I got to talk with Sam Arbesman about The Overedge Catalog, Scientific Discovery, and Complexity. You can check out Sam’s blog at: https://arbesman.net/
Will Jarvis 0:05
Hey folks, welcome to narratives. narratives is a podcast exploring the ways in which the world is better than in the past, the ways that is worse in the past towards a better, more definite vision of the future. I’m your host, William Jarvis. And I want to thank you for taking the time out of your day to listen to this episode. I hope you enjoy it. You can find show notes, transcripts and videos at narratives podcast.com.
Well, Sam, how are you doing today? I’m doing very well. Thanks. How are you? Good. Thank you for coming on. Could you go ahead and give us just a brief background and some of the big things you’re interested in?
Sam Arbesman 0:53
Sure, um, a brief background. Hopefully this will be still not not too non brief. Um, I grew up in Buffalo, I have a My family is pretty came from like a pretty medical or science oriented family. So I was kind of in that mix of like science and interesting thing about science and technology. Certainly one person who actually had a really big impact on my, on my interest was my grandfather. So he actually he is a dentist lived to the age of 99. Oh, nice. He. And he was very, very interested in science and technology in the future in science fiction. And he actually, uh, you read science fiction since like, the modern Dawn of the genre. I like he read. I think he read Dune when it was like serialized in a magazine. He read. And he read like Popular Science for like, I think over 70 years straight. And so it was really interesting, this kind of stuff. And so obviously, it rubbed off on me in terms of my own interest. So in terms of a little bit about my background, as a, I was very interested in science and technology, did an undergrad at Brandeis University in Computer Science in Biology, then continued on for a PhD at Cornell in got a PhD in computational biology. Initially, I was very interested in evolutionary biology, in computational mathematical modeling of evolution. And then while I was there, I actually got very interested in using those kind of models more broadly, to just kind of understand any large complex system. So it’s kind of moving more into the field of complexity, science or complex systems. And so it’s any large system with lots of interacting parts, whether it was in biology, technological systems, social systems, to move more into that began thinking a lot about networks of interacting components. And after I finished the PhD began doing began a postdoc in kind of the network science. I’ve actually, in the Department of Healthcare Policy at Harvard Medical School, I wasn’t doing anything healthcare, particularly related, it was just that that was where my advisor had the most office space. But it was a lot of like very interdisciplinary work, trying to understand the nature of human behavior, as well as trying to understand I also began during this time began getting very interested in how to quantify scientific and technological change. And so people, some people talk about this as like meta research, or meta science or meta knowledge also, there scientometrics, kind of the science of science, I began thinking a lot about this. During this time, I also began doing a lot more writing for general audiences. In addition to writing for other academics, I was also very interested in being able to articulate these ideas to general audiences and kind of to the public at large, I began reading about math and science and technology, which was a lot of fun, began at your writing my first book for general audiences. And then after that postdoc, I actually left academia and joined the Kauffman Foundation, which is a big phone profit foundation devoted to entrepreneurship, innovation, education. And at the time, they were building this almost like a in house Think Tank. And I joined them as a senior scholar and kind of got to think about all the different topics I was thinking about the nature of innovation, scientific, and technological change, did some thinking there around why certain cities are more entrepreneurial than others. And, and then also began doing a lot more writing. And then now at you guys, after I left Kauffman, I joined Lux capital, a venture capital firm about, I guess, almost six years ago now, which is in its early stage venture base in New York in the Bay Area. I’m based in Kansas City, and in my role is, as scientists and residents is to kind of survey the landscape of science and tech and find areas that we should be involved with weather. And then based on that, find companies who invest in find people to some times, we might build companies around these people engage with the public through writing and speaking. So I do a whole bunch of that, as well as connect the ideas and the people involved them that I’ve been exploring to our portfolio of companies that we’ve already invested in. So there’s a lot of almost like, import export of ideas and people which is a lot of fun. And so I would say the kind of themes that I spent a lot of my time thinking behind, there’s many, many different areas that I’m kind of thinking about at any one point but the overarching themes are Around, kind of dealing with and making sense of complexity, either like complex like the complex systems of our technologies that we’ve built, or the wealth of knowledge and information that is around us that we kind of constructed as part of kind of the scientific process. And so in the two books that I wrote, or about different aspects of this, I wrote the half life of facts, which is about how knowledge changes over time, how it has, like, what are the regularities to how knowledge grows, and changes and gets overturned, and how error is routed out and things like that. And then my second book, overcomplicated is about how our technologies have grown more and more complex, and increasingly, increasingly incomprehensible to not just like the everyday user of the iPhone, or whatever it is, but even the experts who deal with these technologies on a daily basis, or might have even actually been the ones who created these technologies. And so I’m very interested in kind of thinking about complex technologies, complex knowledge. And then related to that, I’m also very interested in how science works, how sometimes science doesn’t work. What are the new types of organizations, we might need to actually kind of continue to foster innovation. I’m also very interested in kind of, in addition to new types of organizations for thinking about this, but also the ways in which people are being trained, and they do science. And so do we need more generalist thinking in our kind of age of specialization. And I’m also just arranging to kind of the shape of the future and how science and technology affects society. I’m also I’m affiliated with a long now Foundation, which is an organization in San Francisco devoted to very long term thinking. And so I spend a lot of my time also thinking about, and what are the long term implications for these things? So yeah, so that anyway, that’s kind of a bird’s eye somewhat rambling overview of my background, as well as kind of the things I’m thinking about.
Will Jarvis 6:38
No, I love all that. And I wanted to just jump right in. One of the things you mentioned earlier was, you know, quantifying technological, scientific progress. How do you go about thinking about that?
