so yeah, first and foremost, I just kind of want to say that. Um, I guess the other point I bring out is that sort of tool aspect, right? And we've um, we've always used tools to create things. We've always used um, uh, the we've always built things out of like the broader cultural knowledge base. Um, the, um, the societies we live in, we're we think of ourselves, we really kind of like put on a pedestal the individual. And so we think of, um, individuals as, you know, this kind of privileged category, and I think that's just because it's a useful conception for us to like move through the world, like we are embodied beings. Um, but I think very often it, um, it it somehow misses the full way in which the creative process works. It's even when it's the, it feels like it's a lone genius, it's really not. There's a lot of other interactions with actual other people and with tools that um, co-create anything.

06:25

Thomas: So, um, I mean, I I do see your point, um, that like they are tools and when but they're tools that semi-autonomously are creating, it's not like using a typewriter. That is very much in its it's very clear that that's a tool, the same way that paper and pen. But this time around, we're creating something, it's not Microsoft Word that we've created, we've created Microsoft Word that with a bit of assistance and guidance will create a story that you didn't even expect. When the author of like the the supposed author of a story, him or herself is surprised of the outcome, right? When you're writing in on a when you're typing words into like into Microsoft Word and hours later you read it, like you know that you wrote that, you're not surprised by it. But I think it changes the whole dynamic of like calling, calling AI just simply a tool. It's, it's, it feels, I think this time this time around feels a bit different.

07:35

Mike: Yeah, I mean, I I think that is a very good point. I think it's hmm, I think you're right. I think it's capable of generating actual kind of full content. But your question was more about authorship and I don't, I think if you're using it as a tool, you can still call yourself the author. I think if you're using it, um, if you're letting it do the creation, then um, you're no longer the author, at least no longer the sole author. And then I think it gets much more complex. Like who do we attribute it to? Do we attribute it to, um, Chat GPT? Do we attribute it to the training data that Chat GPT? So like essentially the internet that Chat GPT, um, uh, was trained on because that that didn't come from from nowhere. But, um, I think if you how do I want to say it? So, um, if you think of like, even centuries ago, you have like Mozart, I'm not a composer, but you know, you can imagine like Mozart, um, composing music and you have this like, you know, at least I have this image in my mind of like Mozart like kind of with his piano and like kind of writing notes on a a page. Um, but then in reality, right, like he would, um, bring that to, um, uh, I again, I don't actually know how this works, but I know at some point they would be brought to at least a small group of, um, actual like a a small orchestra and eventually a full orchestra. They would play it. They would have these people, like actual human beings who know how to play the instruments. Um, he would hear that in, um, a certain context actually with the instruments playing. Um, there is a, I think a lot of, um, talk like famous talk about, uh, uh, like renowned composers being able to play the music in their head. But then inevitably, like what you're faced with in reality is different than what is played in your head. And so like your conception comes into collides with the real world, collides with other people who help then shape the experiences they create, their, I think very often their actual explicit input, um, helps to shape that piece. And I think if you are, um, so I I guess one, that has always been a part of the creative process. Um, two, if you are still kind of the composer, if you're the mastermind, if you have the vision, you're saying, this works, try it this way, um, uh, do this, um, uh, maybe somebody gives you a suggestion, either a person or a, uh, an LLM and, um, you, you, you're the one deciding this goes with my vision, um, fits with my vision or it doesn't, then I think you can kind of rightly call yourself the author. If your, um, uh, allowing, if you're kind of giving that creative judgment to somebody else, then I think you are at least sharing authorship, um, if not, um, and you know, it's a continuum, I don't think there's like a clear line, um, but you're yeah, then you're not the sole author.

10:53

Thomas: Okay. There was, there was a case, um, um, I think it was, um, 2023. Um, it was right when, a few months after Chat GPT came out. And, um, Science Magazine, which is, you know, arguably the most important, like scientific magazine in the world, um, said, um, that it would not allow, it would not accept, um, claims of co-authorship with AI, right? A human was ultimately responsible for the content that was being written, and, um, AI was regarded as as a, as a tool. And I, I know, I, perhaps we don't want to get into the, the, I'm probably a lot of why this was said is more a sort of a legal framework, and I'm thinking more of it like the angle at which I'm going at this is, um, more the philosophical angle, right? Um, I do see an argument. So when you say, for example, that, um, in the creative process, even as individuals, we, um, derive, we get so much information, um, you know, when when paint when painting, um, sort of when coming up with a new style, are we really coming up with a new style of a new way of painting? Or are we sort of, um, absorbing everything that we've seen before and then kind of coming up with something slightly different? Um, now, is, I feel that that's a lot of what AI does. Um, I think AI is able to come up with genuinely new, um, ideas, right? So, leaving the legal framework aside from a, from a purely, um, philosophical standpoint, do you really think, um, that there could be sort of a relationship of co co-authorship with with AI?

