Kevin Kelly on Multiple Futures, AI, and Becoming Better Humans

Until recently, there were always other varieties of humanoids on this planet. So maybe there's kind of like, we're just returning to the time when we had multiple sentient beings. I call these AIs artificial aliens, and that's fine, but that's not the same thing as kind of allowing a forking in your own species. The danger there is, you know, the most effective argument we have against racism. is that there is no difference between us. But what if there really was? What do we do with that? There might be a very good reason to not permit speciation. If you care about the long-term human species, for every woman who doesn't have any children, another woman has to have 4 to just keep the current level. Maybe if you do calculations and you figure that every baby born is worth $1 million to the economy over its lifetime, That it's worth to pay a million dollars. So maybe at the right price, it would begin to work. So we have been inventing ourselves and we will continue to change ourselves. And so we will, we want to become better humans, but is there multiple futures for us?

This week, our guest is Kevin Kelly, the founding executive editor of Wired Magazine, technologist, and author. He just released a book called Excellent Advice for Living late spring that jumpstarted this wide-ranging and thoughtful discussion. We go into Kevin's predictions for AI, interspeciation, and how humans need to be having more children. We also talk about practical and unpretentious advice for succeeding, communicating, and being good members of society. Without further ado, here's Kevin Kelly. Kevin, welcome to the podcast. Excited to have you on.

Oh, it's always a pleasure to chat with you, and I'm really delighted to be invited back. Thank you.

We're excited to dive into your new book, which is an advice book, different kind of book. You read old advice books that are written a long time ago, and a lot of those principles still make sense. Are there certain genres of advice that you're inspired by or see yourself in the lineage of? How do you think about advice as a category?

Advice as a category, like I've never really been that interested in advice columns. A lot of advice is wrapped very, very smartly, I say, in stories. We communicate best in stories. And most advice books will have stories about these things. And that is a really great way to convey it because you're kind of showing it rather than telling it. And then you will remember it. I'm not a very good storyteller, so I decided to do it my way, which is these telegraphic aphorisms, these adages, these lessons, these little bits, these proverbs, which suits me and is how I like to consume the advice. I like to collect little quotes and things, and they work for me as reminders. So that's what I produced because I'm no good in making a story, but I can make a little bit of viral tweeting. And so that's what I did.

Let's go through some of them. You mentioned gratitude. You have one, gratitude will unlock all other virtues and is something you can get better at. Talk about how it unlocks other virtues and talk about how you get better at it.

I think in a certain way, in a weird way, I think gratitude and Trust and empathy are connected. They're almost maybe different faces of the same virtue in some way where gratitude, you kind of are acknowledging your unspecialness, your luck. You're acknowledging that you did not earn it. In a certain sense, that it was a gift. And trust is in some ways in my own mind connected in that same way, or gratitude is a form of trust, maybe it's saying, or it's a form of empathy where you're kind of being able to put yourself in someone else's position. So in a way that I haven't really articulated, I think those 3 qualities are very much joined in a way. And they're kind of all expressions of something similarly deep, this kind of connection that we have, not just in a superficial way between people, but that we are actually of the same life. I mean, the life that we share is the same exact life. We have a common ancestor. No matter who we are. And so I think that's what I'm trying to get at is that being thankful for what it is for our lives, for what we have, which are given to us, is very similar to the kind of trust that we would have another person that they're going to do well, which is kind of the same kind of sense of us Having empathy for others, saying, I'm connected to you, we are in this together, there is something in common that we have. And when you work on gratitude, it's kind of a way to work on those other ones at the same time.

There's this old Maya Angelou quote of, people will forget what you say, but they'll remember how you made them feel. And I used to think that was kind of like a bug, like, hey, isn't substance really important? But people operate that way. And you have a Aphorism here, when choosing between being right and being kind, choose being kind. And it just kind of speaks to what people really care about is, are you with them? Are you part of the tribe? Are you aligned with them? And arguments are abstract. They're kind of changing. They're evolving. I'm taking a little bit of an abstract kind of meta commentary on why that might be true, why people care more about how you made them feel than what you actually say. Or why they care more about being kind than being right.

