The Exchanges

Q Wow, what an enthusiastic voice. I always think no one has as enthusiastic voice as me, and then, uh, you come on, uh, but that's fantastic to hear. I want to get started today, though, with a discussion on how you made the move from quantum mechanical theory for microscopic black holes to being chief scientist at Founders Fund. What was the transition?

A Uh, certainly. So I spent most of my career building quantum computers and researching tiny things like microscopic black holes and atomic clocks, and at some point decided to transition from that into industry. And the opportunity that availed itself was a company called Halcyon Molecular, where we used quantum physics techniques to study DNA and biological molecules. So we built single atom resolution electron microscopes, which turns out to be very similar to building quantum computers. And that you're manipulating the quantum states of individual atoms, but use that for, for biotechnology purposes. And that company was backed by Founders Fund. Uh, and so that's how I got to know the, the partners and the company didn't work out, but we had a good relationship. And so they said, we are looking at doing more of these types of high technology, high science risk kinds of companies. And we really want to have your expertise on board and Within our team to help evaluate them and cultivate some of those opportunities. So why don't you just join us? And so I did. Sounded like a super fun opportunity.

AI assessment note: “that company was backed by Founders Fund... so they said... why don't you just join us”

Q Wow, what an enthusiastic voice. I always think no one has as enthusiastic voice as me, and then, uh, you come on, uh, but that's fantastic to hear. I want to get started today, though, with a discussion on how you made the move from quantum mechanical theory for microscopic black holes to being chief scientist at Founders Fund. What was the transition?

A Uh, certainly. So I spent most of my career building quantum computers and researching tiny things like microscopic black holes and atomic clocks, and at some point decided to transition from that into industry. And the opportunity that availed itself was a company called Halcyon Molecular, where we used quantum physics techniques to study DNA and biological molecules. So we built single atom resolution electron microscopes, which turns out to be very similar to building quantum computers. And that you're manipulating the quantum states of individual atoms, but use that for, for biotechnology purposes. And that company was backed by Founders Fund. Uh, and so that's how I got to know the, the partners and the company didn't work out, but we had a good relationship. And so they said, we are looking at doing more of these types of high technology, high science risk kinds of companies. And we really want to have your expertise on board and Within our team to help evaluate them and cultivate some of those opportunities. So why don't you just join us? And so I did. Sounded like a super fun opportunity.

AI assessment note: “And that company was backed by Founders Fund. Uh, and so that's how I got to know”

Q with you, uh, both undeniable in scale and magnitude to society, I think we'll agree, is self-driving trucks. So what are your views on the societal implications of, I think it's four million truck drivers being unemployed over the next, we can give a timescale here, of Five to 20 years. We can kind of pick it as we choose, but what are the societal implications of this to you?

A Yeah, I think the societal implications are really huge. I think it's almost impossible to understate how big of a deal it is. If you think of, you know, in the last political cycle, how influential and how disruptive the idea of the out-of-work coal miner was, uh, well, there's only, you know, a few tens of thousands of coal miners in the United States, so it's like a relatively small constituency. Population wise, but had a, a very large effect on the overall political narrative. And so if you think that truck drivers, which are, you know, a thousand times more causing the same type of redundancy and displacement as cold splinters experienced as, as a, you know, natural gas came online, then it is a, is a big, big potential for societal people. The way that I look at it is not so much just You can't just focus on the truck drivers themselves, but really on the whole network. So the trucking system, you know, we have Eisenhower's interstate system and the road network on that's, that's built around that. The whole trucking system does serves two purposes at the same time. One, it's a goods distribution system. So we load things on the trucks, we ship them around. That's how we get all the stuff that we buy at the store, but it's also a wealth distribution system. And that you have wealth being generated at a few points of focus, things like factories, where the goods are creat…

AI assessment note: “Yeah, I think the societal implications are really huge.”

