The Exchanges

Q What are the benefits of this kind of radical interdisciplinary thinking with startups?

A So I think if a startup can bring together disparate ideas and carve out a niche, um, and unique way of thinking and provide something of value to the world, that's something that is both differentiable and also a defensible moat, um, because it's hard to jump across different domains. So a startup's openness to interdisciplinary thinking Can therefore be a proxy for its ability to adapt and to thrive. So for example, let's say a company can bring together a team or a set of ideas from computer science, AI, and the tax code, which is something that's reasonably novel and hard to replicate. Most accountants aren't playing with machine learning and vice versa. Or take the idea of, let's see, uh, combining together recipe generation and big data, which is something that's actually part of a field called computational gastronomy, which is this idea of using computational approaches to do things like Rethinking how ingredients and flavors combine to create novel foods, and so these kinds of interdisciplinary perspectives, they're very useful, and a number of tech companies are actually rethinking the future of food through this kind of stuff, but ultimately, interdisciplinary thinking, it's not just a source of new ideas, but it also provides new and defensible modes of operation for startups.

AI assessment note: “that's something that is both differentiable and also a defensible moat”

Q So let's start today with a little on you, and how you came to be in the very unique position of scientist in residence at Lux Capital. What's the story for getting that job?

A And mine is certainly a circuitous path, I guess, as it is for many people in the venture world. I grew up in Buffalo, Obsessed with science, technology, science fiction, surrounded by a mix of Martin Gardner, Legos, Arthur C. Clarke, all that kind of stuff. And I wanted to work at the frontier of science and went eventually to graduate school for a PhD in computational biology, where I was initially interested in using computational mathematical approaches to understand evolution. And then while I was there, I became fascinated by complexity science more generally, which is the science of trying to understand big complex systems, whether in biology, technology, or social systems like cities or societies. And, and so after the PhD, while I was doing a postdoc involved in this area, I also began thinking about the nature of innovation more generally in the future of scientific and technological change. Uh, and so then after my postdoc, I left the world of academia and then joined the Kauffman foundation doing essentially that kind of stuff, trying to better understand innovation, science and tech, and then became enmeshed in the world of startups and their role in essentially unfroling the future. Uh, and then, so alongside this, I also began doing a lot, a lot of writing about science and technology for popular audiences too. And so then after the Kauffman foundation, I I ended…

AI assessment note: “I ended up joining Lux as their scientist in residence to help bring this sort of interdisciplinary approach”

Q Can I ask, in terms of the creativity element, do you think man's appreciation of machines will extend to the extent that they actually appreciate creativity augmenting that? So if you see a piece of art that is part human, part machine made, Do you think their appreciation will be the same with the knowledge that it's been partly done due to machine?

A I think people will appreciate it, and I think we're going to see new and more exciting forms of creativity, but we already have this. I mean, when I was younger, people were on their computers, they had screensavers, and we don't really need screensavers anymore. I mean, they're still around, but they've kind of fallen out of favor, but screensavers, in a sense, they were this beautiful computational creative act, but they were incredibly generative and algorithmic, where you didn't really know exactly what The screensaver was going to look like from moment to moment. And so there was this combination of, oh, it's this algorithm generating something, but the algorithm was written by a person. And I think that kind of thing is perfectly fine. And, and the way I like to think about this is there's a term from Yiddish called nachos, which is essentially kind of like this, like a vicarious pride or joy that you have. So like you might have nachos or the Yiddish term is to shop nachos at like a kid's wedding or, or, or graduation. Like these are not your own achievements, but you can still take pride and joy in them. And I And I think to a certain degree, um, even though it almost sounds a little bit silly, we are going to increasingly need to have a certain amount of nachos for our technologies, where these technologies might be doing things, might be doing activities that they mi…

AI assessment note: “I think people will appreciate it, and I think we're going to see new and more exciting”

Q What are the benefits of this kind of radical interdisciplinary thinking with startups?