It's sort of like, why do we remember smells so much? I think because there is a way in which they're kind of, they're wired further down into our brainstem. The experience is fundamental. So it's not like there's, the emotional component is sort of like not an overlay, it's more like an underlay. Yeah, the way that we have these associations. that color everything. So we can have trigger responses that are way below or way faster than our intellectual intelligence. I think we tend to overrate intelligence. There was a whole bunch of studies done about people making decisions and then how it reflected or didn't their gut responses where they would make a decision without thinking. Like on first impressions. And then they would correspond that with whether they were correct or not. And also particularly long-term things about character evaluations and their first impression were usually the most accurate impression before they even applied logic. It was faster than logic. So I think there's something going on about that, the way they were constructed, where the The primeval circuits of first impressions and responses and feeling, they kind of work faster. They work stronger than the rational thing. And a lot of what our humanity is about is trying to not just overcome that, but I would say manage it or steer it or at times elevate it because a lot of time it's is absolutely right. And lots of times it's not the best thing. I mean, we have anger and other emotions that also kind of have the same kind of power. So yeah, so for whatever reason, we're governed by these things that seem to trump what the logic says. And that's another piece of advice about arguing. You can't rationally argue something. somebody out of an opinion that they didn't rationally get into through rationality. And so when you're having disagreements, the emotional component is incredibly important in changing someone's mind or convincing them or persuading them or whatever.

Yeah, you can't reason someone out of a notion they didn't reason into. I was gonna bring that up. It's fascinating. And what you're usually arguing about is not actually the thing that you're arguing about. There's often a subtext. I'm curious, is your mental model of emotions something along the lines of almost feedback loops of what were more likely to help us thrive within our tribes 1,000 years ago or whenever it was? Some of those sort of emotions are still adapted to modern environments. Some of them are no longer adapted. But is that fundamentally what they were doing was kind of intuitions about what was more likely to raise our reputation among our tribe?

Yeah, I think so. If you look at, there was some anthropological work studying the last remaining intact tribal people. And it was really some of those who went very deep and lived with them for a long time and got to know them very well and were able to observe them for long periods of time in all their dimensions. It was really, really fascinating. There was, I think, Colin Turnbull, one of the famous anthropologists. Lived with, at the time they were called pygmy people. I don't know the proper current proper name for them. And what was interesting was they lived in these grass leafed huts, basically had no walls. So everybody heard everything in the little clan, in the settlement. They would have these, like any humans, they would have these Arguments and fights over all kinds of, you know, things that people argue and fight about, but they're really there. But, but the, um, and everybody could hear everything and they could hear the whole, both sides, the whole, the whole drama of it. But the, the interesting thing was while, while they could be very, very vocal and they almost get to the point where they're like fistfights and almost hurting each other, but they would never go all that way of really injuring one another. It was a lot of releasing. It was kind of like therapeutic fighting in a certain sense. But they always had certain lines. And the lines was they would never do anything that would jeopardize the health of the group. Okay. And so there's something sociobiological in us as a social animal that, as primates, that even though we have individuality, we've evolved to also pay close attention to the health of the group and the status of the group. And, and so maybe that's what makes us kind of unique is that we're social animals as well as individuals.

And we have these competing instincts. You have an aphorism here. A great way to understand yourself is to seriously reflect on everything you find irritating in others. Why don't you unpack that one?

