Business, Innovation, and Managing Life (April 10, 2024)

Sa. Hello everyone. Welcome to another episode of Q and A about business innovation and managing life. And I see a bunch of questions saved up here. Gorgola asks, what advice do I have for young entrepreneurs? Can children be successful in business? Well, I've certainly known some children, I guess you could say mid teenagers, who've been quite successful in business. I think it creates a complicated life situation when people sort of start that young and are successful at things. I think it is a complicated thing if you start a business at age 14 or something and you're still running it for the rest of your life. The chance that you set up exactly the right business at age 14 that you can appropriately run it for the rest of your life is not high. Certainly even the first business I set up when I was what, 20, 20, 20, 21 years old, there are many things about it that were not sort of good for the long term, at least for me. And I'm glad that I got to start my current business when I was 26 years old, I think. And you know, I'd already kind of sort of corrected many of the things that I got wrong at the beginning. But certainly it's an interesting experience for kids to have kind of to be able to be entrepreneurs. The truth is there are certain kids who sort of have the energy and the tenacity and the act togetherness, so to speak, to be very young entrepreneurs, and a lot that don't. And the concept that people should sort of push kids to be an entrepreneur, have a generalized lemonade stand, I'm not sure how valuable that is. I mean, in other words, there are cases where kids just sort of want to do it. And if things can be set up so they can do it, they want to do it, they should do it. If it's kind of like, hey, this will be good for your resume. Everybody who applies to a fancy college should have started a company type thing. That seems to me kind of grim. And I'm not sure that that experience is, you know, by the time it doesn't really matter if the company is successful, for example, by the time the whole thing is kind of scripted in some way, I'm not quite sure how much one learns from that. I mean, if one really has an idea and wants to push it as and you know, that's the best reason to start a company is because you really want to make something happen rather than just I want to have a company kind of thing. And in my experience, if you want to make a product, for example, that you yourself want and that doesn't otherwise exist in the world. That's a really good basis for starting companies and things. I mean I have to say I have seen a lot of, I've seen some quite successful companies started by kids and I've seen a bunch where it's like hey, this was just needed for a college resume type thing. And it's like to me it's kind of eye rolling and not, not terribly useful. My, my comments, at least in terms of businesses, you know, I think that, that young entrepreneurs can pursue. It's. I mean the ones that are kind of realistic, I think are the ones to concentrate on. That is if you say I'm 15 years old and I'm going to start a business to cure cancer, for example, it's kind of unrealistic if you say I'm 15 years old and I'm going to start a business doing data science or writing AI prompts for small companies. That's realistic. You can do that. It's something you can do with your own hands and potentially eventually hire some other people to help and things like that. It's a business that can kind of start small. I think that's true of a bunch of sort of services oriented businesses. I suppose the other thing that isn't directly a business is things. All of the various blogs, social media, websites, all that kind of thing, which isn't quite the same thing as, as the. It's being done primarily to make money. It's being done because you want to put that thing out there in the world and if things work out really well, it might make some money, but it's kind of mostly a creative act rather than a, a commercial act, so to speak. I think. But I, I would say that the, the kind of, you know, things that are. The idea that. I would say that the more modest kinds of business ideas that are more kind of either you just put something out there creatively or you're doing sort of small scale services seem more realistic for most of the kind of the young entrepreneur crowd, with the exception of a very few who kind of really have the drive to make something bigger. That sounds very negative. I don't mean to be so negative, but I have to say that I do react negatively to some of these kind of just made for the resume type kind of young entrepreneur kinds of things. Let's see. Talking of resumes, Utai asks, is it good to have variety in my resume when applying for jobs? How valued are long term employees of one company versus an applicant who has had many different jobs the thing to understand about resumes and applying for jobs is companies are different. What may look really good to one company may look really terrible to another company and vice versa. So it's not a one size fits all kind of story. This whole question about resumes where you see lots of one year gigs or even four year gigs or something versus oh, the person was at one company for 25 years, they're different reactions to that. One thing that can happen if somebody was at one company for a very long time is like that's the only culture they know. They'll never adjust to another culture that is. And particularly if the size of the company that they were at for a long time is very different from the size of the company that's hiring them, that can be an issue. That's one thing that can happen again, in all of these things there's no absolutes. People who've been at one company for a long time that has a particular size and they move to a different company of a different size, they can do great. Well, they can do terribly. But the concern would be, oh, they just know this company culture of this size and so on and they'll never adjust to something different. And certainly, I mean, I've seen examples of people who've worked for very large companies for much of their career and then they say, and now I'm going to do a startup. And it's kind of like, well if you do the startup you kind of have to lick your own stamps type thing. And they say, oh yeah, there'll be no problem. And then when they actually get into it, they're like, oh, but where's the stamp licking department more or less? Oh well, you don't have one of those because you're just doing a startup type thing. And it usually does not end well, not always, but that's a type of issue. Now in terms of the resume that has lots of one year gigs, I mean for a company like ours, that's not particularly attractive because it's like, well, why were they one year gigs? Was it because the person has a concentration span of one year which can happen? Or was it people will say, well, I went from this place to that place to that place. Well, it's not usually for many kinds of positions. It's not a great bet to hire somebody if you think they're only going to be with you for a year. Because if you have particularly a strong company culture and you're with unique technology and so on, there's, you know, there's a Quite a long enculturation and training period. And you kind of, you know, one year is kind of not enough to pay back for that, so to speak. And so if the, you know, person is like a one year concentration span, it's like, well, you know, history is usually a good guide to the future. If people have, you know, done eight jobs that all lasted one year, it's pretty good bet that the ninth job is also going to last one year. And that may just not be something that's worth investing in now in terms of having