Q And speaking about kind of anomalies in data sets, maybe, and being able to handle them, I'm intrigued whether you share the same concern that many VCs I speak to have with regards to access to data sets for startups. Is this a primary concern for you?

A No, not really. Most of the data sets out there or the defensibility data sets out there I think are oversold. There is this very, it's a very compelling, very tantalizing notion that, oh, I just have so much data, and nobody else has it, and therefore it must be special, but most of the data out there tends to be noise, and it's just a burden to process. You're trying, you just, most of your algorithms are designed just to filter through it and not actually learn anything from it, and so that tends to be not Be nearly as valuable as people think it is. The other thing that where people get trapped up about the specialness of their data sets is that you would like to believe when it, when a data set truly is valuable, it has this, it has this property that each new incremental piece of data is marginally more valuable to the whole thing than the last one that you put in. So you get this kind of accelerating effect, uh, rather than a diminishing returns kind of effect. Most of the data sets out there have this diminishing returns effect. A good example is like genomic databases where people try and say, okay, well, we sequenced 10,000 people instead of a thousand people or a 100,000 people, implying that their data sets are therefore 10 times more valuable because they have 10 times more sequences. But if you don't actually see any new genes or new, you know, Those 10 X more seq…

AI assessment note: “No, not really. Most of the data sets out there or the defensibility data sets”

Q It's interesting because I often hear that kind of AI means big data can finally come into fruition. There you're suggesting it's more of a rebranding. How do you think about that then?

A One, one way of describing artificial intelligence or a particular deep learning, since, uh, at the core of it is just this matrix multiplication operations. It's, you know, you're just sort of doing linear algebra at industrial scale. We haven't gotten to the point for most of these things or most of the things that call themselves AI, where you're really doing something that is fundamentally different. You're just making it bigger or just having larger scale doesn't actually change the nature of the calculations. And I would, I would even sort of push back on when people tout their big data credentials, that most of the things that are sort of big data are actually not that big. You know, you would like to, you'd like to think of these petabytes and exabytes floating around there and coming up with really amazing insights that could never have been seen. But a lot of times it's really in the sort of megabytes and gigabytes kinds of kinds of range. And so it's always beneficial to say that you have this big data edge when really it's just more, more marketing and branding.

AI assessment note: “it's really just more, more marketing and branding.”

Q Genius and crazy. How do you determine between the two? I'm sure, I'm sure you've seen both. How do you kind of look to identify either camp?

A So part of it is thinking it through from first principles, so understanding a lot of the, like, you know, the backgrounds and what has been done before, what's sort of possible. You can, should be able to reconstruct someone's argument for, yes, like, we are able to solve all of the significant challenges here, and like, you know, these things are hard, but there are also a couple of interesting heuristics that we use to sort of As a check on the first principles thinking. One is just how thorough is the knowledge and the understanding of the people that are proposing it. Even if you don't necessarily yourself have the complete picture by just, you know, drilling down on specific aspects, you know, whether that person has, like, has all the answers and can go as deep as you want to go is a really good sign. If it turns out that the way we're thinking about it is very shallow, then that typically reveals itself pretty quickly.

AI assessment note: “how thorough is the knowledge and the understanding of the people that are proposing it”

Q And speaking about kind of anomalies in data sets, maybe, and being able to handle them, I'm intrigued whether you share the same concern that many VCs I speak to have with regards to access to data sets for startups. Is this a primary concern for you?