Unknown Speaker 6:52
So I mean, and this is one of those things where I think it comes down to like, you often have to like be, and I’m not saying I’m the one being clever about this. But like, many, many scientists have worked on this, like, oftentimes, the scientists end up being like having to find ways of being clever about this kind of thing. So in one very well known metric is just simply how often science is cited, or even just like the number of journal articles like so more highly cited research would be viewed as more influential. And that the number of papers would mean that Oh, there’s more knowledge out there. Now, of course, I mean, the thing is, like, those, that’s clever, but it’s also at the same time, like we have to recognize those are just like, like, those are easy metrics to actually quantify science and research and discovery, there are a lot more messy than those kinds of things. So just because a paper is being cited doesn’t necessarily mean that someone says, Oh, this is really influential. And so, so like to give an example. And some people might be citing a study a scientific paper, because they say, Actually, I disagree with it, and I want to kind of overturn it. Other people would not be citing paper, because they’re so found that citing certain research, because it’s so foundational, you just don’t cite it anymore. So very rarely do I imagine physicists and upsetting like Newton’s Principia, whenever they’re writing scientific papers, I mean, just kind of like, it’s, it’s sort of in the air, and everyone recognizes that Isaac Newton was very, very important and influential, but they’re not necessary citing it. And so you have to kind of carve out a lot of the different ways of looking at these things. And there are many other different ways of looking at how research grows and return. But also, there’s also even things about thinking about how we like, like, recognizing, like how sometimes these metrics are unnecessarily, I guess, narrow, but also recognizing that not only are they unnecessarily narrow, but in some cases, that people use those metrics, not necessarily just as kind of like fun rules of thumb. But oftentimes, they actually have implications for like hiring, or getting tenure and things like that in scientific academia. And recognizing, though, that these things are still somewhat narrow. And the way I kind of think about this is that there are many, many activities that are actually valuable for science, but unfortunately, only a subset of them are actually valued by scientific academia. And some of that is because that those things are just easier to measure. So one example of this, which is kind of interesting was, there was a paper that came out maybe, I don’t know, decade or so ago, maybe a little bit longer. And it was about looking at researchers whose work was very high, highly cited. So had a lot of citations and references. But then they were also looking at people who were mentioned in the acknowledgments at the end of a paper, so they might and so because oftentimes, at the end of the paper, you’ll say, Oh, hey, like thanks for this person who may be some data analysis or just kind of get some idea or, or they’re kind of just like in the mix in some way. And this paper specifically looked at people who had a relatively mediocre number of citations, but a relatively high number of acknowledgments at the end of papers. And they found that when those people died, the the productivity of all their collaborators actually dropped quite a bit. And so what it shows is that there are these people who are really valid For Science, but we’re not necessarily like measuring this or quantifying or recognizing that as much, and I’m not. And we’ll just have to say that, therefore the solution is to start quantifying and measuring the number of times you get acknowledged, because I think that’ll end up being gamed in some sort of way, as well. But I do think that we need to recognize that there are many, many activities that are important for science. And, and they’re not always easily measured. I mean, if you think about like an innovation, I mean, like patents are certainly very valuable. But and there are many, many profound advances in like, especially in like, like the software world that have like no patents involved, whatsoever. And so recognizing how, how these things are very imperfect metrics. But still, sometimes the best metrics we have are, I think it just requires a certain amount of nuance and kind of how we measure it. But that being said, I still think these measurements are a lot better than nothing. And you you’re actually able to see a lot of really interesting insight in terms of like, exponential growth in number of papers and recognizing, and even the most rough metric I like, there’s just far too much scientific knowledge out there for anyone person to read. So we have know that even if we don’t know, sorry, know the detail. So yeah, there’s a lot there.
Will Jarvis 11:08
Definitely. That makes a lot of sense. This question goes off of that. Where do you fall in like the tech stagnation debate? Like, you know, Ross doubt that Peter Thiel since 1970? You know, we’ve got Twitter, but no flying cars, what do you think about that?
Unknown Speaker 11:21
So I think it’s really interesting, I think, I mean, I, I definitely think there’s some like really interesting, like, compelling data. There. I think one of the things. So I do think in some, in some ways, we kind of, we under value, some of the, like information technologies that have been developed over the past 50 years or so. And those things are really, really powerful and impressive. And, and yes, like, we don’t have, we don’t have flying cars yet. And like it kind of like since the 70s. In some ways, it feels like that. Some of like the kind of like, the innovations of like things in terms of like, we have like a lot of really great advances around like home appliances and cars and airplanes. We’ve got, like, 100 years before that, and then it’s kind of like, slow down. And so but but at the same time, though, I think, yes, some of the things when in terms of the innovation around stuff has slowed down a little bit. But I think Information Technology has, has not or at least is maybe a little bit discounted. At the same time, though. I think people are also recognizing that like over the past few years, and some of these things like some of the advances are beginning to speed up again. So it could just be it wasn’t necessarily this thing that we’re that we got all the low hanging fruit in our knowledge just over I think there’s there’s kind of a certain element of like, maybe things just needed to accumulate. And and now we’re kind of off to the races. Again, that being said, I do recognize that, like, even when you look at these curves, it might require more and more efforts to kind of make these advances. I do think actually, one interesting aspect of this is how we think is in terms of thinking about specialization. Oftentimes, when it when a scientific domain or some topic is new, that’s when you can kind of make a lot of like, easy and not not easy discoveries, but relatively easy, because it’s a lot of new low hanging fruit, like there’s just a lot of new ideas, and then it becomes more and more difficult. And so I think one of the things that we need to spend our effort on is almost like the development of new fields. And I actually think one of the ways in which we can develop new fields is at the intersection of otherwise less connected domains. So not necessarily say, not necessarily arguing that we should keep on making like sub specialties and sub sub specialties and keep on going more and more specialization but all but rather kind of recognizing that there’s still a lot of space for innovation at the intersection of domains. And I think that’s something that’s maybe a little bit less valued, but I think could help overcome some of this kind of like stagnation argument.