that person sort of onboarded, trained, enculturated and so on. I think this question of sort of getting different kinds of experience, it depends, I have to say, it's kind of in the academic setting, for example, different area when people do PhDs and so on, that is actually a narrowing moment because it's like, well, you were learning about all these kinds of things and now you're going to concentrate for a few years on something very specific that you're writing a thesis about. Not all theses are about one topic, et cetera, et cetera, et cetera. There are exceptions to this. But I have to say one of the things that is the good news about writing a thesis or something is you got deep into something, you really figured out how to build a fairly serious tower of sort of content to put into that thesis. The bad news is it's kind of a narrowing thing. So I think there's considerable value in having stuck with something long enough to build that tower. There's a bunch of different things you learn by sort of sticking with something for a long time, which now against that there's well, do you know about lots of different kinds of things? I have to say that my experience has been that people, they're people who are very broadly interested in things, who know about a lot of kinds of things, and they're people who really want to focus on just one thing or a small number of things. And that seems to be a little bit independent of their direct experience. Unless it's in areas where it's like, okay, you're coming into Company X and you're going to do something which, yes, it's a hit, you know, bingo. It hits on something which was one of your previous, you know, three gigs ago, you did exactly the same kind of thing and that's exactly what the new company needs. It's a hit, you know, you know exactly what to do. You can just do it. In some cases that will be a good thing. For example, for our Company, I would say that's very, very rarely a good thing perhaps because our company is kind of weird in some sense and the number of things that we're doing that are like a great match for things that people have done before is not very large. They tend to be things that are sort of new and different and never done before. And so people can't have that specific experience. But I think it also depends, I mean in, in I, I'm, I guess on, on balance again it really depends on the company and the situation. I mean in, in, you know, there are obviously situations where the work of the company is a series of small projects and then having had experience with more projects is probably a win. There are cases where the work of the company is primarily building products over a long period of time, where the experience of having seen sort of the arc of something develop over a period of multiple years is very useful. Let's see, Traveling asks, how does one generate a succession plan for a company? Well, you know, I think depends on the company. Most tech companies are not really are sort of, that are backed by for example venture funds. They don't really have that long a time horizon. That succession is a big issue. It's kind of like the company's either going to sort of go to the moon in 10 years or it's going to die or it's going to get sold to another company or something like that. It's not something where there are many kinds of companies which are kind of long multi generational companies and so on. But the tech industry is not, is not very, does not have many of this. I mean our company is very unusual in being a long term private tech company. And you know, so, so we have issues like this but it's not clear that that many people have these kinds of issues. I think that the, you know, one, one feature of companies is it's, it's always nice when you have a diversity of ages of employees. For example. Mean there are companies I know where kind of like the company really got, you know, was successful at one point and everybody is in their 50s, let's say because that's the moment where sort of things, things advanced and, and everybody came in then and so on. So it certainly helps if you have a range of, of of ages and so on. And then it's sort of a question of, of of of making sure that the appropriate youngins, so to speak, learn what happens. And I really think it makes a difference the periods of time that people have been involved. Because if you're involved with a company for, I don't know, three years or something, you see, even if you're reasonably aware of sort of major executive decisions, you see a certain number of things happen. If you're involved for 10 years or more than that, you see many more things happen. And the thing that is necessary in being able to lead a company, particularly an existing company that has an existing culture and so on, is a lot of the, well, what happened before type thing. And, oh, we've seen something like that before. That's a really important thing to know. So I think it really helps to have people who've been around for a while, but I don't know, ask my successors for our company at some point in the future and we'll see what, what's happened there. Let's see. Wish says, do you think the software market is over saturated? Feels like there are many untapped innovations in areas like material science and hardware. You know, these things go up and down. I mean, in, you know, what tends to happen in every industry is there's some methodological or technological infrastructure advance that opens up a whole bunch of new possibilities like kind of LLMs and the AI things opening up a bunch of new possibilities, kind of enthusiasm about blockchain, enthusiasm about the cloud, enthusiasm about the web. All these things opened up a bunch of new opportunities. And at any given time there's, in the early years of those new opportunities, there's all kinds of energy, there's, and all kinds of things that can be done. And then after a while it kind of slows down. It's kind of like, yes, all the obvious things have been done and what remains is kind of hard work or reinvention or whatever else. And I think there was a period of time when software was kind of it and it's like, oh, it has much lower investment than doing anything with actual hardware and devices or whatever else. And then people in the last probably four, five years or so got more interested in sort of deep tech kinds of things that involve more kind of labs and investment in real things and so on. Partly because the software industry did get things got done that were kind of fluffy and were not. There wasn't a lot of content. It's like, I'm going to make a new social media platform. And sometimes people manage to understand enough about human behavior that they hit a jackpot. A lot of times it was just like, oh, another, I don't know, you know, I don't know, you know, a this for that, you know, a Facebook for cockatoos. A, you know, this for whatever it is a kind of a very much of a follow on thing. And in some cases it's like, you know, really, is this really going to work? Or you know, I remember, you know, what was it? A social media platform I think was called yo that just lets you send the sort of one bit of YO thing to people and you know, for a week it was popular and then people realized it wasn't that interesting. But then, you know, then along came things like snap and it's like, really, is this going to be interesting? Well, yes, it turned out to be. Or you know, many other of these things which kind of depending on the details of kind of what they precisely hit in terms of human behavior, what they hit in terms of the trend of the moment and how much money is getting spent to promote them. These are things that can occasionally be a hit. But I think people reacted to the fact that it all seemed