A So I think if a startup can bring together disparate ideas and carve out a niche, um, and unique way of thinking and provide something of value to the world, that's something that is both differentiable and also a defensible moat, um, because it's hard to jump across different domains. So a startup's openness to interdisciplinary thinking Can therefore be a proxy for its ability to adapt and to thrive. So for example, let's say a company can bring together a team or a set of ideas from computer science, AI, and the tax code, which is something that's reasonably novel and hard to replicate. Most accountants aren't playing with machine learning and vice versa. Or take the idea of, let's see, uh, combining together recipe generation and big data, which is something that's actually part of a field called computational gastronomy, which is this idea of using computational approaches to do things like Rethinking how ingredients and flavors combine to create novel foods, and so these kinds of interdisciplinary perspectives, they're very useful, and a number of tech companies are actually rethinking the future of food through this kind of stuff, but ultimately, interdisciplinary thinking, it's not just a source of new ideas, but it also provides new and defensible modes of operation for startups.

AI assessment note: “both differentiable and also a defensible moat, um, because it's hard to jump across”

Q So let's start today with a little on you, and how you came to be in the very unique position of scientist in residence at Lux Capital. What's the story for getting that job?

A And mine is certainly a circuitous path, I guess, as it is for many people in the venture world. I grew up in Buffalo, Obsessed with science, technology, science fiction, surrounded by a mix of Martin Gardner, Legos, Arthur C. Clarke, all that kind of stuff. And I wanted to work at the frontier of science and went eventually to graduate school for a PhD in computational biology, where I was initially interested in using computational mathematical approaches to understand evolution. And then while I was there, I became fascinated by complexity science more generally, which is the science of trying to understand big complex systems, whether in biology, technology, or social systems like cities or societies. And, and so after the PhD, while I was doing a postdoc involved in this area, I also began thinking about the nature of innovation more generally in the future of scientific and technological change. Uh, and so then after my postdoc, I left the world of academia and then joined the Kauffman foundation doing essentially that kind of stuff, trying to better understand innovation, science and tech, and then became enmeshed in the world of startups and their role in essentially unfroling the future. Uh, and then, so alongside this, I also began doing a lot, a lot of writing about science and technology for popular audiences too. And so then after the Kauffman foundation, I I ended…

AI assessment note: “I ended up joining Lux as their scientist in residence to help bring this sort of interdisciplinary approach”

Q I want to speak, though, first, kind of, before we jump into those modes, I want to speak between the balance of generalist and kind of that interdisciplinary thinking, and just how you balance between specialization and interdisciplinary Interdisciplinary thinking within startups and within yourself? Is it something where organizations should have specialists and then those interdisciplinary thinkers? How do you balance between the two?

A So I think it is a balance. And so there's kind of the organizational level of organizations need to have the ability to bring lots of different people together who are specialists, but you also need individuals who are more generous minded who can do that within a single person. And so this, this often involves actually the ability to overcome jargon boundaries. So like what, let's say one person who's deeply versed in one specific area Has their own sort of jargon, sort of language that they use to talk about something. And there's another area that the same problem arises where they also have their jargon barriers. You have to find someone who actually is somewhat averse in both of these who can almost act as a certain amount of translation and can actually translate from one area to another. Thinking about what a good balance might be. So there's a term that some people talk about called T-shaped individuals. This idea that, I mean, you kind of look at like the actual physical shape of the T. You have kind of the horizontal part and the, and the vertical part. So the vertical part says, You need to be deeply versed in one area, but you also need to have this horizontal ability, this kind of ability to interact with other disciplines, and I think we need more and more of these T-shaped individuals, um, coupled with sometimes just deeply specialized individuals, but then alon…

AI assessment note: “organizations need to have the ability to bring lots of different people together who are specialists”

Q Can I ask, in terms of the creativity element, do you think man's appreciation of machines will extend to the extent that they actually appreciate creativity augmenting that? So if you see a piece of art that is part human, part machine made, Do you think their appreciation will be the same with the knowledge that it's been partly done due to machine?

A I think people will appreciate it, and I think we're going to see new and more exciting forms of creativity, but we already have this. I mean, when I was younger, people were on their computers, they had screensavers, and we don't really need screensavers anymore. I mean, they're still around, but they've kind of fallen out of favor, but screensavers, in a sense, they were this beautiful computational creative act, but they were incredibly generative and algorithmic, where you didn't really know exactly what The screensaver was going to look like from moment to moment. And so there was this combination of, oh, it's this algorithm generating something, but the algorithm was written by a person. And I think that kind of thing is perfectly fine. And, and the way I like to think about this is there's a term from Yiddish called nachos, which is essentially kind of like this, like a vicarious pride or joy that you have. So like you might have nachos or the Yiddish term is to shop nachos at like a kid's wedding or, or, or graduation. Like these are not your own achievements, but you can still take pride and joy in them. And I And I think to a certain degree, um, even though it almost sounds a little bit silly, we are going to increasingly need to have a certain amount of nachos for our technologies, where these technologies might be doing things, might be doing activities that they mi…

AI assessment note: “I think people will appreciate it, and I think we're going to see new”

Q Speaking of the relationship between these two kind of thought processes, I'd love to speak about the relationship between Between AI and man. Now, we have many guests on the show who state their concerns for AI and the potential overhaul of mankind as we know it. How do you view the partnership, then, between man and machine, to start with?