Will Jarvis 13:50
Gotcha. That makes sense. Do you think academia is just it’s just poorly suited to cross disciplinary work? And is there any way around that? So
Unknown Speaker 14:02
I do think to a certain degree, it is poorly suited to cross disciplinary work. And I think one of the major reasons is related to kind of the incentives. And so even if you want to do cross disciplinary work, you still are housed in a specific department and the metrics of success and whether or not you get tenure and things like that, or continue getting promoted is whether or not you can kind of play the game in your own specific domain and your own department. So even if you’re doing interdisciplinary work, if it’s too interdisciplinary, and kind of doesn’t really fit neatly into kind of, like my department plus, but it’s rather like my department, kind of but also this other thing, when you either have to be measured by multiple different departments, or you kind of are fitting in somewhere halfway kind of in between, to get to a can be pretty difficult. That being said, I mean, there’s a lot of interdisciplinary centers and institutes and universities and so I think a lot of academic organizations really Try to kind of make sure that this is not a problem. But and I would say I mean, so there’s kind of, I guess there’s kind of two aspects, there’s whether or not you’re doing interdisciplinary work. So whether or not it’s like, the like, I’m a specialist in one field, and I’m working with a specialist in another field, and we are together doing work that is interdisciplinary, right, or whether or not I am a interdisciplinary individual and whether or not I kind of cross lots of different disciplines. And so I think, and there’s problems, though, in both, but I definitely think that the, to specialize individuals working together, that is much more straightforward in academia, being an interdisciplinary individually, I think is very, very tough, because it’s very hard to kind of know where you fit. That being said, even kind of having interdisciplinary teams, it’s still tough, because I think there’s you have to overcome jargon barriers, like just being even able to know like, are we talking the same thing? Like, are we talking, hey, you can actually see this, like, people have done some interesting research. And they found that because of these kind of jargon, barriers, certain scientific advances or probability models, mathematical models, they’ve been reinvented, like many, many times in different domains, because no one’s realize it. So they were just talking about the same thing. But it was named one thing and one area was named another thing, another area. So you just you see this all the time. And I tried to remember, I think it was when I was a postdoc. I was on this, like, email list around network science. And so because it was network science, we had a lot of people from lots of different domains. Yeah. And so someone would email maybe like a physicist would email and say, like, oh, like, does anyone know about the metric to do the following thing around social networks? Yeah. And and then someone else from sociology would email back and say, like, Oh, this is actually been known for, like, 30 years. And so it was just like, so in this case, we were actually talking to each other. But if you had asked like, people would just would not have known. And so I still think there’s, um, it’s tough with those I, I’ve actually seen some research where I think people have argued, or they’ve, they found some evidence that the more interdisciplinary a team is the, I think it was like the hot the lower the average payoff in terms of like, impact of the research, but the higher the variance. So the idea would be like, for the most part, you fail, or you don’t do as well, but when you do succeed, you succeed in some spectacular fashion. So I think it which in which, oftentimes, in, in research and academia is the question of, okay, is University model well suited towards that kind of like high risk, high reward kind of model, right? And sometimes it is. And you could argue, oh, like, once you have tenure, that you can take lots of risks, but oftentimes, that is not necessarily the case. And so whether or not it’s like, high risk research or research that might take a longer amount of time than kind of the, like, the lifecycle of a grant cycle. These kinds of things just often don’t happen as often. And so yeah, so I think there are a number of barriers to kind of being able to do that kind of thing in academia.
Will Jarvis 17:42
That makes sense. That reminds me of a paper that blew up a couple years ago on Reddit, I think, was on like a Reddit map form, but it was a nursing paper where they invent the way to get an area under the curve.
Unknown Speaker 17:53
Like, mentioned, in the halfway facts. Yeah, I I actually, I actually just, I think this is the there was like his paper where he was like, basically, like reinventing like some, I think, was like the trapezoidal method. It’s not like, necklace or something like that. Yeah. And, and to the credit of that domain, like a lot of people wrote in, I think, into that journal article, like, into the journal after it came out saying, like, this is like, this is like, new vibe, but it’s developed as a novel. But at the same time, though, like, yeah, like that just kind of shows like that sort of, like an extreme example of Yeah, like, it’s really hard to be able to know all that is out there. And, and people have talked about this for decades. And there was a, there was an information scientist in the mid 1980s, who he wrote this paper, about what he referred to it as undiscovered public knowledge. And so that the information scientist Don Swanson, he said, okay, like, let’s do this thought experiment. And imagine that there’s somewhere in the literature, some paper that says a implies B. And somewhere else in the literature, it could be, I guess, somewhat, somewhat different subspecialty somewhere else entirely. There’s another paper that says B implies C, and B. And so even though if you read both of these papers, you would say, okay, maybe actually a imply See, you could actually draw a connection between them. Because of the vast amount of knowledge. No one has actually read both papers. And so he said, Okay, this is a fun thought experiment, but let me test it. And so he used kind of a Venn cutting edge technology of I think he was using like, the, like the medical database and Medline and he was doing keyword searches and saying, like, can I find something that’s like, just out there? Like, it’s it’s undiscovered, but public knowledge, like it’s knowledge that’s out there, but no one has actually connected it together? Yeah. And he actually ended up making some discoveries, I think he found he found a number of different things, including one where he he found a relationship between consuming fish oil and actually helping alleviate the symptoms around some like circulatory disease, circulatory condition. And then he I think, he ended up publishing it in a medical journal. And even though he was just and he was not, he was not a physician didn’t have any medical expertise, but he realized that there was just knowledge that was that was being under appreciated. And and since then, people have tried to create more automated and more sophisticated techniques for this, but there’s still so much out there and I I often feel that we just stopped publishing new papers today and said, Okay, let’s take like a final moratorium and just like, just delve into what’s already out there and try to find interesting connections to things, we would still be making a lot of really interesting advances. And so I think there’s a lot, there’s a lot of space there for just finding what is already out there and finding new ways to kind of navigate and cross pollinate and kind of interconnect different domains within science.
Will Jarvis 20:20
That’s awesome. That’s awesome. I love that anecdote, too. So yeah, just going into these different fields and, and finding information that’s really powerful, and being able to apply that somewhere else. And so this goes off my next question, you know, what do people generally just most misunderstand about how scientific discovery actually happens?