rather fluffy and not very contentful. And then it's like, let's go for deep tech, let's go build nuclear reactors, let's build things that involve new material science or whatever else. All of these things are hard roads, but for different reasons. In the case of, but I think these cases where there's a certain minimum investment, because it is hardware, because it involves a lab, things like that. Of course, our friends at Emerald Cloud Lab have been trying to make sure that there isn't a need for the build a lab type investment even when you're doing kind of lab type stuff. But in general it's absent that kind of technology. If you're going to build something that involves doing biology experiments, chemistry experiments or something, you're investing more than you have to invest. If you're saying I'm just going to build a piece of software now, you know, even in the case of software, it's kind of like with the current ideas about how, you know, neural net training works for LLMs, that's again a, oh, it's going to cost a bunch of money to do this. But I think in this question of whether, you know, yes, there is a lot that can be done and I think what's happened in the hardware area is, well, perhaps a certain amount of opening up possibly because of more services companies that let one say, oh, I want to build this piece of hardware, go figure out the engineering. For me, that's become easier. I mean it's in the case of like there's things which haven't become particularly easy, like I want to build a new microprocessor it's still hard to do that. And it's. And maybe if that became. There was a period of time in the past when people said, oh, there are going to be, you know, silicon compilers and they will make it easy to just say, I want to make a new microprocessor, let me type in some code type thing. And that didn't quite develop that way. If that developed that way, then suddenly there would be the cost of kind of innovation in that area would go down and one would see kind of a flourishing of kinds of activities. But I think, yes, there are. Look, a general principle is at any given time when there are methodological advances, when things suddenly become accessible that weren't before, that creates lots of opportunity. The other thing to say is there's probably opportunity all over the place in things that are not terribly popular right now. It's the things where it's sort of popular and well worked out are the places where sort of the innovation is least interesting. The cases where it's just barren landscape and nobody's looked at that for years because everybody was like, oh, there's nothing interesting in that area. Those are potentially fertile as are the ones which are the sort of sudden moments of oh, something new came on the scene. Now we have a good augmented reality system that's finally going to take off. Maybe I'm not sure we do at this point, but let's say we did. Then there's, oh, there are all kinds of opportunities that, that really opens up and so that then becomes an interesting area. Let's see. Amal asks, what's the best way? This is relevant to what was just asked. What's the best way to get funding for a physics based R and D company? I'm starting one and need help. This person says, well, depends what an R and D company means. If you mean a company that is going to try and develop a product. Actually something completely crazy has happened in the US right now that I could rail on about, which is this R D amortization for tax purposes business, which actually right now makes it utterly unappealing to have an R and D oriented company in the US because usually if you R and D in, well, for many kinds of things, it's mostly about the money that gets spent is mostly paying people to do the R and D. And normally when you have a company and you pay taxes, if you make a profit and things, you pay taxes on the profit of your company, which is the revenue it makes minus the expenses it has. But because of some crazy tax maneuvering started in 2017. It is now the case that in the US if you spend money on R and D in a given year, that does not count as an expense in that year. So if you. It is something that is like buying a piece of equipment, it's amortized over five or more years. And so if you're like paying for R and D folk, you don't get that as a. It's as if you have a huge profit because all the money that you were paying to R D is actually a piece of what you have to allocate as profit, which you pay tax on, rather than something that is an expense to the company. It's a. For a company like ours, it's. That's a big deal. It's a, It's a kind of a mess. And it's like, really, should we actually be doing, you know, operating an R D company in the U.S. i don't know. Maybe we shouldn't. That's a current issue, which maybe it'll be fixed in three weeks, who knows? But people have been saying that for a while, and it's kind of one of these things where you can look up the bill in the Senate that's kind of advancing its way through. That's been where this particular provision about R and D amortizing R and D for tax purposes, or rather not amortizing R and D for tax purposes, is a thing that seems to be politically agreed. Yes, it's broken right now, but the way politics works, which I don't understand very well, gets bundled with a lot of other things and then doesn't happen. But the result is, well, by golly, it actually is very disadvantageous, in fact, very hard to make it work to do kind of R and D in, in that way where you're. Where you're actually reinvesting a large part of your revenue in R and D. Okay, that was arguably a vent about a current situation. But in terms of what does one mean by doing an R and D company? Sometimes one means doing R and D for other people as a kind of a services thing, and sometimes one means doing R and D the sake of building a product. And those are somewhat different things. I mean, in the for the sake of building a product thing, there's a question of how much money is it going to cost? You know, is it something that you get to do and get to the point where you have a serious prototype of your product as a, as a moonlighting operation? Or is it something where. Or do you have your own resources to put into it? Or is it something where you're going to have to get money from other people and in the get money from other people? There are a variety of mechanisms. I mean, there's venture capital investment and there are venture capital firms that say they invest in sort of deep tech kinds of things. I know a bunch of them. I think that there's a certain dynamic of what a venture capital firm is looking for in terms of the arc of sort of developing, development of a company. And if, if the payback isn't for 20 years, venture capital is just not going to do that because they're looking for a payback in a roughly 10 year time horizon, preferably 5. And I think in the other thing you can do is, you know, in the US for example, the government has lots of kinds of grants sbirs, small business innovation, something or other that many government agencies have these things. The problem with those things is that they tend to be rather specifically written for, you know, this sbir is for developing, I don't know what, some very specific thing. And so it's sometimes challenging to say, oh, can we fit the thing we really want to do in our company into the rubric for one of these kinds of things. And there's, and that's true in general with government programs is that there's a, you know, somebody set this program and if you, if the thing you want to do happens to be a thing that aligns with that program, well, then there's a chance. But if not, then it's like, well, we just don't do that. The other thing is sort of investment