A So, I'm going to limit my comments primarily to the partnership, at least in the space of creativity, but I think that there are a huge number of rapid advances where computational approaches mean that actually computers can make advances in domains that we view as broadly creative, and so this can be art and music and Design, even computer programming, or even scientific discovery. This is all fantastic, but it actually can be somewhat worrisome for some people. Um, so I view this as a partnership, where machines actually allow humans to sift through creative solutions, allowing us to do our work in even better ways. But even more than that, I think as the world around us is becoming more and more complex, it's becoming more difficult to fully understand either the technologies that we ourselves have built, or the systems that we're trying to understand in science, like a biological system, for example. And so, which means We might actually need computers to help us explore these systems that we might not fully understand any longer. So this can mean maybe we need the creation of things like automated hypothesis generation in science, or using machine learning and big data to optimize our own technologies, or even generate novel source code when we can't fully understand the technologies that we're building, or grapple with them in quite the same way we could when they were ma…

AI assessment note: “I view this as a partnership, where machines actually allow humans to sift through”

Q I want to speak, though, first, kind of, before we jump into those modes, I want to speak between the balance of generalist and kind of that interdisciplinary thinking, and just how you balance between specialization and interdisciplinary Interdisciplinary thinking within startups and within yourself? Is it something where organizations should have specialists and then those interdisciplinary thinkers? How do you balance between the two?

A So I think it is a balance. And so there's kind of the organizational level of organizations need to have the ability to bring lots of different people together who are specialists, but you also need individuals who are more generous minded who can do that within a single person. And so this, this often involves actually the ability to overcome jargon boundaries. So like what, let's say one person who's deeply versed in one specific area Has their own sort of jargon, sort of language that they use to talk about something. And there's another area that the same problem arises where they also have their jargon barriers. You have to find someone who actually is somewhat averse in both of these who can almost act as a certain amount of translation and can actually translate from one area to another. Thinking about what a good balance might be. So there's a term that some people talk about called T-shaped individuals. This idea that, I mean, you kind of look at like the actual physical shape of the T. You have kind of the horizontal part and the, and the vertical part. So the vertical part says, You need to be deeply versed in one area, but you also need to have this horizontal ability, this kind of ability to interact with other disciplines, and I think we need more and more of these T-shaped individuals, um, coupled with sometimes just deeply specialized individuals, but then alon…

AI assessment note: “So I think it is a balance. And so there's kind of the organizational level”

Q I would love to hear that, because there's always perils to every side. How do you assess the perils of such analogies?

A So I think in our rush to import ideas from computer science, data storage, or other technological concepts, we sometimes lose sight of the fact that biological systems, they're They're different in a very important way, which is they are massively and often non-intuitively complex. So biological systems, they've evolved over millions of years. They might be much messier and much more complex than an engineer or computer scientist might expect. There's actually this great XKCD comic highlighting this. There's this engineer type. He's arguing that biology should be simple because DNA is just source code. Of course, it's not. And engineered systems, they could be orders of magnitude much, much simpler than biological ones. And so when we apply ideas from computer science to biology, We end up failing sometimes. So if we use like traditional linear thinking to understand it, to understand them, we have to take into account the massive nonlinear complexity of these things. And so, so for example, one of, uh, Lux's investments recursion actually uses machine vision in order to find drug treatments for diseases, but it engages with the full force of biological complexity and high dimensional data and a lot of this nonlinear subtlety. Um, and so what this sometimes means is that you don't necessarily entirely understand how a drug works initially, but that's okay. Because it's a compl…

AI assessment note: “when we apply ideas from computer science to biology, We end up failing sometimes.”

Q Speaking of the relationship between these two kind of thought processes, I'd love to speak about the relationship between Between AI and man. Now, we have many guests on the show who state their concerns for AI and the potential overhaul of mankind as we know it. How do you view the partnership, then, between man and machine, to start with?