Yeah, it has to do with some of the other ones, particularly these days with our new invention of AI. Have to do with the fact that we, we humans and individually are very, very opaque to ourselves, that we collectively as a species, we really don't know what makes us human, how our minds work. That's part of the thrill and excitement and the worry about what's going on with AI is that we're trying to make something, replicate something that we don't even know what it is, our intelligence. But there's also individually, this was one of the lessons of The Quantified Self was that we just don't have a very good intellectual understanding of how our own minds work, how we work, where we're coming from, why we even make decisions that we make. And looking at one of the ways we can kind of look into ourselves is find out what kind of agitates us and what kind of where we care and where we're paying attention to ourselves. Again, we're kind of, that's another bit like pay attention to where you pay attention to. And so our consciousness is a very gossamer thing. It's a very slippery thing. It's kind of like the only tool we have to try and probe us, but it's not that dependable itself. And so This is just another way of paying attention to what irks us, is another way to kind of help us to get a view of us. It's not necessarily saying you are like that. It's just saying this is a signal. This is another data point that you can use to try and dissect yourself because we are really hard to self-dissect.

Competing instincts, competing emotions. People get anxious about that, but that's kind of the You know, the default state. You have one, whenever you can't decide which path to take, pick the one that produces change. And I'm curious how you balance that with sort of like commitments and sticking, whether it's a partner or whether it's a long-term job or—

Well, yes, I think there are, you know, I'd said kind of elsewhere that, you know, except for like marriage, always, it's always think about the exit first in some ways. And Whenever I do any kind of business deal, it's exactly where I start with is, okay, what does the exit of this look like so we can structure the beginning of it? Yeah. So I think there are some exceptions where you have a commitment and that's sort of the nature of the commitment is like, yeah, despite all the opportunities, despite all the other choices that I'm not going to make that choice. So, so I think there is, I think there is bounded in that, in that sense, but outside of that boundary, there's still many choices that we have where we have, you know, in terms of occupation, where we live.

Hey, we'll continue our interview in a moment after a word from our sponsors. Secureframe is the leading all-in-one platform for security and privacy compliance. Get SOC 2 audit ready in weeks, not months. I believe in Secureframe so much that I invested in it and I recommend it to all my portfolio companies. Sign up for a free demo and mention Upstream during your demo to get 20% off your first year. Now more than ever, startup founders need a safe place to put their cash. Mercury protects your money and also provides the streamlined user experience that great founders expect. Through partner banks and their sweep networks, Mercury offers up to $5 million in FDIC insurance, which is 20 times the per-bank limit. They also make it easy to invest any cash above the FDIC-insured amount in a money market fund. 100,000 startups trust Mercury with their finances. I've been a happy Mercury customer and have found their team incredibly helpful and responsive. They even got an important wire out of purgatory on Christmas Eve. After all, your Christmas is my opportunity. Visit mercury.com to get started. Mercury is a financial technology company, not a bank. Banking services provided by Choice Financial Group and Evolve Bank and Trust, members FDIC. Make the case for why people should consider having more kids than they think?

Oh my gosh. Well, first of all, there is no overpopulation on this planet. There are certainly places that have seemingly too many people to live, but there are plenty of places where there's nobody living. So there's— so not only is there not an overpopulation, there's actually going to be quite the reverse in the next 50 years and beyond, where there are simply going to be under Population, a population collapse. So, so I just want to, so that should be removed. That, that's not, should not be anybody's reason for not having kids. And maybe even a reason if you care about the long-term human species, because, um, for every woman who doesn't have any children, another woman has to have 4 to, to just keep the current level or whatever level it is, replacement level. But beyond that, I think I am trying to recall, but I don't think I've ever met anybody who had regretted having more children. I would say in modern times or in kind of the developed world. Certainly there were people who didn't have birth control and that might not have been true, but I would say in the modern developed world, I have not met anybody there. And that's also been my own experience. And so I would say that there is a compounding joy from it. So the people who benefit from it is the kids themselves, where we did have our son keep requesting that my wife lay another brother, lay him a brother. Because he really wanted to have a brother. And I grew up in my group with 5. And so we know the joy of that. And that was, I think, that to me is kind of like one of the most important ones.