from strategic partners, companies that could use whatever it is that you're building. And that I think is a, is a good thing when it can be done, particularly when it can be done with multiple companies where you're not in a situation where you just have one sort of umbilical connection to one particular company where really the only arc of the story is and eventually if you're reasonably successful, that company will buy you unless that's what you want. But I think that the idea that you have investors who care about the thing you make, that's a good situation. It's a better situation in many ways than the investors being purely financial investors, like venture capital investors, for example, where their only responsibility is to make money back, so to speak, rather than to see your product actually prosper for the sake of actually being able to use it as a strategic investor. For example, Joni asks, what are the benefits to internships? I think internships have become increasingly useful and valuable when they're properly run. I mean, it's a very good way for you to get to know a company, for a company to get to know you. And an awful lot of people who do internships end up getting hired by the company that they did the internship with. It's also, or if there are multiple companies in an industry that are somehow similar, do an internship with one, you kind of know what it's like to be in another and maybe you get hired by that other one. But it's really, I think, a good way to sort of get to know things on both sides. Now it depends on how well the internship is run and sometimes there's certain title inflation as there usually is. And sometimes when people say internship it really is just a, oh, you're coming in for some little sort of series of lectures or something that's not really a, you're actually doing work in a company. But I think in the. And when it is a question of actually doing work in the company, it has to be a well designed internship so that the work that you can do in some limited period of time with a limited amount of experience and so on, although perhaps with lots of enthusiasm and ability, is something that can actually be useful to the company and is properly aligned with the company. I know in, in, oh, I don't know, years ago when we had interns that often wasn't done very successfully and we ended up with an awful lot of oh well, this project wouldn't be done but for the fact that we have an intern, let's get the intern to do it. The project gets 75% done by the intern and then it's the end of the summer or the end of the, whatever it is, the intern disappears and then what are we going to do with this project? Oh, somebody who's a main part of the company ends up finishing that project, costing the company a whole bunch. And it's just like. But we didn't really even really want to do this project in the first place. It was kind of a somewhat make work project created because, oh well, we have this intern, we've got to give them something to do. So that's sort of a bad way to set up internships. A good way is to have had the experience of knowing what can an intern achieve over a certain period of time and to be able to give them something that is sort of something sliced off in such a way that they can actually achieve it and it will get properly integrated back into what the company really does. I mean, I think it took some time at least for our company to learn how to do that. And that's always a question for internships. Is that really going to happen or is it going to be something where it's sort of just sort of hanging out there and so on? I mean one feature I think the mechanism most companies have for telling interns what happened with the work that they did when they were interns and is probably usually very, very crummy. Like you never know what happened and probably if you really probe, you can find out. But I think it's, or if you go work for that company later, you might very well be like oh yes, that intern project of yours that turned into this and that thing. But I think I would say that sort of genuine, well aligned internships are a good thing. And sometimes people will do internships and they'll go work in some company and they'll say I can't stand this, thank goodness I did this internship. Now I know I don't want to work for a consulting firm or whatever it is, or a software company or whatever. And so that's also useful. It's a kind of, I think a helpful thing on both sides as a way to get to know organizations. Let's see. Well, Emruda Osters. Is the AI development of current times just a short term fashion like in the 80s? I don't think so. I think that what has been done with neural nets in the last 10 years or more, these are part of a tool set that will be permanently useful. That's true for sort of machine learning, for image identification, you know, speech to text, lots of things that were sort of, oh, computers can't do that. And now you know, yes, you can assume computers can do that. Same with LLMs. I mean the question of will, will sort of. Will LLMs eat everything? No, I don't think so. I think that they are another very useful kind of tool component that you can build things from. And I think that's, that's kind of the way to, to think about that. And it's something that kind of saying one particular thing. Oh, you know, in the 80s people said everything's going to be expert systems. That was the particular kind of flavor of AI that people talked about at that time. What's going to happen is you're going to bring in an expert and they're going to encode all their knowledge in, in an AI and then you won't need that person anymore. It'll all be AIs all the way down didn't work out. You know, I, I think it's the ideas expert systems were particularly, I would say thin in terms of what survived from them. I think there are other sort of, you know, fads of the time where again, it's very thin. What survived. My own guess is what's there right now with machine learning and LLMs and so on. It's a nice thick component. It's not the thing that's going to eat the world in any sense, but it's, it's a component of what one can do. Like for example, oh, I don't know back in the day being able to do numerical linear algebra, doing things with matrices and so on. That became a thing in the 1970s and that sort of enabled computer graphics and eventually has enabled neural nets and so on. But it enabled a bunch of things. It's a tool that enables lots of things and so I think is a neural nets and machine learning and so on. It's just something that you build other things on. Zaden asks will owning and running private businesses ever be superseded as an economic form? It's an interesting question. In the past, most people had much smaller businesses than developed. I don't know, 100. Well, I mean there were, there were some big businesses, but they were very government aligned. And then there were lots of sort of small family businesses. The idea of these kind of in, in you know, investment inflated businesses that kind of, you know, are big from, from very early on is a more recent construct I think that aligns a bit with things like venture capital and so on, which is a story of the, well, basically 1980s, really 70s 80s was, was when that really got started in serious, in serious form. And I think this, you know, the, the rapidly inflated big company is, is not, was not a common thing. And I think that, you know, you see different sort of arcs of how companies work from the. It's really mostly a family business or it's very much sort of a, a branch of the government more or less to sort of the kind of financial engineering oriented businesses where it's like buy up a lot of smaller companies and things and make it a