A So, I'm going to limit my comments primarily to the partnership, at least in the space of creativity, but I think that there are a huge number of rapid advances where computational approaches mean that actually computers can make advances in domains that we view as broadly creative, and so this can be art and music and Design, even computer programming, or even scientific discovery. This is all fantastic, but it actually can be somewhat worrisome for some people. Um, so I view this as a partnership, where machines actually allow humans to sift through creative solutions, allowing us to do our work in even better ways. But even more than that, I think as the world around us is becoming more and more complex, it's becoming more difficult to fully understand either the technologies that we ourselves have built, or the systems that we're trying to understand in science, like a biological system, for example. And so, which means We might actually need computers to help us explore these systems that we might not fully understand any longer. So this can mean maybe we need the creation of things like automated hypothesis generation in science, or using machine learning and big data to optimize our own technologies, or even generate novel source code when we can't fully understand the technologies that we're building, or grapple with them in quite the same way we could when they were ma…

AI assessment note: “I view this as a partnership, where machines actually allow humans to sift through creative solutions”

Q I would love to hear that, because there's always perils to every side. How do you assess the perils of such analogies?

A So I think in our rush to import ideas from computer science, data storage, or other technological concepts, we sometimes lose sight of the fact that biological systems, they're They're different in a very important way, which is they are massively and often non-intuitively complex. So biological systems, they've evolved over millions of years. They might be much messier and much more complex than an engineer or computer scientist might expect. There's actually this great XKCD comic highlighting this. There's this engineer type. He's arguing that biology should be simple because DNA is just source code. Of course, it's not. And engineered systems, they could be orders of magnitude much, much simpler than biological ones. And so when we apply ideas from computer science to biology, We end up failing sometimes. So if we use like traditional linear thinking to understand it, to understand them, we have to take into account the massive nonlinear complexity of these things. And so, so for example, one of, uh, Lux's investments recursion actually uses machine vision in order to find drug treatments for diseases, but it engages with the full force of biological complexity and high dimensional data and a lot of this nonlinear subtlety. Um, and so what this sometimes means is that you don't necessarily entirely understand how a drug works initially, but that's okay. Because it's a compl…

AI assessment note: “when we apply ideas from computer science to biology, We end up failing sometimes.”

Q Does this not go deeply contrasting the traditional VC thesis of you have to focus, focus, focus?

A So you certainly have to focus on your success, but if that focus derives from bringing together ideas from multiple different areas, then you should be doing that. At the same time, though, you don't want to be billed as a dilettante, just someone who says, okay, here's a really cool area, but the Wait, there's also this other area over here that's really interesting, and I can also combine these other two things and maybe do something else. If that's not part of your core mission, and you can't see a clear path to success, then you are going to be distracted. So there is a balance between you want to focus, and maybe that focus derives from being able to bring different ideas together in some sort of novel and defensible way, but at the same time, though, you don't want to just kind of be interested in lots of different things because they're simply interesting.

AI assessment note: “So you certainly have to focus on your success, but if that focus derives”

Q No, I agree, and I love that example of the screensavers and the curtains, but I do want to delve one layer deeper into the scientific stack, which is, is worrying for me in this conversation with you, but we come to the integration of bio and CS and AI, so tell me first, What's the promise of analogizing in the face of, of such a complex system?

A So obviously there's some very intriguing similarities between biology and computer science, uh, and I'm far from the first one to notice this kind of stuff like we have, like in biology, it's defined by a discrete language of symbols. So in this case, base pairs of DNA. I mean, computers operate in bits. And in fact, it turns out actually there's some interesting historical analogies where the, the common roots, they're actually from around the same time period. So like, And that was also around the same time when a lot of early kinds of computing were developed, the kind of computing that we're familiar with. We're now able to store data in DNA, which has a certain similarity to, to computation. Uh, modularity is a hallmark of both biology and technology. We've actually seen in the news recently this, uh, that malware was encoded in DNA and then it, then it compromised and took over a machine involved in analyzing the sequences, uh, which is pretty wild stuff. So there's a, there's a great potential for taking ideas from one domain and applying them to the other. We've seen this in many people From the CS and AI side, applying this sort of technological thinking to biological wetware. Of course, though, that's not quite the whole story. There's promise, but there are also a little bit of perils in doing this kind of thing as well.