How do you think we address the population collapse? Because it feels like it's already priced in. You can't just start having, you know, 18-year-olds. Is it that robots and AI just kind of solve for some of the labor stuff? Or how do you think that plays out?

I don't know how it plays out. It's really perplexing to me. And I do know from all these solutions in terms of having human replacements to raise birth rates has not worked, including just directly paying people. Now, it may be that we just They haven't been offered enough sums that maybe if you do calculations and you figure that every baby born is worth $1 million to the economy over its lifetime, that it's worth to pay $1 million. So maybe at the right price, it would begin to work. So far, I don't know what that is. And if there isn't a way to do that, the question is, well, what happens to the economy? And there you could imagine AIs becoming in some ways the audience and the market in the way that we build houses for our cars and call garages. Maybe we make entertainment for the AIs in some ways. Maybe we have to build houses for our robots and pools. I don't know. So So that's possible. That's one possible way out. But it is something that I don't have very many good ideas about.

Let's flesh out more of the AI-human relationship because I've heard you're famously an optimist and you're excited about where things will take us. And I'm curious because some people will say, hey, they're optimists, but they also believe that humans one day will be subservient to the AI in the same way that Neanderthals were kind of the previous form of evolution. We're not the last form of evolution and there will be something after us that might be some sort of hybrid or it might be something else entirely. And that's just how life goes. And so, they don't worry too much about that because that's just kind of the inevitability of it. Whereas other people think that that's a terrible outcome. I'm curious where you net out on that.

Well, yeah, I mean, maybe in between. So, First, I would concede or even postulate that our humanity is malleable. It was something we've invented and we're not done yet. So we have been inventing ourselves and we will continue to change ourselves. And so we want to become better humans, but we don't know what that is. And the question, That we also don't know is, is there multiple futures for us? Okay. Can we become different kinds of humans? And that prospect is really very problematic in many ways. We certainly could imagine there'd be varieties of us that are not going to get anything changed in the naturals. They're simply neither going to do any genetic engineering and not permit very much alteration in their bodies. And that's like damage, right? So, and then we can easily imagine people who are going to be very quick to, they want to eradicate the gene for Alzheimer's and Parkinson's from their germline and all future descendants tomorrow. You know, then maybe we could see a forking, at least one fork, maybe more going on. And I don't know what to think about that. Is that surrendering? Is it kind of like the union of the United States? You can't succeed. There can be no divisions. We must remain one. So I can easily imagine a bunch of people believing that, that the worst fate would be to have speciation among humans. And others who would say, no, this is actually natural until recently, There were always other varieties of humanoids on this planet until probably we got rid of them, whether we don't know. And so that was, so maybe there's kind of like, we're just returning to the time when we had multiple sentient beings, which is still a little different. I mean, I call these AIs artificial aliens and that's fine, but that's not the same thing as kind of allowing a forking. In your own species. The danger there is, you know, the most effective argument we have against racism is that there is no difference between us. But what if there really was? What if we're the real— and that's like, what do we do with that? And so there might be a very good reason to not permit speciation.

And There's a few ways speciation can happen, right? There's the genetic engineering example, there's the putting chips in people's brains example, there's maybe the ChatGPT-50 or whatever just gets so advanced that it— which speciation versions do you feel are most realistic or do you think all of them will happen to some degree?

The cyborgian thing is going to be very, very slow. You know, I just recently went to see the Neuralink stuff where they have the implant, about the size of a quarter, put in the back of your head and allows a monkey to control the computer with their mind. And they're coming close to actually doing human testing on this for quadriplegics to help them walk. And that was a lot closer than I thought. So that's how it starts. It starts with a kind of a medical therapeutic way to help people. But that's, I'm not sure, I mean, that will speed up the evolution in certain directions, but that's a very, very slow process. Whereas the genetic germline thing can happen, you know, at the rate of human generations. So I would say in terms of like literally speciating, And having different varieties of humans. I think that, you know, the genetic side will be really the one to pay attention to. And it can be informed in terms of what we discover from AI and other things in terms of how to do it. But, you know, that's still a slow process, you know, 25 years of life for a human generation to to turnover. It'll take a long time before they're literally like you can't breed kind of speciation too. So that's kind of a more technical biological thing is that you can't breed. And I think that's, I don't know if I, you know, whatever we consider, we probably can genetically engineer some way to breed these hybrids. So speciation means only that more of an identity, I guess.