A No, not really. Most of the data sets out there or the defensibility data sets out there I think are oversold. There is this very, it's a very compelling, very tantalizing notion that, oh, I just have so much data, and nobody else has it, and therefore it must be special, but most of the data out there tends to be noise, and it's just a burden to process. You're trying, you just, most of your algorithms are designed just to filter through it and not actually learn anything from it, and so that tends to be not Be nearly as valuable as people think it is. The other thing that where people get trapped up about the specialness of their data sets is that you would like to believe when it, when a data set truly is valuable, it has this, it has this property that each new incremental piece of data is marginally more valuable to the whole thing than the last one that you put in. So you get this kind of accelerating effect, uh, rather than a diminishing returns kind of effect. Most of the data sets out there have this diminishing returns effect. A good example is like genomic databases where people try and say, okay, well, we sequenced 10,000 people instead of a thousand people or a 100,000 people, implying that their data sets are therefore 10 times more valuable because they have 10 times more sequences. But if you don't actually see any new genes or new, you know, Those 10 X more seq…

AI assessment note: “No, not really. Most of the data sets out there”

Q The hardest element of your role with Founders Fund?

A I would say the hardest element is the context switching. You know, one of the things I liked most about being a research quantum physicist is that you, you can think about one problem all the time and just go all the way, all the way down. And so the paradox about Being in venture is, is both a blessing and a curse, which is that you have to context switch between things that are, they're very different very quickly. And so it's both super exciting and that, you know, you get to weigh in on a very diverse set of topics, but you do sort of miss the intellectual satisfaction of having one particular thing that you think about for, you know, years at a time and really ruminate on it and go, go as deep as it can goes. And for things like quantum physics, It goes pretty deep, and so that's a, that can be really satisfying to be down there.

AI assessment note: “I would say the hardest element is the context switching.”

Q So a multitude of activities, uh, all incredibly value add, uh, when analyzing them. But I, I do want to discuss some of our pre-interview discussion, because you said something that really grabbed me in particular, and you said the AI is mostly a scam. So talk to me. What do you mean by this, and where do you think AI is most misleading? Let's start with that.

A Yeah. So I think that most of the things that call themselves AI out there are really just a rebranding. So originally, many, many years ago, we had actuaries, and actuary science was like the most boring, the most mundane, the most tedious discipline. Actuaries were the people that even accountants made fun of You know, no one, no, no one liked it. No one dreamed of becoming an actuary when they grew up, and nobody would go on dates with them, and so they, they sort of got together and decided that this is a terrible place to be, and in terms of the, the status relative to the value that they thought they were providing, and so they rebranded, and it's gone through several iterations, so first they became statisticians, and that was a little bit cooler than being an actuary, and they, that was the sort of, like, era of Moneyball, where people were like, okay, you know, these, these guys are providing Edges on sports, and sports is cool, so you get a little bit of cachet by association because you're hanging out with baseball players, but, you know, that sort of petered out, and so then they got rebranded as data science, and then as big data, and then as machine learning, and then, uh, now, you know, most sort of, most recently as artificial intelligence, but at the end of the day, like, what most of the things are actually doing there is just actuarial science with, you know,…

AI assessment note: “most of the things that call themselves AI out there are really just a rebranding”

Q with you, uh, both undeniable in scale and magnitude to society, I think we'll agree, is self-driving trucks. So what are your views on the societal implications of, I think it's four million truck drivers being unemployed over the next, we can give a timescale here, of Five to 20 years. We can kind of pick it as we choose, but what are the societal implications of this to you?

A Yeah, I think the societal implications are really huge. I think it's almost impossible to understate how big of a deal it is. If you think of, you know, in the last political cycle, how influential and how disruptive the idea of the out-of-work coal miner was, uh, well, there's only, you know, a few tens of thousands of coal miners in the United States, so it's like a relatively small constituency. Population wise, but had a, a very large effect on the overall political narrative. And so if you think that truck drivers, which are, you know, a thousand times more causing the same type of redundancy and displacement as cold splinters experienced as, as a, you know, natural gas came online, then it is a, is a big, big potential for societal people. The way that I look at it is not so much just You can't just focus on the truck drivers themselves, but really on the whole network. So the trucking system, you know, we have Eisenhower's interstate system and the road network on that's, that's built around that. The whole trucking system does serves two purposes at the same time. One, it's a goods distribution system. So we load things on the trucks, we ship them around. That's how we get all the stuff that we buy at the store, but it's also a wealth distribution system. And that you have wealth being generated at a few points of focus, things like factories, where the goods are creat…

AI assessment note: “big, big potential for societal upheaval.”