AI assessment note: “there's a great potential for taking ideas from one domain and applying them to the other”

Q Can I ask, at what stage does one count that integration of man and machine, if you know what I mean? Where does the, the limits of man's ability reach the beginning of machines? So in terms of, like, aiding and assessing us in reaching the next frontier, where does that barrier start where, kind of, the extent of man's ability and where machines can come into the foray?

A I think we've always been using machines to a certain degree to, kind of, augment discovery and creativity. I mean, this can be everything as simple as using, um, Like, yeah, calculators, or various different tools to make our art, or using telescopes or microscopes to discover new things, and so I think this kind of thing has always been with us. That being said, it does feel somewhat different when a certain amount of the cognitive effort is being offloaded into machine and algorithmic world. We've always been doing this, and so there's always been a certain amount of partnership, and so it's being able to kind of use these tools in service of our own creativity as opposed to At the expense of kind of just being worried about all these advances.

AI assessment note: “I think we've always been using machines to a certain degree”

Q Can I ask, how would you like to see this integrated into the current ecosystem? Is this from a startup level? Is this from a VC level? Is this LP level? Is this integrated throughout?

A I do think we need to have it integrated to a certain degree throughout. I'll contrast it with, like, the startup ecosystem is certainly in a much better shape than, um, for example, like academia, where academia is incredibly siloed. But we need to make sure, I mean, startups are drawing people from lots of different areas with wildly divergent backgrounds and knowledge. But the other thing, though, is you also have to recognize that in many cases, startups, like, you want them to be monomaniacally focused on a certain topic, and that makes them much, much more successful, while oftentimes in the venture world, you have the luxury of being able to think about lots of different ideas. That being said, in the startup world, you might want to be monomaniacally focused on your success, but that success might derive from many, many different areas, and so being able to synthesize these In some sort of productive way, really gives you some sort of competitive advantage. So you really actually need it at both levels, kind of the startup level, as well as the investor level.

AI assessment note: “I do think we need to have it integrated to a certain degree throughout.”

Q No, I agree, and I love that example of the screensavers and the curtains, but I do want to delve one layer deeper into the scientific stack, which is, is worrying for me in this conversation with you, but we come to the integration of bio and CS and AI, so tell me first, What's the promise of analogizing in the face of, of such a complex system?

A So obviously there's some very intriguing similarities between biology and computer science, uh, and I'm far from the first one to notice this kind of stuff like we have, like in biology, it's defined by a discrete language of symbols. So in this case, base pairs of DNA. I mean, computers operate in bits. And in fact, it turns out actually there's some interesting historical analogies where the, the common roots, they're actually from around the same time period. So like, And that was also around the same time when a lot of early kinds of computing were developed, the kind of computing that we're familiar with. We're now able to store data in DNA, which has a certain similarity to, to computation. Uh, modularity is a hallmark of both biology and technology. We've actually seen in the news recently this, uh, that malware was encoded in DNA and then it, then it compromised and took over a machine involved in analyzing the sequences, uh, which is pretty wild stuff. So there's a, there's a great potential for taking ideas from one domain and applying them to the other. We've seen this in many people From the CS and AI side, applying this sort of technological thinking to biological wetware. Of course, though, that's not quite the whole story. There's promise, but there are also a little bit of perils in doing this kind of thing as well.

AI assessment note: “great potential for taking ideas from one domain and applying them to the other.”

Q Can I ask, how would you like to see this integrated into the current ecosystem? Is this from a startup level? Is this from a VC level? Is this LP level? Is this integrated throughout?

A I do think we need to have it integrated to a certain degree throughout. I'll contrast it with, like, the startup ecosystem is certainly in a much better shape than, um, for example, like academia, where academia is incredibly siloed. But we need to make sure, I mean, startups are drawing people from lots of different areas with wildly divergent backgrounds and knowledge. But the other thing, though, is you also have to recognize that in many cases, startups, like, you want them to be monomaniacally focused on a certain topic, and that makes them much, much more successful, while oftentimes in the venture world, you have the luxury of being able to think about lots of different ideas. That being said, in the startup world, you might want to be monomaniacally focused on your success, but that success might derive from many, many different areas, and so being able to synthesize these In some sort of productive way, really gives you some sort of competitive advantage. So you really actually need it at both levels, kind of the startup level, as well as the investor level.

AI assessment note: “I do think we need to have it integrated to a certain degree throughout.”