sort of financial story of can you make all the loans and this and that and the other fit together. That became another kind of business. I think that the idea of 20, 30 years ago kind of the arc of the story tended to be you get venture capital, you puff the business up for a bunch of years, then you take it public. That was kind of what was. And then eventually it either goes to the moon or it gets bought and usually it gets bought by a bigger fish, so to speak. And that was, you know, had been the longtime story and sort of going public was the holy grail. In more recent times, you know, it got more complicated being a public company and more expensive, etc. Etc. Etc. It's like why are we doing this? Why not have a private company? Because most of the investors you really care about are very large investors. You don't really care about the small retail investors. It's going to be sort of a, you know, exchange by titans type thing for stock in the company. And so that became more of a, more of a trend, I think that there's a small trend right now of people doing things more like what our company has been, which is a closely held private business for a long period of time. There are several, I would say good and prominent tech companies now where I know their founders and kind of their plan at least is to do something similar to what I've tried to do, which is to have kind of long term private business. Now in terms of is there going to be some alternative to what does it look like? I suppose you could say that some aspects of the open source software world are an alternative to the idea of companies doing things. In the end that's all a bit of a fake because in the end in most cases somebody is paying people to make the software. It's probably, it might be a big company that says, well yes, we don't care about putting the software out to the public. We're not going to make money from the software as such. I mean this started very early with things like unix, Bell Labs, where the telephone system was the telephone system and the fact that, and they were making all their money from selling telephone services and the fact that a piece of software was produced. It's like, well that's just, might as well just put that out there in the world. It might be good for some promotional purposes. It's like whatever doesn't really matter. And these things that are, I mean, you know, for a company like ours, there are things that we're putting out there that are sort of educational content and things where for us it's like, well it's good for the world. It, you know, helps people understand more about our technology. That's not where we are trying to make our living, so to speak. Whereas somebody else would say, oh my gosh, we're a, you know, textbook publisher or something. That's where we make our living. It's shocking that you're giving that stuff away but, and oh, as a textbook publisher, I'll just give the software away. And it's like, but for us that's how we're making our living, so to speak. So I think the kind of the idea that somehow things magically get produced within the end, nobody paying for it. I just don't think that's a realistic idea. And usually when you sort of follow the money, there's some kind of thing, oh, it's a thing where we'll give away the software but of course it has patents associated with it, but which will end up coming after you if you actually use the software. Or we give away the software but the version that we give away doesn't really work and you have to buy our services to get it to actually work. Or a whole bunch of different things. But. So one model would just be that people out of the goodness of their heart or for reasons of sort of personal self promotion or something, say I'm going to make a bunch of stuff and just give it to the world. That would be a possible model. I don't think that's a model that seems to have been. It has some small areas of success but on a broader scale. It's something where in the end there are usually deep pockets, often vested interests to having things be produced. I mean one of the things that's sort of an exercise for people is like the, well in the case of their countries now that are like, let's put up an LLM and give it away. And it's like, why would a country do that? Well, you know, if you're a country that wants to have a point of view in the world, that's a great way to advertise yourself is have the LLM, you know, know the, you know, the, the, the opinions of that country. And, and when, you know, when things come up, it's like, okay, we'll just, you know, speak it. The, the way that's, that says, you know, country X is great type thing. It's kind of like when people buy tourism advertising for a country. It's a somewhat deeper version of that and maybe it's good, maybe it's, maybe the world is prepared to go for the bargain that let's get cheaper LLMs or something in return for having a certain amount of promote country X type stuff. I'm not sure. But you know, these different models of these things are. I mean, I suppose the other thing to say about companies and people and so on is there are industries in which sort of the entity that is making something happen is much more loosely associated with people. I mean, movies, for example. A lot of the people who make a movie are kind of wheeled in for a week or two to do just this particular part and then off they go. And the actual sort of core organization is quite small. And it's a lot of kind of sort of piecework kinds of things that are done. And certainly as an increasing. It's been something of a trend in companies sometimes to say we're just going to be the little tiny piece that is sort of making the big decisions and everything else is kind of outsourced in some way. I don't happen to think that's a terrifically good idea, but that's a thing that is a slight change in the way one thinks about companies. I mean, another slightly perhaps cynical point of view is that there's sort of an increasing degree of abstraction in the world. It's like at the beginning everybody was growing their own food type thing. Then there's sort of the economic development to the point where there are centralized food companies and people are doing what they do for a living and then they're buying food from other people who actually grow the food and so on, and it becomes more abstracted and sort of what one does to contribute economically can be sort of further and further away from growing your own food type thing. And I suppose the increasing elaboration of the financial world is another step in that direction where you're kind of abstractly trading derivatives in some high frequency thing and you're doing this and you're sort of leveraging this and that and doing some complicated sort of piece of financial engineering which is far away from the actual content. It's kind of like, well, I'm doing some quantitative trading and these ticker symbols, I don't know what these companies actually do. I have no idea whether they make, you know, bricks or potato chips. It's just a ticker symbol to me type thing. And I know it's volatility and I know this and that and the other. And in a sense that's a new level of abstraction in the world. And it is certainly something which I think is very bizarre. And I don't fully. I'm not sure I know that the full import of it as more and more economic activity becomes more and more abstracted and more and more sort of money is getting made in the machinations of that financial system rather than in doing things that seem to one to be more immediately of human use, so to speak. But I don't know how that. And one of the features of that, those sort of financial machinations is they tend to involve rather few people. It's not the case