Unknown Speaker 20:39
I think in a pretty good amount, to a certain degree. Certainly, the things I mentioned in terms of like that, how we sometimes value only a certain subset of things that are important for scientific discovery, and because of that, we end up we end up like incentivizing those activities. And then sometimes the other things that are so really valuable in terms of maybe like, building software tools for science, or, or sharing data, these things. They don’t just like fall through the cracks, but they’re sometimes under valued. But I think one of the I would say one of the broadest things that people sometimes forget about is that scientific discovery, it’s a human process. It’s done by scientists. And so therefore, it’s going to be sucked. It’s not just like, oh, like brilliant hunches and quick, like quick analyses, and then just kind of like you write this perfect paper, and then it’s done. It’s there’s a lot of ups and downs, there’s dead ends, there’s like complex back and forth arguments, there’s work being done by irrational people, there’s the fact that like, like, a postdoc might have done some research and then decide not to continue on in academia and then left, and then no one actually knows how to, like replicate that kind of thing. Like, it’s just like science is done by people. And then I think, but more broadly than that, I mean, also, we need to recognize that in science, it’s not just a like, science is not a body of facts that we just memorize. And science is really, it’s a rigorous process of querying the world. And, and so if science is a process done by people, and so if we kind of say, oh, like scientists is perfect, pristine thing, and then we see evidence, otherwise, people are going to be a put off by that and surprise and concern versus saying, like, no, it’s actually this wonderfully human endeavor, that, that people get to participate in. And it’s not perfect. And we’re going to keep on improving things. And we’re going to be like, modifying things are having arguments about stuff and, and constantly figure out what we know and what we don’t know, like, and that’s wonderful. And I love this, this story that former former professor of mine in grad school, told me that he was he was teaching this. He was teaching a course on it came in on Tuesday gave some lecture. The next day, he read a paper that actually invalidated what he had taught. So he went on Thursday, the next time he was in class, and he said, Remember what I told you on Tuesday, it’s wrong. And if that bothers you, you need to get out of science. And so he wanted to kind of almost like, like, celebrate the fact that like science is constantly in draft form, like we’re always learning things new. Hopefully, we’re asymptotically approaching the truth, but like, but like the fact that, like, there’s, there’s a lot of things we don’t know. And that’s the reason scientists work at the frontier, like, we don’t want to work at all the things that are kind of known in the textbooks you want to work out, like, where they’re still all the arguments in the back and forth, and where there’s things that maybe are real, or maybe just like, like, statistical noise, and that’s where it’s really exciting. And so I think, I would say, and people sometimes forget that scientific discovery is done by humans and scientists. And, and it’s and it’s an amazing endeavor, it’s always in draft form, we kind of need to like revel in the fact that we’re always learning new new things and overturning things. And yeah, just working with like, as people. So yeah, I would say that that’s certainly one thing.
Will Jarvis 23:46
Yeah, that’s really important, like people are working under all these different constraints, you know, and it’s just like, it’s not a perfect process. You know, it’s not just like this perfect diamond that works all the time. The overage catalog, you know, what is it? And how did you first have the idea to put something like that together?
Unknown Speaker 24:02
Sure. So yes, the overhead costs. So it’s kind of stuff that take a step back. And so when, and when, when people often think about like science or innovation, and how research is done, they often think of it as being done in a few different places like it might be done in like in a university, or in a corporate industry lab, or maybe even sometimes, like in a tech startup. But the truth is, those are really just like those kind of organizations are just a few different points in some like what is really like a vastly high dimensional space. And the truth is like that high dimensional space is often under explored. And so the overage catalog is really my attempt at identifying some of the organizations that don’t really fit kind of the traditional institutional models. And but I would almost say they’re kind of like the Misfits for organizations and I’m using misfitting like the best possible way. And then some of these are for profit, some are nonprofit, some are residency, some are not so like some are like kind of like distributed collectives, but they’re all trying to kind of rethink institutional design and kind of doing research or kind of thinking about, like the progress of knowledge in some kind of interesting way. And so, and yeah, and so the way I kind of came up, came up with this, I’m not really sure if it kind of came up with it in any sort of specific moment. But over the past number of years, I’ve been thinking a lot about new types of organizational structures. And I, I’ve always been intrigued by these kind of weird outlier groups then. And by the way, so the term overage comes from like, this term in cartography, where you have a map, you have a border, but occasionally some like the information on the map kind of like seeps over the border, where maybe like a little, little bit of a river needs to go over the edge of some mountain range kind of goes around goes beyond the border. And those are referred to at least in this one glossary that I found as overages, I thought, Oh, this is great. He’s interesting outliers. And so and so I’ve just been trying to kind of find as, and I’ve always been intrigued by these kind of outlier organizations. And so I’ve been talking a lot of them trying to catalog them. And a certain point, I realized, I had a whole bunch of just like information and knowledge about these things, I should just kind of put it online. And there are other people who have kind of done somewhat similar lists or kind of collections. From their own perspective, I must admit that like, my, like, my listing of the organizations within the orange catalog is a very subjective list. And so since I began, I began publicizing the word catalog maybe a few months back. Since then, a lot of people have kind of written to me with suggestions of like, you know, their own organization or other organizations that that they’re just familiar with and saying, like, oh, maybe this is one that you should be including, and then some of them fit, some of them don’t, not. And whether or not they fit or don’t is very much like kind of my own judgment call. And I’m still trying to find like, almost like, are trying to figure out how best to articulate what these organizations are. But I have kind of view them as, ultimately, these organizations where when you kind of describe like, if you were running one of these organizations and trying to describe what you’re doing, there’s kind of like, there’s going to be a certain amount of like, struggle and confusion initially. And like, if there’s kind of, if you’re kind of struggling to figure out where you fit, that often means you’re doing like, at least from my perspective, you’re doing something really interesting, right? And so, when I developed this catalog, I wanted to publicize it one. And and I would say, first of all, just because I think these are the kind of places where interesting people with interdiscipline interests can actually thrive. But I also want in these organizations to realize like, there’s a community here, like these organizations are not alone. And the truth is, a lot of these organizations are familiar with each other already. So it’s not like I’m like, like, introducing them to this whole new world. But I but I think for many people, as people think about starting organizations, they can realize that even if you’re kind of going along a path that is less trodden, there still are people that you can talk to him and kind of work with, right and discuss things with until Yeah, and I’ve had some interesting conversations with people who are kind of in the beginning of starting these organizations, and kind of thinking about all this kind of stuff. And yeah, it’s been a while since I kind of began publicizing the catalog a few months ago, it’s been this like, wonderful experience, just talking to all these interesting people who are running these organizations who are thinking about these organizations. I just want there to be more I kind of view it as like, we need to begin populating this high dimensional space. Whether it will these organizations continue, or they’re going to all be be still out there. And like 10 or 20 years, I don’t know. But I kind of view it almost as like evolutionary process. We’re in the beginning, I want there to be a Cambrian explosion of all these cool organizations. And and hopefully some of them will survive. And like maybe we will hit on some really interesting attractors within this high dimensional space, that then will be the new kinds of sort of established models that people can actually emulate. So yeah, truly exciting stuff.