Q Genius and crazy. How do you determine between the two? I'm sure, I'm sure you've seen both. How do you kind of look to identify either camp?

A So part of it is thinking it through from first principles, so understanding a lot of the, like, you know, the backgrounds and what has been done before, what's sort of possible. You can, should be able to reconstruct someone's argument for, yes, like, we are able to solve all of the significant challenges here, and like, you know, these things are hard, but there are also a couple of interesting heuristics that we use to sort of As a check on the first principles thinking. One is just how thorough is the knowledge and the understanding of the people that are proposing it. Even if you don't necessarily yourself have the complete picture by just, you know, drilling down on specific aspects, you know, whether that person has, like, has all the answers and can go as deep as you want to go is a really good sign. If it turns out that the way we're thinking about it is very shallow, then that typically reveals itself pretty quickly.

AI assessment note: “drilling down on specific aspects, you know, whether that person has, like, has all the answers”

Q It's interesting because I often hear that kind of AI means big data can finally come into fruition. There you're suggesting it's more of a rebranding. How do you think about that then?

A One, one way of describing artificial intelligence or a particular deep learning, since, uh, at the core of it is just this matrix multiplication operations. It's, you know, you're just sort of doing linear algebra at industrial scale. We haven't gotten to the point for most of these things or most of the things that call themselves AI, where you're really doing something that is fundamentally different. You're just making it bigger or just having larger scale doesn't actually change the nature of the calculations. And I would, I would even sort of push back on when people tout their big data credentials, that most of the things that are sort of big data are actually not that big. You know, you would like to, you'd like to think of these petabytes and exabytes floating around there and coming up with really amazing insights that could never have been seen. But a lot of times it's really in the sort of megabytes and gigabytes kinds of kinds of range. And so it's always beneficial to say that you have this big data edge when really it's just more, more marketing and branding.

AI assessment note: “when really it's just more, more marketing and branding.”

Q Can I ask, would that kind of ability to handle ambiguity not kind of, ah, waver into the general artificial intelligence framework, and is that not kind of what's so commonly touted as the 20 to 40 years out?

A Well, I think there's, there are sort of layers to that ambiguity, right? There's different levels of abstraction that the ambiguity can present itself. And at some point we will be able to deal with the full layer, which, which includes things like moral ambiguity and be able to make rational and reasonable decisions about that. And so I think when you get to the point of human level, artificial general intelligence, uh, then that's the kind of thing you're talking about, but there's still lots of problems that I think we can be able to solve where things are just sort of You know, visual ambiguity, mechanical ambiguity, these types of things that hopefully we'll be able to make serious progress on, on the way to the human level general intelligence.

AI assessment note: “when you get to the point of human level, artificial general intelligence”

Q Now, this one might be challenging, but quantum computing in 60 seconds, why so exciting, and where are the opportunities?

A Super exciting because of the, the computational potential is dramatically more than classical or von Neumann computing as the, The laptops and cell phones that we have in our pockets are known. And the reason is because in quantum mechanics, we have these strange counterintuitive principles that we call superposition, which says that things can be or exist in more than one place at the same time, and entanglement, which says that things can be connected to each other in an information sense, even if they're not physically connected, even if they're, even if they're far apart. And using those principles We can operate on information, on pure information at the quantum mechanical level, and learn things that are sort of core to the natures of the universe that would be impossible with a regular computer.

AI assessment note: “Super exciting because of the, the computational potential is dramatically more than classical”

Q So a multitude of activities, uh, all incredibly value add, uh, when analyzing them. But I, I do want to discuss some of our pre-interview discussion, because you said something that really grabbed me in particular, and you said the AI is mostly a scam. So talk to me. What do you mean by this, and where do you think AI is most misleading? Let's start with that.