that you'll have some giant company of, of tens of thousands of people figuring out how to do this particular piece of quantitative trading. It might be 100 people or something who are just sort of using their smarts to leverage huge amounts of money and so on. So to this question of what, I don't know, people in the time of, in the sort of, well, the last heyday of blockchain, the previous heyday of blockchain, I should say that was kind of a big thing, was can we invent other kinds of governance structures and ways that companies can operate and kind of notions of somehow connecting sort of the people who want to buy something directly to the producers of the thing and there's no kind of company in between. But, but then there always ends up being just for, I mean, as soon as you want to scale things up at any serious level, I think you end up with these structures that are basically just like companies. Let's see. Harry comments. I think in, in response to previous remark of mine, I think the best advice for young entrepreneurs is don't do it. If they do it anyway, then there actually should be entrepreneurs. Yeah, maybe. You know, sometimes people get put off by being told don't do it. They really get put off and they actually would have done okay. But I agree that there's a certain, There is a certain. If I look at my own life, for example, the number of things that were really good ideas that people told me were really good ideas was very small. Mostly people told me it was a very bad idea. Although I suppose the only thing that I've benefited from is having worked with some of the same people for a long time after they've seen things which they said were bad ideas but turned out to be good ideas, they are slightly resetting their kind of is it a good idea when they hear about a new idea? Let's see. Okay, Brash is asking, how would you, what would you say to individuals who are foremost, mostly interested in making a difference in the world rather than monetary incentive? How would you weigh the choices between pursuing traditional academia, working on the, on the cutting edge with the private sector, or pursuing research in one's own time, independently? Actually, I think I probably talked about this before, so maybe not. Let me not say too much about that right now, but this, you know, how do you make a difference in the world? Well, you Know, you can write exactly the right thing and it gets out there in the world and it sort of picks up steam. And everybody's heard this, this one sentence that you said or whatever, you know, you can, that things like that can happen. I think that many things in the world ultimately require a certain degree of scale, which typically requires a certain amount of money. And the extent to which you can. I mean, as a practical matter, if you're always asking other people for money, then you get into this situation where any new thing you think of doing, everybody's going to say, oh, don't do that, and you won't be able to do it. And so you do have to have a certain amount of ability to do things financially yourself. I think. I think that there are people for whom the motivation is just make a lot of money for the sake of making a lot of money. As a matter of kind of as a sport, who can make the most money type thing, which I'm not going to criticize. People have different life goals and that's a possible life goal. And I don't think it's in any sense any less honorable than a goal of like, I want to run the marathon in the shortest time type thing. It's a kind of sporting value type type idea. Another thing is, you know, I want to make a lot of money because then it will allow me to do xyz. Sometimes that doesn't end well because sometimes once people actually get to do xyz, retire, you know, give money to such and such a cause, whatever they discover once they're in the thick of it, that that wasn't actually such a good idea. But it's really hard to discover that before the event, before you're actually in that position. But I think in terms of sort of where can you have an effect on the world? I think it varies a bit over time because it's. Where is there sort of a distribution channel to get out the things that you can contribute to the world? I mean, that distribution channel might be you're writing things, it might be you're producing products, it might be you're even writing academic papers or something. It's the different distribution channels and it depends a bit on the ideas and it depends on the moment in time. I mean, for example, I don't think academic papers are in most areas the most immediately high impact way to get sort of majorly new things out there. They're primarily oriented towards sort of incremental progress. And occasionally that incremental progress hits a jackpot of Some kind. But mostly it's incremental progress. And the systems in place are about delivering academic papers that make incremental progress. And I'm afraid a lot of these heavily institutionalized areas, which now includes a lot of startup stuff that works through venture capital and so on, is also, it's quite structured. It's kind of like if you have a business plan that fits this model, well, yes, then you can get venture capital. If you're doing something, oh, it's really different. You know, occasionally some senior person in that area will say, hey, this is really cool, I'm going to support it. But if you're going through the kind of rank and file machinery, it's hard for it to get, get traction there. Let's see, Question here. Memes is asking about how about nonprofit R and D operations? You know, I'm, I've got one of these. We have a Wolman foundation and we have our Wolfman Institute, which are examples of this that are still in sort of fairly early phases. It's complicated. I think that if you want that to work, it has to have very clear leadership. It has to be there's a person who's trying to drive that or there's a very, very clear mission that's being driven. If it's just like, hey, we got this thing and it's a nonprofit and it's going to do R and D and it's going to be great, it's mostly going to dissipate its efforts. I've seen that many times. Actually. It's something where if there's a particular person who's kind of pulling it in some direction, that may be a direction that is not immediately commercially important. It may be something where, yes, it's going to be very important in the world over the next 300 years, but there's no immediate sort of payback this year, this five years. I'm going to start talking about R and D amortization for taxes in a moment if I'm not careful here, because that's exactly. A completely crazy disincentive to do R and D that will have sort of future value and so on. But any case, the, the. I think, Yes, I, I, you know, I, I think the management of nonprofit organizations may be even more difficult than the management of for profit organizations. Because at least in a for profit organization there is, I mean, at some level, well, at some level, both organizations, there's. How are you going to make a living for a for profit organization? It's like, well, we're going to sell a product to customers. I always like situations where the money you make comes from value you add to some set of customers rather than the money you make comes from some third party who is trying to get the data from the customers who are getting the this and that and the other. I mean, I always believe that it is a healthier business when it is a more direct link between the money you make and the products you make, so to speak. But I think in the case of a nonprofit, the way it's making a living is by raising money typically. And that's yet a different kind of strange, sometimes rather misaligned set of objectives. I