Will Jarvis 28:34
That’s awesome. Are there any that are particularly interesting that you think? I don’t know? I don’t need to pick favorites.
Unknown Speaker 28:41
Yeah, I’m not sure. I want to, I don’t want to pick favorites. I think that like certain organizational models, like revenue models are kind of interesting, like whether it’s like, trying to have like certain like corporate sponsorships, or kind of, or trade creating certain types of like, reports that people can subscribe to, I definitely think that it’s very interesting to see some of these organizations that are kind of trying to blend both research as well sometimes like education. And so there’s like these interesting organizations that are doing residency or education model, like or having an educational component or having research components. And really, I would say, as long as they’re kind of doing a non obvious combination of things, I think that’s really interesting. So if it’s like trying to act as like, like, basically acting as a startup and building a product, but also actually incorporated as a nonprofit, that’s interesting. If they’re a for profit, but they’re like, they really don’t have a product and they’re just kind of doing a lot of research that’s also really interesting. And so, so I just think I just need to like non obvious combinations within this high dimensional space,
Will Jarvis 29:52
but fascinating. That’s awesome. I love that. That reminds me of, you know, used to be Xerox PARC Bell Labs, you know, big names. And the big tech companies do have research labs like this. I think Google does Facebook does. I’m not sure about the others. But it seems like they’re not producing the same. And maybe they are. We just don’t hear about it. But it does it. You know, I hear Bell Labs, it hears Eric’s part. I’m like, oh, man, like, you know, I know, they produce some really important stuff. But the big tech companies today, that seems to be less the case, do you have a general feeling? Or do you think that’s just kind of misplaced, just a PR thing?
Unknown Speaker 30:26
So I definitely, I mean, Xerox PARC, and Bell Labs, like they made a lot of really interesting innovations. And the truth is one of the things I wonder about when people kind of think about these kinds of things, is, like, if there’s a certain amount of like survivorship bias or hindsight bias, we’re at right for all and I imagine there are a lot of research labs in like, the mid 20th century. And I don’t know to what degree I like, I’m just not aware of many of them. So it could be right only hear about the successful ones. That being said, I mean, the reason I Belen Xerox were able to kind of do so well is because they they had these like pseudo monopolies, or right after monopolies. And so they had a lot more flexibility. And so, that being said, I mean, Google and Facebook are pseudo monopolies as well. It’s like, yeah, so maybe they could, they could do that kind of thing. And so I think, um, and there are, I’ve also read a lot of analysis, that kind of things are like, what distinguishes Xerox PARC and Bell Labs from the other kind of things from these other kind of more modern organizations? And so I would say two things. One is, I do think some interesting research is coming out of the current labs, it might take a certain amount of time to kind of recognize all those kinds of things. Sometimes it also could be that some of it is being productize in a way that we don’t necessarily associate with research. Like, for example, like, I’m pretty sure the new like newest version of like Google’s translate function, came out of like Google brain and like all their research arm. And, and that’s it, it’s unbelievably good. But we kind of think of it as like a product. But like, it’s, it’s unbelievable. I like when I think about, like the translation from like, I don’t know, like, was like the late 90s, early 2000s. There was a Babel fish.com. And it was terrible. And now it’s like, you can read newspapers and other languages and like, like, it’s not perfect, but like it’s extremely serviceable. So I’m not sure we should like undercounts some of what’s what’s happening, but I but I do think there are kind of interesting institutional structure components around like labs and charts. And so like, certainly Xerox PARC, like one of the things that I think people, maybe sometimes like, under appreciate is not just like, the collection of raw talent, and just kind of like the brains that are there. But also the people who almost like ran interference between the lab and the corporate structure. And so like the people who are willing to say, okay, like, I’m going to run interference and let you do your thing, and kind of protect you from all the corporate bureaucracy and other kind of other kinds of things. Those kinds of people I think, are really, really important. And I imagine to a certain degree, those kinds of people exists, like the large corporate industry labs up today. But But I definitely think being very deliberate about that kind of thing is extremely important. And then something we we sometimes like, like, Don’t focus as much attention on.
Will Jarvis 33:18
Gotcha. So maybe protecting and like, building this little enclave where, you know, you’re safe from all the paperwork, and everything’s really important, and just having freedom.
Unknown Speaker 33:26
Yeah, having the freedom and sometimes even just like the time because I think a lot of things with like, in with research, research just takes time. And, and it’s very hard to kind of say, okay, like, I like I’m doing research, I don’t really know what’s gonna come out of it, it’s gonna be really interesting. And I also will have you stopped that, like, I’ll have something to you in like a year, or even six months, it’s like, there’s a slow burn, like, it takes time. And so, and I think being able to kind of protect Group, a group of researchers for a certain amount of time is also very important that and that being said, I mean, a lot of these organizations like the ones, even the successful ones, like they had, like, a certain sweet spot in time when they were doing a lot of things and, and then, and like, some of them are kind of still around to varying degrees. But their minds are churning out the same kinds of, like, Nobel Prize winning research or things like that. And so yeah, it’s Yeah, this is a long way of saying, I think there’s a lot of factors. Some of its could be some of it could be just simply like the individuals who are being aggregated, it’s the, the institutional structures around them that is potentially giving them the freedom to do other things. And some of it can also just be survivorship bias, like we kind of just only look at the successes and forget that. And the truth is, I don’t know, there could be it could be that during this like mid 20th century period or whatever. There weren’t that many research groups. And so the fact that we only hear about the successful ones is that we only actually hear about all of them, but it could also be that we just hear only about the successful ones and there were a lot of ones that were just less successful. And and I think it’s kind of interesting, and I feel like I should have better data about that. But um, but yeah, I’m not sure though. That’s all Yeah, it’s
Will Jarvis 35:05
a complicated question. No doubt. If you read Don braven scientific freedom by any chance and his concept of habit, yeah, I had a chance to go through it. It’s really interesting. That’s awesome. Yeah. So we had Don on the show a couple of months ago. Really awesome guy. Yeah, he’s like, 85 he’s still kicking. It’s really cool, dude. How important do you think scientific freedom is, is it you know, it Don talks about it as it’s like, you know, it’s it’s like, you know, the air we breathe, it’s like really important to give people flexibility to pursue, you know, the wacky ideas they’re interested in. And I tend to agree with them. You know, just think about Max Planck 20 years to thermodynamics. It feels like it would be difficult for that to happen today. What do you think about that?