12:54

Mike: Oh, um, I I think that that is a I would, it's a hard question to answer. So, I guess for me, and I usually, I mean, I should say I'm not like, I I create art from time to time, but I'm not mainly an an artist. I, I do, like, I played around with creating with writing, with, um, creating content and videos now and podcasts. Um, I would most, um, try to stay away from having AI as a quote unquote co-author, unless it's like a statement or an exploration of what AI can do itself. And I've actually thought of doing that. Like, try and you, you actually, you'll see these like TikToks with, um, um, people like having AI, having ChatGPT in voice mode and having a conversation and it's kind of funny or, and, you know, that that I think is interesting. I think that's like a commentary on, um, how this technology is evolving and how we're using it like it's kind of a a companion or a person even though it's not. I think I would put that kind of case in a, um, in like a special category. Um, separate from that in my own life, I try not to use it as, um, so I yeah, I haven't thought that, I'm I'm kind of exploring this out loud. I think in those cases, the thing that makes it interesting is like that right, like I can watch that TikTok video, like I see ChatGPT there in some sense, like it it is clearly, I can see what it's producing. Um, the thing that I think is problematic, so if you're like explicit about that and you're somehow showing the audience that you're using it, how you're using it, if you're playing with that, I think that's actually really interesting. I think what gets, um, what I want to shy away from, uh, is using it, quote unquote, as a co-author and then claiming sole authorship. And I think like, I don't think this is just me, I think we shy away from this as as like humans, like our sense of taste is somehow offended. Like there's so much content that you see on LinkedIn where it's just like my brain, you know, we're really good at detecting patterns and my brain just pings like I can see the like we've talked about this a little, you like the word quiet now or chaos or this, not blah, not blah, just blah. Like there there are these tropes that, um, that AI repeats over and over again and it's boring. Like it's, it's not even that like I, I think we should be able to do whatever we want. Like I, I think I, I'm fine with people playing around with it, but so much of what makes life interesting, so much of what makes talking to other people interesting is the, um, the idiosyncrasies, the imperfections. I think it's very interesting that like all of even, even when you're using different models, they all seem to kind of, uh, settle on this least common denominator of like ways of talking and style that, do you think it will be possible to, well, do you think it's possible to prompt those models away from behaving that way and if not, do you think it will be possible in the future to have, um, these models replicate human imperfection?

16:39

Mike: I I do think you can do some a bit of that with prompting. I think it's way harder than you think. And I think this is there's so much and we talked about this a little bit last time, there's so much that we do as humans that we don't understand. Um, and the way that um, the way that we the way that we think is affected by our differential histories, it's affected by the time of day, the place that we're thinking, the like even the thoughts in my own mind change so much from uh, and like there's studies and whatnot that like it's not just my kind of perception, like they it actually, I mean, everybody has this this, um, I think experience of trying to think through something particularly difficult when you're tired versus like when you're, you've slept well or like you go for a walk and it clears your head and then you have a new idea. Um, there are so many things that go into the idiosyncrasies of our thoughts and how they unfold. And AI is not, I mean, it's not even close to that right now. Like it, it makes sense that it's, um, it gets trained, like there's a training run and then it gets puts, it gets put out in the world. It doesn't, it's not remembering, it's not learning from any of the the interactions that it's having. It's not able to walk around in the world. It doesn't get tired, it doesn't. And we assume that that's like a negative about humans and like a positive about AI. Like it can just do as much as we want and like it doesn't have to like feed itself or like, you know, go and clear its head. But I think that's actually a, I think that's a beautiful thing about humans, and I think it's the thing that makes, um, a lot of the content that AI produces much more boring.

18:37

Thomas: There's been cases where um, it's happened to me and obviously they're documented all around the internet where people have seen that AI suddenly becomes lazy. They're like, oh, I'm sorry, can't can't do the can't help you right now. It's like, excuse me. Like, I don't know if if it's due to uh, traffic overload, but there is a sense that AI even gets tired. It's happened to me many on many occasions where I'm trying to get a very simple answer, it's just gives me like the laziest, like cheest answers. It's like, sorry Thomas, like you're when you're asking, it's like it's even, it's even told me at time, like, you know how these models they they tend to be very um, sort of people please, right? Yes. It's even told me is like, sorry if you ask me like a version of if you ask me a question like that, I'm not going to be able to answer because it's super unclear. Like,

19:32

Mike: so it's funny, I actually, I think I, now that you say this, I have heard this a bit myself, but I haven't run into it. Um, but I go back to how are these things trained? You know, and what are they trained on? They're trained on human language. So things that humans do and ways that they respond in certain situations are patterns that it's going to pick up on and it's going to replicate. So like the reason I would think that it's doing that is because humans do that and it's seen that language before and in that situation, it feels that that's

20:10

Thomas: But isn't that how humans behave as well? Like, don't we, like, aren't all our responses also a consequence of what we've seen in in society? Like we wouldn't, I mean, even language itself, like it's acquired. And so our can't like if I, if I say specific like, if I dismiss you with a joke, it's because, you know, my sort of my cultural understanding of how I can dismiss you is like with a joke.

20:34

Mike: I, I mean, I'm just going on my own model of being a human, which I swear I am one. Um, uh, like at least my feeling in those situations is like, I am actually tired and like I want to go away or I want to, I want to to end the conversation or like I don't want to do the thing that you're asking me to do. Or that like I want to make you laugh. Like that's why I'm telling you a joke. Like I want the social feedback. Like I feel that would be appropriate, I feel that would be like interesting for me, fun for you. I'm going to get like a, uh, um, a rush of dopamine from, from like having somebody laugh at my super funny intelligent joke. So, so yeah, I I think, no, I think it's actually quite different when humans do it than when LLMs are doing it. Like very, very different. Just and this is like a general thing. Just because it converges on the same behavior, doesn't mean that the underlying mechanism and, you know, very importantly motivation for why it's doing it is the same.

21:37

Thomas: It's it's funny because you've mentioned something um about, you've specifically mentioned the example of going out, going for a walk, clear like clearing our heads just to be able to think better. And um, you know, and and I was recently um, I think it was two people and you were one of them obviously, uh, who recommended me this book about the the extended mind. And um, one of the, one of the ideas that is explored is the idea that, you know, we talk about productive work and we think of productive work as like there's a very specific version of what that means, right? Like it means being in front of the computer and it means like answering emails and doing this, etcetera. And sometimes actually some of the most productive work happens when you go out and clear your, clear your mind, you can sort of take a step back, right? So what, um, I want to explore the concept of productive. What does it mean to be productive um, now and in in the coming years and in the future in the like, you know, with AI powered in a in a in an AI powered world?