mean, again, the most directly aligned is probably you make money by selling a product to people who want the product and find it valuable. In the case of a nonprofit, if you're lucky, there are people who say, I really care about this mission, I want to see it succeed. Here's money to help you make it succeed. But if you're not so lucky, the incentives are bizarrely sort of misaligned. And it's kind of like, well, I want to contribute to this because I want to be part of this club and I want to do this and I want to do that. And it's kind of like it's all a bit at odds, so to speak. So you know, I think that that has, that has its issue too. But in terms of the pure management of the organization, I mean, my own feeling is it's a little bit more difficult to manage, well, a non profit effort than a for profit effort and has a higher chance of just sort of spiraling off into doing things that don't make sense. Because in the for profit case, if you do that for a while, nobody's buying your product and you're kind of toast. The. Let's see, There's a question here, and I should probably wrap up in a moment here, but Junior asks, how much do you think businesses will have to adjust to account for AI workers if they take off? Well, okay, let me tell you one non secret secret about my life in business, so to speak, which is I've tried to build a company where we are in the business of automating things. And among the things we've tried to automate are our internal operations and our internal development efforts and so on. So for the last 35 years we've basically been sort of recursively automating things and by the way, providing that automation as tools that other people can use too, but for our own internal purposes, progressively automating things. So, you know, I would like to think that the productivity we have is the equivalent of a company vastly larger than ours, ten more times larger. And that's enabled by the fact that we've automated a lot of what we have to do. And I think that's. So that's in a sense a model for what does it look like when AI is in the picture? What it looks like is the people that you have get to be the sort of the puppeteers of that automation, but they are no longer as much in the trenches doing things because that's been automated. And I think that's the kind of, sort of, that's something you can just do more if you've automated more things now in terms of how that's managed internally within companies and how people get to sort of be the ones being the puppeteers of the AIs rather than doing the work that the AIs take over. That's an interesting question. I tend to think that it's really a management issue of understanding your human resources and what people are able to do. I mean, it's if somebody did something for a long time that is now automatable, what do they do next? Do you lay them off because you don't need them anymore? Or do you say, hey, wait a minute, this is a person who has these skills which we haven't really made use of because the infrastructure didn't exist to make use of those skills. But now that the infrastructure does exist, that person can be highly valuable and useful in this slightly different role. So I think that's, you know, in a sense that's the challenge is to figure out, given that new things are possible, what can you do with people that you have, for example, and where they don't have to do what they were doing before, they don't have to, you know, dig the trenches, so to speak. They can be sort of operating the bulldozer kind of thing. It's, you know, and to what extent do you take the skills that, as I say, weren't really exercised previously because they had to actually dig the trench themselves and let them use those things in a sort of more leveraged way? I mean, I tend to think that well managed companies, that's what's going to happen in other cases. People will say, oh, we don't need all these people because their jobs are taken over by automation and we'll just get automation specialists. That doesn't make a lot of sense to me. I mean, the number of people who just need to know about the Automation itself is actually small compared to the number of people who need to know what the automation is good for, so to speak, and what the ultimate sort of things that should be done are. Let's see, Jack asks about one person corporations and zero person corporations. I mean there are plenty of one person corporations. They're people who are, you know, sort of content creators, kind of consultants and so on, who are basically just sort of with their own hands creating whatever it is that's being created. I mean that's, that's not a particularly uncommon thing. And where sort of everything that they need, whether it's accounting or whether it's, I don't know, website building, whatever else, that's all contracted out. So one person corporations, pretty common. The zero person corporations. Well that's not here yet. I mean I've always sort of wondered about the AI that kind of owns its own corporation by some kind of circular chain of LLCs or something. And where the AI is making a living from entertaining people or whatever and using that to send payments to its sort of compute providers or whatever else. I mean that's a very weird situation that we haven't seen yet, I think, where it's really just a piece of sort of separated automation. I mean, of course there are zero person companies even right now where it's just a question of how certain resources are being bundled up. And it's like there's this company and it owns these things and it has these responsibilities and so on, but there's nobody who actually works for that company. It's kind of a, well, you know, you might call it unkindly a shell company, but sometimes it has a, a particular purpose in terms of, I don't know, it could be operating in some particular locality or it could be doing something else that requires it to sort of be a separate company even though it needs zero employees. All right, the last thing here, very different topic. Peter asks. Seems that public libraries should offer access to journals as well as books. Thoughts? I know some of them do. I mean even some states in the US had these whole deals to get sort of access to, for all library holders to things like academic journals. I, I'm not sure. I don't really know the status of academic journals in the modern world. I mean there are sort of free preprint versions of in, in many fields there are free, free print versions of everything and the, oh, it got in a journal and got, got stamped as being included in a journal is not really the most important thing. That's more true in the Kind of physical, mathematical type sciences. Also the. Nowadays in computer type sciences, more so I think, than the biological sciences, which have had more of a tradition of sort of actual sort of medical journal publication type things. I don't think there's as much of a culture of sort of the preprint type world. I mean, I have to say, I think one of the things that's sort of a weird, weird feature is the whole machinery of journal publishing and academic journal publishing. It's a very strange business because in a sense it's part of the mechanism of what does it really mean when a paper is published in a journal, given that it's been circulated as a preprint, et cetera, et cetera, et cetera? It means, oh, it got reviewed. Well, depending on how good or bad that reviewing was. And it's often not a very convincing thing. It's kind of more of a. I often view it somewhat cynically as a sort of natural selection moment where, you know, if a paper, if the reviewer is benefited by the paper being published, then accept the paper. If the reviewer isn't benefited, then reject the paper. I mean, that's not a bad