Unknown Speaker 35:47
So I think, definitely, freedom is important. Like having a certainly, like not having as many constraints is certainly useful. At the same time, though, I still think probably some constraints, like mild constraints are actually useful in terms of like, causing people to get more done. And my sense is, like, if you look at like, like an organization, like the Institute for Advanced Study in Princeton, young, that they’re there, I guess, there was a burst of really interesting work in like, the late 40s, around like the development of like, the modern computing infrastructure. But by and large, a lot of people who end up being there, like long term, yeah, they’ve often had, like, kind of like their best work behind them. Not necessarily always, but like, but but it’s often like a great place to kind of like, just hang out and discuss interesting things. And, and I wonder, and I guess this could hypothesis that could be tested. And I wonder if it’s because they’re almost there are too few constraints, and kind of too few pressures that it’s suddenly just like, you don’t have to feel like you have to kind of generate any thing. Now, that being said, I think probably a lot of the best scientists have a certain amount of like, like, internal drive, and so whether or not they have constraints, they’re still gonna get things done. And so I’m not kind of arguing both sides of this. But but I do think scientific freedom, certainly in kind of the world we are in were, especially in like, in kind of, like, within universities, where there’s certain incentives and certain pressures on publishing and things like that. And, and grant cycles, I definitely think scientific, like, we need a lot more scientific freedom. So I think we’re nowhere near the point where scientific freedom is going to be a problem. But I do think like, if we were in that wonderful world, where everyone has huge amounts of sign up, then we might want to be a little bit more deliberate about how we think about it. Because it could eventually be an issue as well, kind of on the other extreme, but I definitely do agree that we need to create more opportunity for scientific freedom, whether or not it’s through grant granting agencies, how we think about structuring universities, things like that, I think there’s a lot of different ways in which we can kind of generate greater sense of
Will Jarvis 37:55
freedom. Gotcha. So probably important, but at the end of the day, incentives do matter. Yes. That’s cool. Love it. I wanted to kind of take a left turn a little bit here. So, you know, in general, how skeptical should we be of, you know, really easy gains in like, our, in biology, especially like, so no tropics and, and things that make us smarter? You know, it seems like we’re already pretty well optimized, you know, in general, like, is that just a bad place to be looking for, you know, advances?
Unknown Speaker 38:25
So I think there’s nothing wrong, necessarily with looking like, and yeah, it would be great if we kind of, like find some something that, like, helps us like, never have to sleep or, like overclocks our bodies and our brains. But I do think I mean, it’s a less that we’re kind of like, an AI. Not that humans are in or biological organisms are like, extremely optimized things, but like, we’re like, we have been optimized over long stretches of evolutionary time. And oftentimes, that involves, like, creating, like, optimization between extremes, basically, like managing a lot of trade offs. And so and and you can see this with and, and and we’re, we’re really more like in work like these, like, messy collages that have like, evolved over long evolutionary time, optimizing trade offs dealing with things and were enormously complicated. And so the idea would be so so the argument that, like, you can find something, they can immediately kind of cut through complex metabolic pathways. And will, like, allow us to kind of change or be improved really easily, I think, is I know, it’s exciting, but I think it also betrays a certain amount of almost like engineering naivete where like, it’s like, like, oh, like, like, because we have these engineered systems in technology that we can kind of understand or overclock or tweak then therefore, humans which are systems as well. They can also be engineering modified and like but we are, we are, like we’re not engineers. I think there’s a there’s Um, there’s some trees in my backyard that actually have These like giant like dangerous looking spikes on them. And the reason they’re they’re the center of the evolutionary argument is that they’re there because they’re there to protect the tree against like the fruit or the seeds being eaten by all the megafauna of North America, which, of course, are extinct like these, these, like most of these giant sloths around these things don’t exist anymore. They’ve been extinct for 10,000 years plus, but this tree still has them. And so I think, interesting to argue that like, we are these like, well honed machines that maybe we just kind of need to be like, tweaked a little bit more. Like we’re not, we’re messes and so so I mean, I think it’s, um, I think we can we can make improvements, and we should certainly try. But it all needs to be done with a certain amount of like, humility in the face of complexity, rather than a hubris of like, oh, like, we understand how to engineer systems, like you don’t know how to understand like, you don’t know how to engineer systems, or like modify systems that have been, like subject to constraints over like, millions of years have like, insane numbers of like, interconnecting parts, like this is something that I think a lot of like, people from the engineering world who are enamored of nootropics can sometimes forget, like, biology is like, orders of magnitude more complex than technology.
Will Jarvis 41:14
I love that. I love that. Yeah, it sounds like unless you like, there’s some like, easy constraint, you could see like, maybe, you know, there’s this drug you take, it makes you burn quite a few more calories. And then it’s like, well, that makes sense that that might be a trade off, you
Unknown Speaker 41:27
know, right. Yeah. Right. It could work. But it’s like you also have like, it could be trade offs, like you have to. And this is, this is like when I think about like, was it like Michael Pollan’s dictum of what is it, like, eat food, mostly plants, not too much. And he’s like, okay, like, this is like, this is what he kind of describes what all these different parts mean. But it’s basically the argument of like, we have these traditional diets that have served us well for millennia, or maybe not quite millennia, but for a long time and like to mess with them and kind of argue that we’re going to kind of break things down into individual nutrient components. I don’t think that’s a bad idea. I definitely, I am of the belief that like, we should obviously be learning more and more about our own biology, right. My initial training isn’t like evolutionary biology, like these things are great. But to immediately say that, like that we can kind of like just want like, like, wipe away the kind of the slow, incremental tinkering approach that we’ve had for a long time. As just like, kind of like, like old wives tales. I think it betrays a certain amount of hubris. Rather than saying, like, we can learn from it, maybe we can do better. But like, but I think there’s something there and kind of recognizing, we need to have a certain amount of humility here.