22:45

Mike: Um, I want to like put a pin in before we move on to this, can I put just a pin in the the authorship conversation because I feel like we started on authorship and then we kind of got into the way that um that we interact with AI. I I think that we should be, I think it's actually very interesting to use AI as a tool, as a creative tool. And I think that and this relates a little bit to the question on productive work, right? Like, um, whereas before I would need to learn so many, like if I wanted to, um, to produce music, right? And I don't play instruments and I, you know, even with electronic music, like I don't know the software, etcetera. Um, I still, I love music, right? Like and I, um, I know enough that I can, um, and I have a like I have my my sense of taste and the things that I like. And just like somebody who, you know, actually knows music well, like I think I can, um, think of interesting riffs, take take something from one song, take it from another song and combine them in ways that are interesting. Um, previously, I would have to learn a lot of just like very like technical skills, um, in order to do that effectively. I think the softwares that we are creating and a lot of, you see this with like, you know, programming, etcetera, like the vibe coding is is, you know, another version of this. Suddenly, um, you can create something without having to go through all of that, all of that like technical training. Um, and I think that's actually great, you know, I I still think that the people that are going to be the best at creating music are going to be the ones that like stay immersed in it and maybe, yeah, maybe now they don't have to like, you know, learn as many instruments or learn as many um, uh, how to manipulate certain pieces of software in the ways that they did before. But they're still going to have to like learn a lot about like how sound combines, etcetera, um, in order to make good music. And I think that's like actually 100% fine. And then I think we are also going to see, as you suggested and was a really good point that you brought up, um, that uh, like the generative, the truly generative element of AI, um, we're going to start to find interesting ways to play with that where it's like in conversation with a person, we're going to find ways to actually like maybe make AI output more interesting and less like predictable. Um, and I also think that uh, that is like very interesting and fine. And I would not, um, I'm like you at the end of the day, this it still like it's this is being made for humans to consume, right? And like it has to be interesting to humans. And I think as long as um, I think our our taste is going to be able to differentiate between good uses of AI in creativity and bad uses of AI.

26:08

Thomas: Because I mean, I know you said that this was just going to be a pin, but like we're just going to go ahead. And um, because you, um, how many times have I gone to museums, right? And I've seen especially with modern art, um, people say, oh, I could have, I could have made that. Like my child, my five-year-old could have, could have done that, right? And um, how many, how many people have been sort of completely um, inspired or impressed by something that's I mean, it doesn't matter if it's beautiful or not. It's difficult, right? The they're impressed by big cathedrals because they took 200 years to make. Um, they're they're impressed by the Mona Lisa because it was, it took, it was very complicated, it was very, um, very craftful. Um, and so what happens, how do we even value the output of AI if we don't have the craft dimension of it?

27:12

Mike: But I think you still do have the craft dimension, right? Like I, I love the uh, the analogy to modern art. Like, yeah, maybe like where there's a famous one of like the banana tape to the wall. Forget who the artist was, but and every like when he when he like he's done it in a bunch of places and it's literally just a banana duct tape to a wall and he goes and gets the banana like from a stand by whatever museum he's doing it in and like goes in and duct tapes it to the wall and I think it's like some sort of commentary on like what is art and like what counts as art. But like so much at least when I look at that, like, yeah, your child could do that. First of all, your child didn't think of that idea and it's actually, the reason it resonates is because it's sort of it's both funny in a way that like, you know, resonates broadly and it's a crazy um like the fact that he convinced, uh, he built himself into an artist who could like walk into like the Moma or something and like tape a banana to the wall and have that be okay and have people like large swaths of like the art community like just be like this is really interesting. Like that is part of what the art is. Like it, it is true that anybody could produce that, but not anybody could get it into the Moma, not everybody could convince like broad swaths of society that this is actually like the forefront of art and that's that's that's a hard thing. And that's going to stay hard, right? Like, um, yeah, people will be able to like I can, you're seeing on, um, on Amazon, I was just reading one of the another creator that I saw was like kind of making, I think I sent it to you, was making fun of the fact that like there were a bunch of like clearly AI generated books that were either about him or like, I think alleged to be written by him that were like put on Amazon. And um, yeah, like that happened, like but they're not actually by him and people are going to realize that they're they're fakes and he's not going to sorry, those those books are not going to be bought in the same way that this actual creator's like content and books are going to be bought and consumed. And like that that is a really important part of the the story. Like human taste is still going to be inserted and is still going to decide what actually is considered good art, good content versus, um, you know, like uh, AI created bullshit.

29:57

Thomas: But do you think I understand that um, I understand that there's a sense of human taste being able to even if it's if when we look at something that's been um, produced by AI, um, obviously there's there's a sense of a human looking at it and thinking it's good and and then um, in a way sort of having our validation not like it my point being, I'm not talking about so much about how it's perceived by the person who looks at it, but um, how influenced that perception is by the fact that it was actually created by a human, that it was, that it was difficult to make. And I'm wondering if we're going to put a premium on things as we go forward. We're going to put a premium on things, we're going to proudly present things that have had, that have been just done with without with AI. If it's even we're going to become a luxury in the future where, you know, something was designed, you know, like clothing brands that, you know, present their models like and and it's actual people wearing the clothes and it's not the AI generated and stuff and it's if we're going to pay an extra premium, we're going to appreciate it more because

31:15

Mike: Yeah, 100%. Like I mean, the this is like this we've already seen happen, right? This is like handcrafted whatever, you know, like leather goods or or I don't know, a million things, right? Like we mechanized a lot of production and then the suddenly the um, the way that it was done before, which was by hand, which had many more imperfections became um, a sort of a luxury version of that thing that was thought of to be better because of its imperfections, because of the care and the craft and the work, the labor that went into it, um, rather than like the mass produced, um, uh very predictable, very because right like also mass producing means that there are way more of them and anybody can, you know, can have that thing. Um, so I absolutely think that's going to um, uh happen with AI. I also think like we we're always in conversation with these as creators, with the these new technologies. Like a lot of why um, a lot of what birth more like impressionism and like the, you know, subsequent schools of more modern art was the fact that we invented the photograph, right? Like that we invented cameras, technology that could like before that artists were rewarded by their skill at faithfully reproducing whatever was in the real world. Um, suddenly that's going to that was commoditized and so they started um thinking about, you know, okay, what well, what what can't, what is it that a camera can't do? What is it, how do we deconstruct the image and And I think well like I mean I've said this to you in many ways and I never thought it as connected to art, but it absolutely is. Like um, I am so excited about AI because like suddenly it can do these things. And then we have to think about like what do we do that's different? And I think there's a lot of things that we do that are different.