The Eliezer Yudkowsky argument, as far as I understand it, is something along the lines of, The idea that it's not that AI needs to be conscious in order to get rid of us. It needs to be indifferent to us, which it is, and then just understand that our atoms are something that they could use for anything else, to advance its own goals. And then at some point, it will become in its own interest to eliminate us. Maybe I'm botching his argument. You seem to be not as dismayed by either his argument or the kind of existential concerns that some of the AI safety community has. How would you comment on that?

The argument of the many arguments out there about existential risk, the only one that sort of makes sense to me as a thing to worry about is that basically that we over time allow and engineer AI to do more for us. And at some point we give over more and more of, we voluntarily engineer more and more control to the AIs on purpose to run things. And that at that point of kind of giving them power, then we're at their mercy should it awaken to some kind of sense of its survival being threatened. That's the closest I can understand it. I think it's a fantasy in many, many dimensions. And one of them is that that's not going to happen fast. So the one fantasy is it happens fast and we can't back out of it. 2, the fantasy is that there's a single AI or incredible collaboration among Several really big AIs, and there's no evidence at all. The evidence is going to be that there's going to be millions, hundreds of different AIs, just like we have different engines and tools and machines. We don't have one big machine. And then the third one, I think the most serious fantasy part is this complete overestimation of the role of intelligence in achieving things, accomplishing. There's this idea that whatever is smartest will dominate. But we know if you put a human and a tiger in a cage, we know which one's going to live. It's not the smartest one. There's just so many other things that are necessary to accomplish things in the real world. Being the smartest person in the room does not necessarily mean that you are the dominant person. No matter how smart you are, you have other things to, to, you need other things to get done, including access to things and cooperation and collaboration with many people doing many different things or many AIs doing many things. And that middle-aged guys who like to think a lot really, really put a high emphasis on thinking. And they think that if they could think faster and better than They would be running the world. And they can think of all the ways in which they could ingeniously figure out how they could take over. But it's a fantasy because that's not how the world works. That's not how reality works. You have all these things and they break down and they don't work on the first attempt. And by the way, the instinct for survival usually will dominate the The attempt to kill. Survival is a more powerful motivator than trying to get rid of somebody. I think it's a fantasy, kind of like Superman. It's beautiful. It's mythic. It makes comic book sense.

But why isn't the Neanderthals to humans kind of like evolution? Why isn't that a good analogy for what could happen or will happen?

Well, yes. Okay. I think it is an analogy. And so that took, I don't know, 10,000 years. Yes. If you would say that over time, that they could replace us over time, I could buy that. But to say that, you know, all at once overnight, overnight, that's just comic book fantasy.

Right. And we do see, you know, in time, there are jumps in, you know, sort of economic growth, for example. So things aren't always as linear, but even still, you know, 10,000 years is a long time. And so to expect this to happen in 20 years—

And the analogy with Neanderthals, by the way, is they're much more likely to have been bred away rather than murdered. The point is that there was a willing merging going on. It wasn't that they were killing us off. There was— and so in that sense of like, yeah, we may merge with them, but only if we decided it was in our interest to do so. And that's one of the scenarios for Neanderthals, joined us rather than we murdered them. And it took a long time. And if that's the scenario that we're talking about, okay.

You wrote a book a long time ago called What Technology Wants, and you've written a lot about the topic since then. And obviously you co-founded Wired. You've been thinking about technology for many decades, both past and future. Do you feel it's kind of a precursor in some ways to Marc Andreessen's software is eating the world thesis? Or how would you kind of contextualize your mental model of technology and what it wants, where it's going?