A Yeah. So I think that most of the things that call themselves AI out there are really just a rebranding. So originally, many, many years ago, we had actuaries, and actuary science was like the most boring, the most mundane, the most tedious discipline. Actuaries were the people that even accountants made fun of You know, no one, no, no one liked it. No one dreamed of becoming an actuary when they grew up, and nobody would go on dates with them, and so they, they sort of got together and decided that this is a terrible place to be, and in terms of the, the status relative to the value that they thought they were providing, and so they rebranded, and it's gone through several iterations, so first they became statisticians, and that was a little bit cooler than being an actuary, and they, that was the sort of, like, era of Moneyball, where people were like, okay, you know, these, these guys are providing Edges on sports, and sports is cool, so you get a little bit of cachet by association because you're hanging out with baseball players, but, you know, that sort of petered out, and so then they got rebranded as data science, and then as big data, and then as machine learning, and then, uh, now, you know, most sort of, most recently as artificial intelligence, but at the end of the day, like, what most of the things are actually doing there is just actuarial science with, you know,…

AI assessment note: “most of the things that call themselves AI out there are really just a rebranding”

Q Can I ask, would that kind of ability to handle ambiguity not kind of, ah, waver into the general artificial intelligence framework, and is that not kind of what's so commonly touted as the 20 to 40 years out?

A Well, I think there's, there are sort of layers to that ambiguity, right? There's different levels of abstraction that the ambiguity can present itself. And at some point we will be able to deal with the full layer, which, which includes things like moral ambiguity and be able to make rational and reasonable decisions about that. And so I think when you get to the point of human level, artificial general intelligence, uh, then that's the kind of thing you're talking about, but there's still lots of problems that I think we can be able to solve where things are just sort of You know, visual ambiguity, mechanical ambiguity, these types of things that hopefully we'll be able to make serious progress on, on the way to the human level general intelligence.

AI assessment note: “when you get to the point of human level, artificial general intelligence”

Q Okay, then taking this in our stride, and if we were to kind of break all these different buzzwords down, where do you believe in the differing components there is true scientific breakthroughs where you do get really honestly very excited?

A Yeah, so the place where I look for the real breakthroughs in artificial intelligence and where I would like to see the field progress more is in the way that it deals with ambiguity. I think that the biggest distinction between Human intelligence, which is what artificial intelligence is meant to model after or meant to work towards, is that humans have a much higher tolerance for ambiguity. In computers, where you have things are explicitly programmed, you have logic, whenever there is a sort of corner case that, you know, wasn't envisioned by the programmer or a kind of boundary condition that is not explicitly handled, then things break down. And so you, it becomes very fragile. And that's exactly the kind of situation that our modern homo sapien brains were evolved to process through. And that gives us a huge evolutionary competitive advantage. So being able to, when you have things you haven't seen before, to be able to think about it in a higher level of abstraction and deal with the ambiguity in a reasonable way and still be able to, to move forward. And most of the AI applications that we are using out there, Deep learning, pattern matching kinds of things can't actually do that. Can't use abstract thinking to deal with ambiguity. They can only sort of say, well, this is just the closest thing that I've seen to it before, and so we're just going to go with that.

AI assessment note: “the place where I look for the real breakthroughs... is in the way that it deals with ambiguity”

Q Well, I have to ask, because you're the first chief scientist I've ever had on the show, which is probably one of the reasons I'm so excited. But what is the role of chief scientist? What does a day in the life look like? And do you think every VC firm should have one?