33:12

Thomas: That is actually that that makes me feel very excited about what might happen in the realm of AI music and whatever. We see we seeing like new like these models and um and you know, full on SaaS platforms that allow you to interact with with AI to create like new pieces of music and and you know, like we've we've seen Spotify um like put a lot of AI produced uh music to um so that like they they don't have to pay royalties etcetera. And so it makes me very excited to see how all of this will evolve, like what is what is what are aspects of music that AI cannot replicate, the same way that camera cannot replicate about um like reality, right? Of course the limits get a little bit blurry now because like we're we're talking about like again, the main difference here and I I need to explore this idea a little bit more is that a camera really is a tool. A camera, the same way that a that the typewriter is a tool and and Microsoft Word is a tool. AI is able to come up with original ideas.

34:17

Mike: Oh, so you said that before. I don't think we know that. I mean it's like what is an original idea is, I think it's actually it's very difficult to say. And there's um, there's ways in which it seems like AI despite all of the knowledge it has, despite um, like we it it can recombine things in ways that seem interesting, but it doesn't seem to be able to create wholly new things. And like the the area that I've heard this being most salient is with like new discoveries, like despite all of its knowledge of like math and physics, like it hasn't come up with any breakthroughs. You know, like um and and there are labs, you know, researchers trying to push this specifically and like it literally has access to all of the books, all of the knowledge, it has perfect recall and it still isn't able to do what human minds can do to really push the forefronts of any field. It seems limited by the things that it was trained on. And of course, like it can recombine those in so many different ways that feels like creativity, that feels like generating a new thing. And it might be new in the sense that like yes it didn't exist exactly in this way before, but it I think there's a sense as well where it's really a recombination of what existed, um, without anything really new inserted and that there's still something, I'm not saying AI won't eventually be able to do this, but right now there is still something that our minds are doing as humans that and I I don't think this is proven, but I feel from the things I've read and from how I interact with AI, I intuitively feel it's true, that um, that we are able to really come up with genuinely novel ideas to like that are not merely some recombination of what we've been taught and what existed already, that AI can't do yet.

36:25

Thomas: Hmm. And I think we'll like, you know, part of the creative process is like continuing to do that, you know, and then we'll do it together with AI, but yeah.

36:34

Thomas: I'll tell you um it's slightly it's vaguely related to um, I mean because we've explored the idea that um, how when um the we we came up with industrial processes to to be able to, you know, create um like jackets and clothing etcetera, then we we started putting a premium on like luxury brands that um pride themselves in being able to or like having people have the like create the whole thing from scratch by hand, right? And um and also relating it to art, I'm going to tell you a story of um of a friend of mine who works at um Tokyo University and um he him being in the linguistics department, he um he had a lot of contact with art students, Japanese art students that went, you know, from Tokyo University, the most prestigious Asian university, right? They go to places like Paris to study and they came back very confused because they said all I've done for a year, for a whole year is explain my art, right? There was nothing about the execution of it. So I think it's a very specifically Japanese way of looking at it. Um kind of

37:48

Mike: But these were students from where? From

37:49

Thomas: Tokyo University, Japanese Tokyo University. And they were studying what? They were studying art.

37:56

Mike: Okay, yeah.

37:57

Thomas: Right? And so they'd go and and and live in Paris for a year and they they they came back very confused saying all I've learned all year was justifying myself. I like it's and sort of relates to the to the man who

38:09

Mike: Having lived in France, I I I bet that was true.

38:12

Thomas: But like it happens in the Western world, right?

38:14

Mike: Yeah, yeah, yeah, yeah. But in in France in particular.

38:19

Thomas: And there's there's I mean, obviously cultures look at it in different ways, but there's this appreciation of the of the craft, the process. Often times it's not so much about the the meaning and significance of it is like it's just a deeper appreciation of the process. And and and that sense of like producing something with your hands, etcetera. And so and it ties to the question of productivity, right? Um, because in what it the same way that this is a this is a way in which, I mean, it's it's proven in the past that this is a way in which, um, our perception of productivity will evolve. It's not we we actually pay more for less productive processes, right? Like more communion example.

39:04

Mike: Yeah, like a sort of an expensive leather jacket that could have been produced.

39:09

Thomas: So yeah, okay that, that, right? That so that it's been proven, like how do you see this going?

39:15

Mike: So I, I think I it's funny, I didn't understand the connection at first, but there is a clear connection between um, what we were talking about in authorship and the and productivity. I think very often, like we we think of um, the the human as a worker is and this this proceeds AI. The human as a worker, like the model is like of the best worker as as being the human as a computer, right? So like it just, we just need to compute and produce and produce output, um, uh, and if you're not producing enough output, like you just need to work more hours, you need to like stay in the office longer. Um, you need to like, I used to be a consultant, you need to make more slides, you need to like, you just literally have to, um, produce more information, produce more things, like whatever, whatever it is. And I, I think we've already seen and again, before you bring AI into this, that like one of the things technology allowed us to do before AI was to produce a lot more of like useless shit, right? Whether that was like, um, uh, like useless conversations in Slack that like weren't that feel like busy work, it feels like you're like solving a problem, but like no, you're just you're like just writing. Like you might as well be writing to an app.

40:58

Thomas: I've had a few of those.