You mean, does it completely take over? Do we become kind of like completely techno-monolithic, techno-planet? Is that what you mean by takeover?

We become the tools. Our tools shape us literally.

Well, yes. So in a certain sense, we are becoming, as I have long said, we've invented our humanity. We invented ourselves. We are an invention and we will continue to do that. But there are many, many attributes that we have of living in these kind of wet cell biological self-reproducing things that, you know, it's as we're making these robots, We've come to understand that the kind of power density of a human, we're a quarter horsepower and 60-watt brain. It's going to be a long time before any kind of a robot will be able to operate at those low energy levels. So yes, so we may continue to invent ourselves, but we may not necessarily diverge very much from this miraculous machine that we have. I think we're likely to populate our surrounding environment with all these artificial alien beings of all different varieties and stripes. But beyond that, in terms of the AI and just the general drift of the techium, which is kind of to make— it's basically is to make organisms and forms and kinds of being that were possible, but not possible to be made with living tissue. On the long trend, you know, there's all these forms that could be, that could be doing things, but you're not going to get there if you have to make enough cells that are mostly made out of water. You can get to those if you can make them out of other elements, and you can only get to there through a mind. So we have first a human mind, and then we're going to make all these other AI minds that will be inventing new ways to make forms, new ways to make a living, new ways to exist. exist to be that are technological and we can fill up that. I don't think it's going to necessarily replace biological forms because that's generally not what we see. We see evolution much more additive. There are things that go extinct. However, in technology, we don't see extinction. And that's one of the differences is because they're idea-based and we can carry ideas forward. So we have shifted the evolutionary arc because now we don't have to have as much extinction. And so we can imagine going forward where we retain as many of the biological species as possible while adding on additional technological species. And we make more and more. So we have a world, like a planet full of all the biological species that we have today, and they continue. And millions of more technological species. Um, and so that's what I think the general pattern is, is yes, it's a world filled with all kinds of technologies that we don't have today, but not at the cost of the biological species.

And to make this a bit more concrete, you said a few times, you know, we invented our humanity. Unpack exactly what that means in terms of like, what were the biggest inventions or just flesh it, make that a bit more concrete for us.

One of the, our biggest inventions is the invention of language, which we did invent. I mean, it was meaning that I think that primitive, you know, primates trying to communicate, try things in their mind, try to do something and they worked. Those who were able to do that survived and try it again to make something work. And that kind of, that battery of things and techniques that they discovered were passed on and became the basis of our language. And that language, what language really gives us is, there's 2 things. One is it's communication between members, which is very powerful in how we can hunt better so we could, clan could survive. But there's something else really important about language, which is that it gave us access to our own thinking. So the only way we know what we're thinking is through language. So the language was a two-pronged thing that gave us access to try and figure out what we are. So it was our origin of our consciousness, basically. That was very, very powerful because it kind of then could give us to be purposes and directions and intention that we didn't have before. So that was something we invented. And then from then we invented things like domestication of herding animals, which we could milk. And once we figured out that, the human body rapidly evolved in certain populations, adult tolerance of lactose. That happened within, I don't know, 7,000 years or something. It was really fast. We invented cooking and fire, which was an external stomach that could digest stuff we could not with our primate stomachs. We could then access nutrition that we couldn't get to, which changed our teeth and jaws very, very quickly. So how we look right now is something that we invented. And so we're continuing to do that right now. Actually, biological evolution has not slowed down with cultural evolution. It's actually sped up. So that's what I meant by inventing. And then all the things that we think are important to us, like fairness and the moral judgment and all these other things, we invented. Idea of law and our idea of, even though there's some primitive traits we can see, like fairness among primates, We have invented elevated and much more stricter levels of that. And some, a lot of us communicated through culture, which is something else we invented. But when we think of an elevated enlightened human, that's something that we've invented.