A Sure. So there's three parts. One is providing a scientific context for For the investments that we make. So that's sort of keeping an eye on what is happening in academic and government science so that we have an overall idea of where things are going. Some of that is, um, for sourcing, but a lot of that is just we don't want to invest in something that we think is going to be obsoleted by the next thing coming out of the labs. And so having that sort of overall framework, that sort of voice in the room where we, we have Some idea of where things are going is, is super helpful. Uh, the second part is once we get interested in something into, into a company doing the scientific diligence so we can say, you know, this, yes, this is going to work, or yes, this is the team that has the capability to get this sort of thing to work. Uh, and there's a very thin line between genius and crazy, right? So there's a lot of discrimination that goes on between the ideas that are just crazy enough to work.

AI assessment note: “So there's three parts. One is providing a scientific context for For the investments”

Q Okay, then taking this in our stride, and if we were to kind of break all these different buzzwords down, where do you believe in the differing components there is true scientific breakthroughs where you do get really honestly very excited?

A Yeah, so the place where I look for the real breakthroughs in artificial intelligence and where I would like to see the field progress more is in the way that it deals with ambiguity. I think that the biggest distinction between Human intelligence, which is what artificial intelligence is meant to model after or meant to work towards, is that humans have a much higher tolerance for ambiguity. In computers, where you have things are explicitly programmed, you have logic, whenever there is a sort of corner case that, you know, wasn't envisioned by the programmer or a kind of boundary condition that is not explicitly handled, then things break down. And so you, it becomes very fragile. And that's exactly the kind of situation that our modern homo sapien brains were evolved to process through. And that gives us a huge evolutionary competitive advantage. So being able to, when you have things you haven't seen before, to be able to think about it in a higher level of abstraction and deal with the ambiguity in a reasonable way and still be able to, to move forward. And most of the AI applications that we are using out there, Deep learning, pattern matching kinds of things can't actually do that. Can't use abstract thinking to deal with ambiguity. They can only sort of say, well, this is just the closest thing that I've seen to it before, and so we're just going to go with that.

AI assessment note: “the place where I look for the real breakthroughs in artificial intelligence and where I would like to see the field progress more is in the way that it deals with ambiguity”

Q Before we go into the quick part, I'm going to toss one interesting topic into the mix, being universal basic income, potential for implementation, effectiveness of doing so. What are your thoughts?

A Yeah, so I think the potential is probably there. My biggest concern, and to first order, I think that it's sort of correct, or maybe something more like negative income tax might be sort of a mechanically better way to do it. We do have trouble decoupling wealth and virtue in our political system. One of the consequences of that is we like to put lots of requirements on certain types of welfare. So you think things like drug testing and worked well for and like Limitations on what can be spent with welfare money. The hope is that if you can get rid of a lot of those social programs and the overhead and administrative costs associated with them, and really free people up to do whatever they want with, with those resources, you end up with much more efficient system, much better outcomes. I think in terms of actually getting to the scale that's required, it may not be enough, may not, there may not be as, as much efficiency to be gained as we think. It's sort of answering the wrong question. It's like, well, you know, so we, we changed the goods distribution system. We changed the wealth distribution system that has been a consequence of the, the trucking system. And so now we're just trying to sort of force a replacement wealth distribution system that we can manage. And I think that there's sort of a lack of creativity there. What I think is actually a better way to do it, a m…

AI assessment note: “so I think the potential is probably there. My biggest concern”

Q I'm sure you've thought about it before, but how would you then look to approach it from a societal perspective? Perspective in terms of dealing with this kind of destabilizing of wealth distribution.

A Yeah. I think the best way to do it is to use that goods distribution system as to your advantage. So if you think about other times in the past where this level of societal level of people has happened, and a couple of examples are, you know, we used to have 90% of the workforce involved in agriculture. Now it's less than two percent. And so because we invented tractors and combines, it was very disruptive, But it created a better situation in the long run because now we had much more, much cheaper food, much more available agricultural products, and so that freed up a lot of human potential to go and do other things, you know, to invent cars and become yoga instructors and, like, all the great things that we have in society that weren't possible when everyone was, like, forced to be involved in agriculture because we didn't, we didn't have enough free human, human capital, human potential. So, The best case scenario is we use that good distribution to our advantage. And there's a few things that are sort of set up to do that. If you think about businesses like eBay and Etsy, where you have folks that are creating things that people want or selling things that people want, but the biggest barrier to getting them to the, to the customers, even if they can find them over the internet is still costly and to, to do the shipping. So, but if you have a much better, much faster, safe…