40:59

Mike: Yeah. Um, uh, I mean, but and it's not just, it's not just like information actually, it's also physical goods, right? Like I think of all of the, you know, we we, um, moan about like consumerism and rampant consumerism, like all of the, you know, Trump like ranting about the cheap shit coming in from China. Like, yeah, maybe some of that's overblown, but some of it is true, right? Like we, um, I don't think that, um, output for output's sake is productivity. You actually have to, like the way that economists measure productivity, you actually have to like produce things that are valued. And some of the things that, um, that are most valuable, like are, you know, may may not even be like exchanged for money. Um, and I think I think there are a lot of things that humans do that we produce, that we do in the firm, in the work place, that um, that aren't about just that sheer output. And I think a big part of that, like you think about knowledge work is the quality of the ideas that we produce, right? Um, and so you have this, I've heard a lot of, um, podcasts, a lot of, I've read, um, some essays, uh, about the the fully automated firm, right? And they're they're just talking about, imagine if you had like, uh, Steve like 3,000 Steve Jobs, like, you know, you just you could because of course AI, you can just like re you can just copy that. And yeah, it's going to take some computing power to um, uh, to to like run those copies and you're gonna have to pay for them and make sure they're worth it, but like you could just have as many Steve Jobs as you want. And like, I don't, I like I don't actually know if Apple would have worked well with like a thousand Steve Jobs. Like a lot of what's happening with and I said this in our last podcast, but first of all, like, um, uh, organizational behavior and like social behavior is one of these, did we talk yet about Moravec's paradox?

43:17

Thomas: No.

43:17

Mike: Okay. Um, uh, Moravec's paradox is this idea that, um, the things that computers can do well, um, are the things that are hard for humans. And the things that that humans and animals can do well are things that are actually the most hard for computers. So you can think that like it is really easy for a computer to like multiply like 30 digit numbers by each other and very fast, whereas like there's no even with a pen and paper, it's going to take you a very, very long time. Um, whereas, uh, we tried to design robots that can like navigate the real world, that can like, you know, they're seeing a scene and we talked about this last time like, you know, uh, just pinpoint or home in on that like one little thing that's that's moving in the corner. Um, though that's very, very hard. And the reason for that is because, um, rational thinking, that's more this part of our brain that can like reason through things that can, um, uh, that can compute like, you know, difficult math concepts and equations. Um, that's that we evolved that very recently. Um, so like our brains, our bodies are not our genes, like are not mostly like coding for that ability. Whereas the much more difficult things like, you know, um, moving through the world, like eyesight, um, and and, uh, connecting sort of sensory perceptions with motor activity. Um, that has been, you know, um, selected for by evolution over billions of years. It's very deeply coded in us. Um, and so we take it for granted, but it's actually very, very valuable. And one of the things that, one of the things that falls into that bucket is our ability to, um, uh, I'm going to connect this to what we were talking about before, but, um, is our ability to, um, to cooperate as a species, our social capabilities fall into this kind of bucket that's really hard to, um, rep, it's hardcoded in us and it's really hard to replicate with machines. So like I am not, I don't think anybody's really tried it at scale yet, but I'm not convinced that once you kind of have a bunch of copies of AIs in a firm kind of working together, that they're going to do a really good job like making good decisions and creating shit because like one of the things that organizations are doing, I think is, you know, there there are a bunch of people in there who, um, think differently, who like have different motivations and we think of that very often that is modeled in like classical in classical economics as a, um, a downside of, um, of the firm, right? Like ideally everybody would just be a perfectly rational actor. But like I don't think that's actually true in practice. I think firms are, um, breeding grounds for new ideas. There's like a good thing that ideas need to be tested and proven within like somebody somewhere in the hierarchy has an idea and then they they want it to become a, um, uh, the next product that the company produces while they have to convince their colleagues, they have to then like create a prototype, put it out to the world, they have to convince customers. Like I I I think we're really making a lot of positive assumptions about how easy it is for AI, is going to be for AI to do that.

46:51

Thomas: It's assuming that is that sort of irrational thinking and the disagreement that happens within that um, pushes us closer to true innovation and like thinking differently etcetera. Um, you just put an idea in my mind so I'm just want to explore it with you, right? Which is I get that um, AI comes from a in a way collective mind, if we can talk about it, um, that way. Um, so it will be very difficult for it to disagree with itself even though like but still you could program different agents to think in different ways, to think from a, I mean and and I mean I'm I we've been there's been there's already thought thought experiments online with people um, forcing or having um different agents have a conversation between different agents that one has been programmed to think more like Kant, right? So even though it's it's, it's based of what what a human has said has um has said in the past, still thinking and like different from a person who's um, you know, made to think like Adam Smith from a more um economic perspective. And there is disagreement. So is solely disagreement that pushes us closer to innovation or is there something else? Like you're talking about sort of that

48:18

Mike: It's something else. I, I think it's something else. Um, so you are right that we can, you know, you're kind of are getting at this with like, oh, you can prompt AIs to, you know, think in different, quote unquote think in different ways or. Um, yes, but and my my answer to you was, yes, but like it's actually the different, it's the sum of the different experiences that people have, the different context in which they have different ideas that really kind of drives that idiosyncratic thinking. So I don't think it's going to produce the same thing. I think when we're talking about people, they're we are all a, um, a mix of this kind of lizard brain and I'm not even just lizard brain, the the like the body, you know, the the very sort of primal thing that we um, uh, that that causes our our our emotions, that causes our gut feelings, that causes our, as we talked about last time, our survival instinct. So we have that core with a, um, with a with a rational um, like our cortex attached to it. And very often, especially and it goes back to what I was saying before, very often like we assume, like we always talk about like the biases that that lizard brain brings to us and we like kind of assume that it's just negative, right? And that like ideally everybody would be that, you know, classical economic perfect perfectly rational actor. Um, and I just don't think that that is true at all. I don't think that the lizard brain is negative and the and the um rational uh, uh cortex is is is solely positive. I think that there's probably something very important because right, again, that lizard brain contains much of that um, literally like billions of years of evolutionary knowledge that um, is embedded in our genes, right? Um, and it that is then connected with this more rational um, more um, systems pattern-based thinker that can, uh, that that is not just rational by the way because it can also think very irrationally itself. Um, or irrationally. Um, uh, and I think there's a magic in that interplay. And I think if we, I don't I'm not a physicalist, like I think we could probably can um, model all of, I'm not a physicalist, so that means that like I don't um, uh, is it a physicalist or a dualist? I'm not a dualist. So like I I I do believe that, um, the the brain and like what produces our thoughts, etcetera, is is physical. I don't think that there's some sort of like magical other essence that like needs to be injected. Um, so I think if we modeled the full brain, I think we could also model the lizard brain, but I think that's going to be, I don't think we've done that yet and I don't I think it's going to be much harder. Um, and I yeah, so I I know so the the the the long answer to your question is, no, I don't think that we're gonna just be able to easily kind of prompt our way or um, like uh, you know, if you're if you're if you're training a, uh, an LLM to be like Kant, but it's not attached to Kant's like body and brain. No, I don't think it's gonna be Kant.