In one of our first conversations almost a decade ago, I remember you said something like, we will need a new mythology or new mythologies that help us kind of make sense of the new world that we're entering.

Yeah. I mean, at the Long Now Foundation, we're building this clock to tick inside the mountain for 10,000 years. And that's, we hope that to be a mythic thing. This idea of a clock ticking in the mountain for eons, for generations. That's the kind of mythology that I think is helpful for us as we Kind of reimagine ourselves.

Yeah. And our new purpose in this world. It's interesting because, well, just on the point of purpose, I mean, some people say, hey, people will do art and poetry and we'll find all these new things once a lot of knowledge work has been automated, but the AI will be better at that kind of stuff too. How do you think about, in a world a decade from now or 2 decades from now, how humans think about purpose in a different way once a lot of things that they were the only, to use your words of don't be the best, be the only, could perhaps be done with AI? Or feel free to dispute the premise.

I just think we know so little about what our own intelligence, our own being is like. And so AI is going to be disruptive and instructive for decades because it's going to help us experiment on us in some ways. I think we're going to learn more about our own brains from AI and making AI than we have from neuroscience or psychology together so far. And I think we're going to be having this discussion about what is it that we are about? Where are we going? What is it for, for the next 50 years, at least, as, as all these things come along and we kind of re-register them and— I think one of the things that we're— one of the things we've done just this year is demoted creativity. Again, this has been talked about. It's like, well, the thing that humans do is creativity. Computers are the opposite of that. And now we know, wrong. Computers can do creativity at a lower minor level very easily. And so now we're saying, well, yeah, creativity's not the this high-order supernatural thing is actually very, very primitive. And so we've kind of changed our ideas about creativity very, very fast. And I think that kind of thing we're going to continue to do, not always necessarily demoting it, but just shifting and maybe having more subtle, nuanced differences. Because I think there's what I call capital or major creativity and then a minor. So we'll have We'll have to devise new language for distinguishing between what this kind of everyday novelty is versus a kind of a breakthrough where you are, where you're trying to do something that's outside the average. And so we probably maybe have 2 new words instead of just the one word creativity.

I'm fascinated by the, you know, we'll learn more about ourselves than we have with neuroscience and psychology. We still have no idea what consciousness is and maybe we'll figure it out in the process. One idea you've been thinking a lot about for a while is this idea of a global government or the need for one. And also this idea of covalence, I think you call it. There's the idea of kind of two-way transparency.

Well, yeah, it's not surveillance, but covalence.

Yeah. And it's interesting because when some people talk about existential risk, I mean, one of the existential risks we've introduced in the the last century has been nuclear weapons. But the competitive dynamic in terms of multiple countries having nuclear weapons seems to have staved that off. And so I wonder if similar to right now, we're having a conversation about AI and centralization and decentralization, and if there should also be a competitive dynamic that might stave that off. And one worry people could have with a global government is, does that reduce the competitive dynamic? Does it centralize power a bit too much? And thus, why would they be covalent, so to speak? Why would they allow 2-way transparency? How do you think about the— do you see a need for decentralization there or do you agree that competitive dynamic is what's preventing perhaps this abuse of power?