Q with in terms of we've always been using machines to augment our ability. One thing that has changed is kind of the speed of integration into society, and I don't think we've ever seen such rapid technological advancement so quickly as we do today and integrating that so fast into society. Do you have concerns around how quickly it can be integrated and the kind of repercussions of such speed?

A So I think there definitely are certain concerns. In terms of the speed, in terms of like automation, destabilizing job stability, and things like that. But I think when it comes to certain areas within creativity, ideally, I think over the next like several years, it's going to involve building systems that actually make creative knowledge workers more productive versions of themselves. So rather than plateauing when it comes to creative output or discovery, we can actually continue doing more. For me, there's going to be certainly some destabilizing effects, and I think we need to be very, very mindful about that. But for me, one of the things that I think Are really, really exciting. Are the ability to actually make novel types of advances and creative outputs in a way that were just not possible before. And so like, for example, going back to what I was talking about, the incredibly complex systems around us, like whether or not it's like things in biology that are hard to understand, or even just dealing with the technologies that we ourselves have built in terms of machine learning and things like that. There's a certain amount of limitation in terms of all the stuff that we can hold in our head and actually truly understand. And so we're going to need these kind of Machine partners in order to continue making advances, and so rather than viewing it as humans are being di…

AI assessment note: “there definitely are certain concerns. In terms of the speed, in terms of like automation”

Q Can I ask, at what stage does one count that integration of man and machine, if you know what I mean? Where does the, the limits of man's ability reach the beginning of machines? So in terms of, like, aiding and assessing us in reaching the next frontier, where does that barrier start where, kind of, the extent of man's ability and where machines can come into the foray?

A I think we've always been using machines to a certain degree to, kind of, augment discovery and creativity. I mean, this can be everything as simple as using, um, Like, yeah, calculators, or various different tools to make our art, or using telescopes or microscopes to discover new things, and so I think this kind of thing has always been with us. That being said, it does feel somewhat different when a certain amount of the cognitive effort is being offloaded into machine and algorithmic world. We've always been doing this, and so there's always been a certain amount of partnership, and so it's being able to kind of use these tools in service of our own creativity as opposed to At the expense of kind of just being worried about all these advances.

AI assessment note: “I think this kind of thing has always been with us.”

Q with in terms of we've always been using machines to augment our ability. One thing that has changed is kind of the speed of integration into society, and I don't think we've ever seen such rapid technological advancement so quickly as we do today and integrating that so fast into society. Do you have concerns around how quickly it can be integrated and the kind of repercussions of such speed?

A So I think there definitely are certain concerns. In terms of the speed, in terms of like automation, destabilizing job stability, and things like that. But I think when it comes to certain areas within creativity, ideally, I think over the next like several years, it's going to involve building systems that actually make creative knowledge workers more productive versions of themselves. So rather than plateauing when it comes to creative output or discovery, we can actually continue doing more. For me, there's going to be certainly some destabilizing effects, and I think we need to be very, very mindful about that. But for me, one of the things that I think Are really, really exciting. Are the ability to actually make novel types of advances and creative outputs in a way that were just not possible before. And so like, for example, going back to what I was talking about, the incredibly complex systems around us, like whether or not it's like things in biology that are hard to understand, or even just dealing with the technologies that we ourselves have built in terms of machine learning and things like that. There's a certain amount of limitation in terms of all the stuff that we can hold in our head and actually truly understand. And so we're going to need these kind of Machine partners in order to continue making advances, and so rather than viewing it as humans are being di…

AI assessment note: “So I think there definitely are certain concerns. In terms of the speed”

Q Does this not go deeply contrasting the traditional VC thesis of you have to focus, focus, focus?

A So you certainly have to focus on your success, but if that focus derives from bringing together ideas from multiple different areas, then you should be doing that. At the same time, though, you don't want to be billed as a dilettante, just someone who says, okay, here's a really cool area, but the Wait, there's also this other area over here that's really interesting, and I can also combine these other two things and maybe do something else. If that's not part of your core mission, and you can't see a clear path to success, then you are going to be distracted. So there is a balance between you want to focus, and maybe that focus derives from being able to bring different ideas together in some sort of novel and defensible way, but at the same time, though, you don't want to just kind of be interested in lots of different things because they're simply interesting.

AI assessment note: “you certainly have to focus on your success, but if that focus derives from”