52:06

Thomas: I mean, I I I understand your your point, but you've also when, um, you've painted a very extreme picture. You've gone from, you know, humans interacting in the workplace to 3,000 Steve Jobs, right? Is there is there sort of an in between, is there sort of a combination where we, um, like what what what is the best version of our like and I'm and I'm this is a segue to the to my next question, right? It means um, there's there's moments in history, um, and and we've talked about this in the past about the the extended phenotype and this the extended mind. Um, what are interesting ways um, that we can use, um, what AI, the tools that AI gives us to extend our minds, to um, to help us, you know, be more creative or do you think, is there a version of the future where it's inevitably, um, human intelligence will become obsolete because, you know, AI will inevitably replace us.

53:23

Mike: Um, I, I mean, I I guess based on what I just said, like I I do think that we will we will eventually be able to create a solely synthetic mind that has the lizard brain and the um, rational brain and kind of figures out the interplay. So like I, I think that that's possible. Um, I think like we're going to learn a lot on the path to that. And my hope is that um, I think there is a version of the world where we actually just like create, there's nothing about like the laws of physics that forbids us from creating a super intelligence that like replaces us and takes over. Um, I don't think that we're headed in that direction and I don't think that we're headed in that direction because I think that the the the um, before we kind of create um, those more like synthetic minds that might replace us, I think we're going to um, learn to use, coming back to the beginning of the conversation, to use AIs more as a tool. We're going to see that um, the the greatest um, improvements, uh, whether economic or and within companies or like within our lives comes from the interaction of AIs and humans. And so like I I think there's um, there's a very famous futurist slash AI um, technologist, Ray Kurzweil, who he's wrote a bunch of books and um, he I think he works for Google now. Um, and he really is bullish on us like kind of combining with AI, like where we're going to like use the breakthroughs that um, we're eventually able to come to interacting with AI to figure out how to, um, integrate it with our brain. To so to the ways that we're like already using technology as an extension of our mind. Like I think this is true for like the phone, right? Like we're where we think of it as a separate thing because we're interacting with that information through our senses rather than like it being plugged into our brain. But like I think a lot of us, maybe negatively sometimes, like have this intuition that, um, uh, when our phone's not there, you feel like somebody took a part of your brain from you. And I think that's like a real intuition because you've outsourced parts of your brain to that phone. And I think that that's we're going to keep on doing that, hopefully in a better way than we've done it with the phone because as we discussed last time, like there's just, I mean, the phone's like, I think hijacked our brain because we don't think, uh, integrate it in a responsible way. Um, but I think we're going to figure out how to integrate technology in a responsible way. I think we're going to figure out how to integrate it um, more directly, right? You have these companies like Elon Musk's Neurolink that's already experimenting with how to create um, direct interfaces between the brain and technology. Um, and yeah, so I think I think we're gonna, we will create super intelligence and we will create kind of new forms of life even if you know, I if you'll permit me to be so like kind of forward thinking and future thinking. Um, but I really think it's more likely that it will be some sort of combination rather than a replacement.

57:02

Thomas: Yeah. That's like it's starting to get a little bit wacky, right? Because we're thinking of course, sort of integrating um, um, our minds with AI. And I think a lot of I don't know if in my mind this is getting wacky, right? Like I we think of kind of all this uh,

57:21

Mike: I think the world is about 70s, 80s to get wacky. Like I actually, this is where I like, I think we're in for a wild like 20 years.

57:31

Thomas: But but we feel like I think of like these films from like the 80s, 90s that explored uh, themes of um, science fiction and um, and then you you just mentioned a more like a subtler way in which, um, you know, iPhones, like telephones have taken over our our own brains, right? Like we now delegate just so much of of of I mean, silly example, like this is a silly anecdote. I, um, I usually go to the gym with no phone, right? Like it's my way to detox and so I just do my exercise there because like I am completely addicted to my phone as you know. And um, and I like last year, um, I had an accident and I like and ambulance like I almost died. And um, and ambulance had to come pick me up. It was blood all over the place.

58:21

Mike: You told me about this.

58:22

Thomas: And for like 12 hours, I didn't have a phone with me, so and I didn't remember anyone's telephone number, right? I still remember the like the different telephones for like my grandmother's uh, place and my my aunt's place and my friend's thing, but I don't remember any like mobile phone number from my my current contact. So for 12 hours, I couldn't communicate, like they're that I had plans to meet with friends. They were worried. They they came to my place. They were like, they they talked to the neighbors. Nobody knew what was happening. I had a thousand phone calls on my phone, but I couldn't communicate with them because I had delegated that part of my brain to

59:00

Mike: we've all done that. I think with like Google Maps, right? Like nobody can figure out how to get places anymore without Google. Exactly.

59:05

Thomas: I don't know how to drive without Google Maps. And so yeah, so we think about this this wacky ways in which we're going to integrate with uh, with AI, but it is true that I mean, I don't know if this was your part of your point, but there is a point in thinking that the way in which we're going to integrate is going to be the subtle in those ways. Or is it going to be totally wacky? Like are we going to have implants in our brains and then like those are like other implants are going to be able to make autonomous decisions for ourselves. Like how is this going to look like? Like what are you picturing?