Yeah, I would say in general, what we know about systems, again, me taking the whole systems approach, is that There's a tremendous power in the bottom, in a very flat-ish, bottom-up, decentralized, distributed system, which is a large part of evolution, a large part of ecology, a large part of living systems, a large part of the mind, but it's not the only part. And that's the lesson is that most of the systems we see are combinations of lots of decentralization in many aspects and some centralization. Well, we know that the advantages of decentralization and distributed systems is agility, flexibility, adaptability supreme. They're just the best ways to adapt to changing environments, changing circumstances. changing goals. But we also know that there's a cost to that. It's, they're incredibly inefficient. I mean, just by the nature, you're duplicating everything. There's no sense of efficiency whatsoever. There's no mechanism for efficiency. As soon as you introduce efficiency, you begin to centralize. So the question always is, there's trade-offs and these, all these systems being hybrid is you're going to, pay the costs for certain aspects of the decentralized system, pay the cost of inefficiencies and slowness and other stuff because the adaptability or flexibility that it does is so valuable that you're willing to pay the cost. And other aspects of it, it's not worth paying the cost. So you want to have something more centralized. So authentication could be Decentralized, and there's reasons to do that. And there'll be some cases where it's going to be worth the pay, the cost of completely decentralizing it. But in many other cases, it'll make sense to have a more centralized version of it. So there's no free lunch in that way. The way I would say is, you know, the bottom-up is always to take you further than you can go decentralized systems. take you further than you ever thought. And they're usually the best way to start, but they don't take you all the way. And so most of the systems that are kind of highly oiled and well-working will be some combination of mostly decentralized with little bits of top-down centralized control.

This has been a fascinating conversation. I want to close with one more of your book. Which is you have this one aphorism which talks about instead of thinking about can do or can't do, think about I do in terms of internalizing something, identity. This kind of meta concept, which I'll say is a plug for the book, which is there's this basketball player, Giannis Antetokounmpo, who just lost in the first round of the NBA playoffs. And before the series started, if you were to ask him, hey, if you guys lose this series, is it a failure? he would have said, yes, it's a failure. We have to win at all costs and got to have our head in the game. But then afterwards, when someone asked him, is this season a failure because they lost? He said, no, it's a learning experience. And it just talks about the idea of certain proverbs might serve you better in certain times. And maybe life is about adopting the right mindset at the right time while not getting caught up in the contradictions.

Yeah. I think you agree. I'm not sure if it was the same guy, but there was somebody else recently who said, there's no failures in sport.

It's the same guy. Same guy.

Yeah. I mean, that's just very, very briefly, but that's one of the base, one of the newest. Well, occasionally there are new things under the sun and there are one new thing under the sun that I think Silicon Valley can take credit for is demoralizing failure to understand that failure is seen now as a necessary component for science, for innovation, for entrepreneur, for the economy in general. And that what you want to really have are systems that manage failure. Okay? They're failure management systems where you have your failures in small doses and you manage them to prevent the cataclysmic failures that you want to avoid. And so I think there's been a change, complete sea change. where if you lost money or if you had bad grades or if an experiment failed, that was considered a disgrace. And now it is seen, as you say, as a learning experience, as a way we go forward. So the fail forward idea. So yes, fail forward.

That's a great place to end. The book is Excellent Advice for Living: Wisdom I Wish I'd Known Earlier. Kevin, thanks so much for writing this book and for coming on the podcast.

I really appreciate your great questions, Erik. Thank you.

Upstream with Erik Torenberg is a show from Turpentine, the podcast network behind Moment of Zen and Cognitive Revolution. If you like the episode, please leave a review in the Apple Store. Secure Frame is the leading all-in-one platform for security and privacy compliance. Get SOC 2 audit ready in weeks, not months. I believe in Secure Frame so much that I invested in it and And I recommend it to all my portfolio companies. Sign up for a free demo and mention Upstream during your demo to get 20% off your first year. Now more than ever, startup founders need a safe place to put their cash. Mercury protects your money and also provides the streamlined user experience that great founders expect. Through partner banks and their sweep networks, Mercury offers up to $5 million in FDIC insurance, which is 20 times the per-bank limit. They also make it easy to invest any cash above the FDIC-insured amount in a money market fund. 100,000 startups trust Mercury with their finances. I've been a happy Mercury customer and have found their team incredibly helpful and responsive. They even got an important wire out of purgatory on Christmas Eve. After all, your Christmas is my opportunity. Visit mercury.com to get started. Mercury is a financial technology company, not a bank. Banking services provided by Choice Financial Group and Evolve Bank and Trust, members FDIC.