Will Jarvis 42:35
That makes sense like that. Are you down for a route of overrated? underrated? Sure, let’s do it. Let’s do it. So complexity, overrated, underrated.
Unknown Speaker 42:44
So I, at this point, you probably can tell my obsession with I would say it’s probably underrated. Mainly a sense that I think when people hear the word, they like, often just think it’s like, kind of this like, thing that’s bad or something to avoid. Versus like, for me, I think it’s just like, this is the world we’re in. And we need to like, we need to recognize that I actually think it’s underrated and sensitive. I think people all almost don’t even always realize that we are living in a world of sufficient complexity, whether or not through these nootropics as well as and even just the technologies we build. So um, I remember when, like, around the time, like the Apple Watch came out in the Wall Street Journal, a few months later had this article and like their style section around, like whether or not like mechanical watches are still going to be a thing whether or not people are going to start buying smartwatches. And so they there’s a quote by this one person they interviewed who was saying like, of course, I want a mechanical watch, like when I think about the mechanical watch and all of its complexity as opposed like a smartwatch, which is just a check. And I’m like, Yeah, I get like that mechanical watches are complex, but like, just kept like, these things are like, so much more complex. But you’ve been shielded from that complexity. And I think oftentimes, when we think about complex, like, we’ve just been, we’ve been shielded from it. And we, we don’t realize how much is out there. And then also related to that, I think we just mean, like, yeah, it’s wonderful. Like, we desire simplicity. And like, that’s not a bad thing. But I think, actually, I think that when the world is more complex, it also just makes the world more interesting. And so yeah, so I would say, probably a little little bit underrated.
Will Jarvis 44:12
Definitely. Yeah. And the example you gave, you know, apples, one of their unique advantages seems to be their ability to make things seem like they’re, if they’re very simple, you know, and it there’s unhide, all these complex processes behind what
Unknown Speaker 44:24
you do, you often don’t even realize how complex they are until they fail. And they’re like, oh, there’s there’s a lot going on under the hood here. Which is, which is not bad. But but but I think, um, but I think we need to have a little bit better sense of the vast complexity of like the technologies and just the world that we’re in. Because, yeah, in a world where sometimes even the experts don’t fully understand the technologies they built, they can’t grapple with just a massive number of interconnected parts. it’s incumbent upon each of us to a certain degree to kind of just be aware of the complexity around us and maybe have ways of like peeking under the hood and computer like throwing up a terminal every now and then doing some like command line stuff. Like having that ability to see what’s, what’s in there and how complex things are. Think can actually be a good thing or a bad thing.
Will Jarvis 45:06
That’s awesome. specialization, overrated, underrated,
Unknown Speaker 45:10
I would say overrated. Um, I mean, people in specialization is very important. Don’t get me wrong, I think I need a specialization is something we need. And we’ve been able to make a lot of advances. And people who are experts in this very narrow field have allowed us to actually learn a lot about the world. And I think, but but related to that, though, is when we specialize too much, we lose the ability to kind of see connections between different domains. And those and or whether or not it’s like, overcoming jargon barriers and finding things that have been like rediscovered, like been discovered and some other areas that we need people who can jump across domains, run that kind of import export business of ideas, as well as kind of connect, like finding connections and saying, like, Oh, these two fields should actually be they can be productively connected. I definitely I, I think I mean, specialization is obviously very important. But but one of the other things is that like, as we’ve started doing larger and larger projects, whether it’s technological projects, scientific projects, that the only way to do them is to kind of combine lots of different specialists together, which means though, as a result, no one individually has a sense of what’s going on in some large product or some scientific endeavor. And, and that’s not great. Like, if we kind of understand things, but only partially, that’s, that’s not good. Which is not to say that Oh, therefore, like a generalist, and understand everything, but I think we should strive to and whether it’s like the T shaped individual where people can kind of cut across different domains and maybe have some have some specialization in one domain. Or simply just being comfortable in like, learning about new domains. I think that is something that is that is really important. And and then it could also just give you like a competitive advantage in the world with like, you can kind of understand multiple different domains and no one else can that’s, that’s great. Like, you can do things maybe other people can’t. So I would say a specialization by a little bit overrated.
Will Jarvis 46:58
Was it 1980s computer magazines?
Unknown Speaker 47:02
So, so old magazine, old computer magazines are fantastic. I think they’re, I think they’re, they’re underrated, not just for like the nostalgia component. I certainly like looking at old computer magazines as interesting from nostalgic one. But I actually think they’re really interesting source of like, like working raw material for innovation. I think a lot of people kind of in the tech world and kind of like the whole, like Silicon Valley Community. They’re often pretty ignorant of technological history, sometimes, like, probably ignorant, I would say that’s probably more the exception than the rule. But most people are just not aware of like, the deep history of what’s happened in technological history. And, and we often because of that, in the same way that sometimes we can cut across jargon barriers, we end up reinventing things, we often end up doing the same thing in technological history, like we end up like, like reinventing something that has been done maybe, like, decades before. And so being able to kind of just look at what has come before us, and and seemed like, what were the like, what were the tools people were trying, people were trying to develop? Or what were the problems people were trying to solve? Or were that the killer apps people were trying to make, like, these things that like, in general, the people the problems that people had, like, they’re not necessarily going away. I mean, anyway, when I think about, like, the current, like, no code, low code moment, and that’s not like a new thing. I mean, like and like, like, like, the old Macintosh is like the night like late 1980s. Like, they came with a computer program called hypercard. And like, that was basically like a proto like End User Programming, no code, low code interface. And like, so these things, were not an amateur, there’s a lot of people kind of work side by no code, no code, they often like us kind of hypercard as a touchstone. But like, it’s not like these things are not necessarily new. Like they’re kind of things come in cycles, there’s always like, Ecclesiastes, there’s nothing new under the sun or whatever it is, like recognizing that, like we can actually learn about, about what it’s like we’re learning about what are kind of the things that we can be doing and rather than just kind of constantly reinventing, actually build upon what came came before us, I think is really, really important. So yeah, 1980s computer magazines, they’re amazing and definitely underrated.