59:34

Mike: Um, I I so in preparing for this, I did, I reread some of um, Ray Kurzweil who I was just mentioning, um, his stuff. And I, like he's very quick to remind his readers that like he's been predicting this stuff for a long time. And there is a lot of, I think sometimes he overstates that, but there is a lot of consistency in in um, in his predictions. Um, and he's been, his track record is pretty good. Um, and so he thinks that eventually, and he thinks eventually is like within the next 10 to 20 years, that we're going to figure out ways to non-invasively, um, integrate, so not invasive in that like it will involve nanobots that like kind of go through the, it sounds crazy, but like there's again, there's nothing, yeah. It's just an engineering problem really. Like there's there's nothing theoretically precluding this technology. Um, so nanobots will go in through your bloodstream, so you know, they're injected or something. Um, and then they go to the capillaries like in your in like the outer layers of your cortex and then they can communicate with like some with the cloud in some way and that yes, you'll be able to essentially build more layers of cortex. And if you think like the way that the cortex is um, structured, uh, the lower layers have like kind of simpler symbols and thoughts and and were it simpler concepts. Um, and the higher layers uh, are more kind of um, like big picture, uh, um, what's the word? Um, abstract thoughts. So you'll be adding more abstract layers that can make kind of more disparate connections. Um, you'll you'll extend your memory, etcetera. Um, with with this technology and yeah, I mean, it will, I mean he makes the point that and I don't know if I really think that this has been like fully sussed out that like the biggest difference between like the um, our closest relatives like apes and us is the these like additional layers of the cortex. And so like with those, um, we're going to kind of become beings that are as, you know, superior in our ability to think and you know, see connections, etcetera, um, uh, um, versus current humans now as we are from apes and even more so because then there's no like we we are um, our brain size is limited by like actually what um, what head size can fit. And it's one of the reasons that humans are born so um, like helpless and early in their gestation process. Um, and so like suddenly your the brain size isn't limited so you can just make it kind of artificially or like synthetically, digitally as big as you want. Um, and yeah, like the then who the hell knows what that unlocks.

01:02:58

Thomas: Okay, yeah. So as promised, it's pretty wacky. But um, perhaps I want to ground this a little bit more because um, I want to understand your your current experience. And I I love it when you when we talk about the future, but um, um perhaps as some last question, um I want to ask you about how AI, specifically, has changed the way you think, the way you behave, the way, what are some of the ways in which this is already happening? And and in which you're extending your mind into this tool?

01:03:37

Mike: Absolutely. So I think um related to this conversation, I'm going to give you one good and one bad. I'm going to start with the bad. Um, so I, I think that there's a danger that we outsource too much, right? It is so easy to use AI, use LLMs in the way that you were suggesting up front, right? Which is like do this for me for me, generate this thing for me. Um, and I've done that and I'm still tempted to do that, but like I I guess I'm like so scared and on a on like a personal level, like I don't want my creative skills, I think mostly about writing here, but I it's it's other skills as well, I guess, to atrophy, right? Um, and I I find that I need to be like very purposeful in um, in ensuring that that doesn't happen. And I I think that also makes the quality of my kind of like output and writing better. So I've actually started to, I will force myself, I don't do this every time, but I try to, um, even when I'm going to use like I I always iterate when I'm writing, I always iterate with AI. And usually like the first version I get back is kind of shit and I find that I have to do like a lot of the writing myself before it's can really help me get to like uh, before it's actually being an edit like it only works when it's editing things that I've already produced. If I just give it like amorphous ideas and ask it to give me back output that I actually feel comfortable sharing as my own, what it gives me is like not good enough. But eventually like I kind of iterate and I get to a point where I'm relatively happy, I still force myself to retype it. Like I'm literally just reading the AI output instead of copying and pasting, I'm retyping it because that forces me to like really think through each word and be like, is this my word? And I think that that is very important. And so like, um, so one way that it's, it's sort of it changed my behavior in that like I did start outsourcing a lot to it and now I'm trying to counteract that. Um, the way in which I use it for the better is there are a lot of things that I just don't know how to do or um I just wouldn't do, especially if it's like, you know, we've played around with music recently. Um, figuring out uh like a new software. Um, if I had to do that myself, like it's the type of thing that I, you know, without a tutor, without like a class, without um, a lot of time to invest in like learning um, how to produce music. Like I just I I wouldn't do it. Like I don't have like it's a it's a hobby. It's not like something that I'm going to like do for life. And and now, now I can, I mean, I'm not saying that like I can produce music really well, but like I can actually sit and like take screenshots and be like, I want to do this and like send it to ChatGPT and like it will tell me what to do. And so and I love that because like I'm able to acquire new skills without paying a tutor, without taking as much time that I couldn't before.

01:07:05

Thomas: Can I tell you which um a way in which um AI's already changed my the way I think it's made me so like first of all, more confident um in in in asking questions that I would have like before I would have thought um thought they were just dumb questions and also because I'm I'm saying this, I use this feature um AI Google Studio has this um feature, some of you might know it. Um it allows you to uh record your screen and then you can talk to it and the and then Google AI Studio, like Gemini's like latest pro model will see everything that's happening, right? So you need to take in a lot of information at the same time, but it actually, anyway, so I just ask it like, um, hey, I'm trying to um, like use this specific piece of software. I I write a lot of music and so I recorded my voice and my voice sounds like slightly muffled, etcetera. And then it just, if I if my question is too open or like I haven't given it like enough context on what is it that I want to do, it will just kind of tell me, sorry, excuse me, like what are you like you're trying to um, compress it? It's like, I don't know if what I'm trying to do is compress it necessarily, like I want it to sound like a specific way. And so because I interact with it, I need to be way more precise in the way I I make requests.

01:08:35

Mike: I love that. Yeah, yeah, yeah, yeah.