AI assessment note: “provide a wealth distribution system for the, for the folks that are out there”

Q Well, I have to ask, because you're the first chief scientist I've ever had on the show, which is probably one of the reasons I'm so excited. But what is the role of chief scientist? What does a day in the life look like? And do you think every VC firm should have one?

A Sure. So there's three parts. One is providing a scientific context for For the investments that we make. So that's sort of keeping an eye on what is happening in academic and government science so that we have an overall idea of where things are going. Some of that is, um, for sourcing, but a lot of that is just we don't want to invest in something that we think is going to be obsoleted by the next thing coming out of the labs. And so having that sort of overall framework, that sort of voice in the room where we, we have Some idea of where things are going is, is super helpful. Uh, the second part is once we get interested in something into, into a company doing the scientific diligence so we can say, you know, this, yes, this is going to work, or yes, this is the team that has the capability to get this sort of thing to work. Uh, and there's a very thin line between genius and crazy, right? So there's a lot of discrimination that goes on between the ideas that are just crazy enough to work.

AI assessment note: “So there's three parts. One is providing a scientific context for For the investments”

Q Before we go into the quick part, I'm going to toss one interesting topic into the mix, being universal basic income, potential for implementation, effectiveness of doing so. What are your thoughts?

A Yeah, so I think the potential is probably there. My biggest concern, and to first order, I think that it's sort of correct, or maybe something more like negative income tax might be sort of a mechanically better way to do it. We do have trouble decoupling wealth and virtue in our political system. One of the consequences of that is we like to put lots of requirements on certain types of welfare. So you think things like drug testing and worked well for and like Limitations on what can be spent with welfare money. The hope is that if you can get rid of a lot of those social programs and the overhead and administrative costs associated with them, and really free people up to do whatever they want with, with those resources, you end up with much more efficient system, much better outcomes. I think in terms of actually getting to the scale that's required, it may not be enough, may not, there may not be as, as much efficiency to be gained as we think. It's sort of answering the wrong question. It's like, well, you know, so we, we changed the goods distribution system. We changed the wealth distribution system that has been a consequence of the, the trucking system. And so now we're just trying to sort of force a replacement wealth distribution system that we can manage. And I think that there's sort of a lack of creativity there. What I think is actually a better way to do it, a m…

AI assessment note: “Yeah, so I think the potential is probably there.”

Q I'm sure you've thought about it before, but how would you then look to approach it from a societal perspective? Perspective in terms of dealing with this kind of destabilizing of wealth distribution.

A Yeah. I think the best way to do it is to use that goods distribution system as to your advantage. So if you think about other times in the past where this level of societal level of people has happened, and a couple of examples are, you know, we used to have 90% of the workforce involved in agriculture. Now it's less than two percent. And so because we invented tractors and combines, it was very disruptive, But it created a better situation in the long run because now we had much more, much cheaper food, much more available agricultural products, and so that freed up a lot of human potential to go and do other things, you know, to invent cars and become yoga instructors and, like, all the great things that we have in society that weren't possible when everyone was, like, forced to be involved in agriculture because we didn't, we didn't have enough free human, human capital, human potential. So, The best case scenario is we use that good distribution to our advantage. And there's a few things that are sort of set up to do that. If you think about businesses like eBay and Etsy, where you have folks that are creating things that people want or selling things that people want, but the biggest barrier to getting them to the, to the customers, even if they can find them over the internet is still costly and to, to do the shipping. So, but if you have a much better, much faster, safe…

AI assessment note: “provide a wealth distribution system for the, for the folks that are out there.”