Will Jarvis 49:09
That’s, I love that and it sounds like it’s a it’s a great place to look because there are things that people could they could have been trying to do in the 80s that were just not possible due to technology constraints that now are possible and
Unknown Speaker 49:22
right and I remember this is maybe was like in the 90s but um, but like in a lot of computer magazines, there were just like tons of ads for like simple neural network software and all these kinds of things. And we do it in parallel parallel sort of like the development with an AI where was like it was hot for a little while and kind of became less hot and like it kind of hit ran into some some barriers and now of courses may be able to make a lot of advances, but But yeah, seeing just the fact that like there was it wasn’t just like research being done in academia like there were like, people were selling like consumer software, like there were advertisements for like a neural network software, in computer magazines. Like it was great. Yeah, that’s fantastic stuff.
Will Jarvis 50:00
polyphasic sleep overrated underrated.
Unknown Speaker 50:03
So, um, I guess I kind of go back and forth on this one, I would say probably it’s like, probably properly rated in the sense that like, most people ignore it and kind of think it’s kind of wonky and weird. Yeah. But then they’re like the people who kind of think of it like, but at the same time, like the people who are think it’s amazing are probably like overselling it. Um, I my sense is, I mean, I guess I would say one of the one of the times I kind of looked into it, I never actually experimented with it myself in terms of like changing my sleep habits. But when, when my son was born, he was slept horribly. And I was like, Oh, I wonder if there’s a way of kind of like, overclocking my sleep or kind of like, finding some, like, quick fix to my need for sleep. And so I looked into it, I mainly realized that, like, it would just be like, it would wreak havoc on like, my ability to, like, interact with my family, and like, like, do things like that. So So I eventually I ignored it. But I would say actually, one of the things that I think about it is, the main issue is almost it’s like a less whether it works and more about what it says with our like, cultural obsession with like, the nature of like being busy and like always doing things and and I actually think, like, the more now that I’ve actually, like, my son now sleeps very well, it’s been many years since he since he slept very, very poorly, I now can get sleep. But I also realize, like, I see the increasing importance. Like, rather than trying to kind of like overclock my sleep and get as much awake time as possible. There’s actually value and kind of going the other way. And like taking breaks, or and people making lots of arguments for like working shorter, like shorter work weeks, or shorter, shorter hours. These things are all, they allow us to kind of not only sometimes work more efficiently or work better, like there’s kind of like these, there’s the efficiency argument. Yeah. But they also give us a better sense of like, helping us figure out what is actually important and what we want from our lives. And so I think, and for me, polyphasic sleep sometimes is a way of almost like, ignoring, or kind of like, like pushing aside that entire deeper question of like, Okay, do I actually need to be as awake as I think I need to be? Can I prioritize certain things like keep room that I think are important? Can I cut back on my work? Do I have to? Can I realize that maybe I don’t need to get as much done as quickly. And so I would say I mean, it might very well work. And I’m sure there’s many people who swear by it and think it’s great. Right? But and so so in that sense, I don’t think it’s like, underrated. But I also think it’s overrated in the sense of like, we’re, it forces us to kind of, it prevents us from actually grappling with the things that are truly indulgent in like humanity.
Will Jarvis 52:34
Right, right. Yeah. No, it does. That’s a really good point. It’s almost frenetic, right. It’s like, Oh, God, like, maybe you should just sit down and think, well, what should I really be doing? What’s the highest value activity? I can contribute? A little bit? The efficient market hypothesis? overlay? overrated? underrated. Ah, so this?
Unknown Speaker 52:51
I’m not really sure. I, I guess I do think I’m probably most people like who are not like not like professional investors, certainly, like most people operate as if it were true. Because I think well, while there are inefficiencies in the market, like they’re really hard to find. And so like, you should kind of operate as if it is accurate. And so I don’t know if it’s like underrated or overrated. It’s probably like, it’s, um, it’s something good that people should incorporate into most of their investments rather than kind of being obsessed with like, trying to concentrate and beat the market or finding efficiencies. Because unless that is like, the majority of your job like, that’s right. It’s, it is a very tough thing to do.
Will Jarvis 53:32
Definitely. Yeah. I think that’s, it’s a great point. And I think it’s perhaps underrated in the sense, over, it’s underrated by most people investing in public markets. And it’s like, perhaps underrated, overrated in certain areas where, you know, people don’t have, you know, there’s this inadequate equilibria. And it’s
Unknown Speaker 53:52
like, depending on like, which community you’re talking to you.
Will Jarvis 53:56
Yeah, definitely. Love that. Well, Sam, thank you so much for coming on. Do you have any parting thoughts? And where should people go find your work?
Unknown Speaker 54:05
Well, first of all, thank you so much. This is fantastic. I had a blast talking about all these different things. In terms of parting thoughts, I don’t think I have any, anything particularly insightful to say I’m certainly I definitely I kind of like reiterate the whole thing of like, recognizing that like science is a very human process, I think is an important, maybe maybe underrated kind of thing to think about. And also kind of recognizing that complexity is around us, but we can actually kind of sometimes harness that or kind of appreciate it, rather than just trying to like avoid it. I think it’s really important in terms of like, how to find find me online. My website is Arbus min dotnet. I’m just my last name and then dotnet I have a newsletter ombudsman.substack.com but you can also go to our bisman iMac to subscribe to it. I have left kind of other social media. So that’s kind of my major way of interacting is either through the newsletter, where you can also kind of see like other places where I where I write periodically, I often link those and then my my way Epson has a whole bunch of like older writing and other other things that I’ve done as well.
Will Jarvis 55:04
Awesome. Well, thanks so much, Sam. We really appreciate it. Thank you very much. This is fantastic. Awesome.
Thanks for listening. We’ll be back next week with a new episode of narratives.
Transcribed by https://otter.ai