01:08:36

Thomas: And it's just already helping me. I can see also when when interacting with humans, I'm way more precise in the way I communicate what I need, right? It's not this sort of this fuzzy, make it sound better, right?

01:08:49

Mike: Yes.

01:08:50

Thomas: Um it's actually like because you need to be very precise with the way in which you interact with AI with AI in order for for you to be able to get the best outcomes. It's just helped me structure, it's already helped me structure my thoughts and it's helped me, it has an impact in the way I communicate with other humans, not just with AI.

01:09:07

Mike: I never thought of that, but I I think I see that too. And I mean it makes sense like these are machines, like you have to be more precise with machines to like get the output you want. I mean that's the whole thing like programming languages are like very you can't be imprecise at all. Um, yeah, it makes makes a lot of sense. I love that.

01:09:27

Thomas: Yeah, so that's that's a way in which um, it's already sort of changing, like it's I can see um, a positive effect. Now, are there negative effects? 100%. I I've delegated uh, I mean and a way that like you um, the way the same way that you said that you need to check each word that you're writing, etcetera. I've I've last year, um, I was working for this German Fintech and after like a three-month round, uh, my manager there, one of the the piece of it has stayed with me and has stayed with me ever since, like, he very explicitly said in the politest, nicest possible way that I rely on AI too much to

01:10:13

Thomas: that you do. Yeah, to like generate ideas, right? So I'm thinking like we have a problem, like we say, I we need a an outreach plan and like

01:10:23

Mike: ChatGPT, please do an outreach plan. Exactly,

01:10:25

Thomas: like I just give it like a little of context, etcetera. And it's just forced me like but why would he why why did he tell you why he didn't want you to do that? No, not necessarily because the ideas were bad. I think like there's a there's a sense of appreciation for the craft. Like if you're good like there's going to be a bit of resistance and um, there's it almost undervalues the the work you do if like if you know that everything you like you've like the output of your plans, etcetera is just done with AI and then sort of it's like like slight revisions. Do I agree with it? I'm still sort of grappling with the idea that um it does like and it is getting better to like if you give it enough context and guide it like it it is, but I still feel like you have to inject so much of your. I mean there's a sense of like you're your brain atrophying, like in a sense that will I like to use if I go down this road where I'm just delegating thoughts into the, will I be able to tell if the output is good?

01:11:31

Mike: But there's I've seen there's been like a few memes. It's like, you know, like me or like us humans like 10 years from now and like like AI's talking and like it's just like somebody acting like a monkey and like banging on the computer. Yeah, yeah, yeah. Um, no, I like I I actually like I worry about that and I worry about like I I think it's a very interesting challenge. And so like, right, like everything is is a double-edged sword, right? Um, there we were just talking about like how easy it is to learn new things with with AI. Um, and I think that like, I I really think that kids are, there's a version of this world where like every kid has the best teacher, the best tutor, the and not just in like math, but like in whatever the hell it is that they want to learn. You know, the thing that and and we'll also like increasingly we can increasingly move to a world where the thing you want to learn, like that is what you should focus on because like the the basics sort of material things that, you know, like that's going to be sort of taken care of um by machines and we're gonna have to figure out how to like distribute wealth etcetera, but we will. Um, and people get to like focus on learning their passions and like um, uh, again with like sort of the best instruction, no matter how esoteric that kind of uh passion is. Um, the negative version is um, is we atrophy, is like we if we keep the education system the same as it is now and like we just like allow students to use AI or try to stop them but they use it anyway. Like I do think kids will grow up like a bit atrophied intellectually. Um, uh, because you will they'll they'll, I mean I was writing essays because I was forced to and I don't even think that was necessarily like the best way to learn. Um, but like it forced my brain to like figure out how to grapple with ideas, to absorb things, to like recombine them, to like make my own arguments. Um, and you can absolutely outsource that entirely. And so like we but you know, I think this is it's a solvable problem and it's a choice. And I think unfortunately, we've like to date, we've used these technologies in detrimental and negative ways, especially for children. Um, but I don't think they're, that's coming back to the beginning, I think they're tools and I think they can be actually like amazing tools for even for the future generations. Um, but we have to like people have to start building products and building systems that incentivize um, using these tools in the right way and not for not in ways that hook our base instincts, that allow our brains to atrophy and um, uh, allow us to like, you know, be subjugated to the algorithm.

01:14:51

Thomas: Yeah. And it's funny like, um, you've mentioned how we might be headed um into a future where all the like the basics, food and accommodation and whatever are provided by by the machines and then we are sort of on the that top layer, right? Which is something that um, I wanted to explore and uh, we'll we'll surely explore it on the next chapter.

01:15:21

Mike: Yeah, yeah, yeah, yeah.

01:15:22

Thomas: Um, we we wanted to explore, um, how economies might be managed, but also in because many people actually might not be aware that economics originally is it's it's belonged to, it still belongs to the realm of philosophy. Yes. And uh, intermingled with philosophy, there's also the idea of morality. There's a lot of moral judgment that goes into deciding how to manage an economy, right? So in the next chapter, um, we probably going to be exploring, um, the the like when economies are being managed by uh increasingly as it's as it's already been happening.

01:16:03

Mike: Or not or not managed or yeah.

01:16:04

Thomas: Or just sort of um, you know, um,

01:16:08

Mike: I think there's a well, I'll talk about it later.

01:16:10

Thomas: We'll talk about it.

01:16:11

Mike: I don't think it I the point I wanted to make is that I don't think an AI being like an integral part of the economy means central planning. I think you can have decentralized economic systems like more like a free market that also rely on multiple AIs that are acting in a decentralized fashion.

01:16:29

Thomas: but yeah.

01:16:30

Mike: Yeah.

01:16:31

Thomas: Well, we'll explore that. All right, thank you very much, Mike. And uh, catch you on the next one.

01:16:35

Mike: Yeah, this was a great conversation. Yeah. Yeah.