model, I think, for how a lot of that works. But in any case, it's something where when it's published in a journal, okay, maybe somebody actually read it and checked the box and said, yes, I accept it. Maybe if you're lucky. I'm not sure that always happens. I think it can be very much this natural selection theory that's just like, look at the citations. Do you like them? Do you not? Is the person a friend of yours or not? Just that's the idea. Not supposed to be that way, but I think that's how it's evolved. And certainly the anonymity of typical peer review really sort of supports skulduggery. I think it's kind of a bad idea. Many years ago, well, I haven't reviewed papers for a long time because it's not a good use of my time. But there was a transition that I made at some point of insisting on signing any report that I wrote on the grounds that, well, I was going to say something sensible and hopefully constructive. Even if I was saying, look, this paper doesn't make sense to publish in its current form and I was going to take responsibility for that. And actually, somewhat paradoxically, a number of journals said, well, if you're going to sign your reports, we don't want you to write these reports. It's like, really? Why not? That doesn't make any sense. But in any case, I think the thing that has sort of evolved in that area, it used to be the case before the kind of pre print system came into existence, first in kind of printed form and then in electronic form. In printed form. It was in place in the 1970s, in electronic form by the 1980s. But journals had a certain value in just distributing something because there'd be thousands of subscribers, all the various university libraries, and it will be a thing that got physically sent to all those places. I mean, I certainly found it satisfying when I was a kid publishing papers and academic journals. And you can kind of like go into any university library anywhere in the world and you could find your paper. That was a big deal in those days. Today, of course, if you put something on the web, you can go to any web browser and you can find your website, so it's kind of less of a big deal. But I think that the. Then the other thing that was true at that time was the very act of printing something, doing the copy editing, the typesetting, the printing, that was all a significant investment. And so the very fact that somebody had gone to the trouble of doing that meant that probably somebody believed in that thing. And in addition, the kind of original, typed or even handwritten versions of papers were much harder to read than the nicely typeset ones that actually were published in the journals. But by the 1980s, it was, well, at least now, I guess I was early to that business. So it must be at least by the end of the 1980s, it was much more routine to just say, we're going to make a camera ready copy. We're just going to produce what the journal is going to print. And so that whole sort of segment of investment that had to be made by publishing companies just went away. I have to say, you know, we published a little journal called complex systems since 1986 that I think has a, you know, has a certain community and certain utility. But I have to say that this theory of people just send you the electronic files and then you publish it. I gather that's not really as real a theory as it might be. I mean, if you actually want it to look good, you have to do some work still. But be that as it may, certainly the work dramatically decreased. I think what should have taken its place is kind of computational publishing. You know, when I write things that I post on my writings website and so on, I make a point of making sure that every picture I have, you can click it, you get the Wolfram language code, you can run it, you can See how it came to be. You can modify it, you can build on it. That seems to me to be a very powerful mechanism. But that's not a mechanism you see pretty much anywhere else. It's a mechanism you see some people using with our tech stack. And that's nice, but it's a mechanism that is not nearly as widely used as I think it should be. If one wants to sort of take the academic mission seriously as a thing where one's progressively building kind of capabilities and intellectual results. But I think the thing that is non trivial to get those sort of that computational essay format with code that actually runs and so on, it's non trivial to have make sure that code really runs and to make sure it's all clean and such like that, in my opinion is where publishers should be adding value in journal, journal publishing. I'm not sure quite where they're adding value right now other than in brand development. And you know, but the ecosystem is very complicated because at this point universities have made use of the fact that oh, there's sort of a, A, there's sort of some value maybe to the peer review process. Maybe there's a sort of value, oh, did you manage to get into this prestigious journal type thing? And that then becomes the mechanism for are you going to get promoted in your job, are you going to get tenure, are you going to get a raise? All those kinds of things at universities. And there's now long after I've been in academia, but there are all these complicated calculations about H indices and impact factors and a lot of sort of quantitative, you know, how cool is your paper kind of thing type measures that are very directly used by universities to decide. It's kind of like a quota for a salesperson is kind of like what was your, you know, H index, I think it's called, you know, this year type thing. Well, if you have a good one, then you'll get a raise and if you don't, you won't. It's kind of like did the salesperson make quota or not? Academics might not like that characterization, but that's really what I think it has become. I think the. So this question of access to this kind of material, I don't know how important it really is. I think it's more a question, you know, the web has an awful lot of stuff on it these days and there are also these complicated kinds of republishing mechanisms and websites that I don't quite know how they work from a copyright point of view, but I think maybe they're related to the fact that authors retain certain kinds of rights to their pieces. So if you put them there as an author, that's an okay thing to do. I'm not, I'm not quite sure how that works, but in any case. So I think the, you know, it is the case nowadays that a lot of kind of the raw access to research, the web is a pretty good place to find it. Now. The, you know, what did fancy journal X pick this week as its papers? I suspect you can even find the table of contents somewhere on the web. And I suspect if you just do a web search for those titles, you'll find their preprint versions and they'll be pretty much the same. You know, is it, you know, I don't really know what the role of public libraries is or should be in the modern world. I think they are. I know in the town where I live a lot of money was spent on a new public library and I've been meaning to go in there sometime for the last 10 years. And actually we were trying to organize some event there and it was like, oh no, we can't do this. So that wasn't an uptick either. But so I'm not, you know, I think they have more evolved to be this kind of community place than to be a thing which has, you know, books and such. But I have to admit that I've been really negligent, I suppose in not really having been in public libraries much in a long time. I always used to really enjoy kind of going to a big library and looking at all the various current journals and looking at was great when you could get access to the stacks. Actually I have been in some libraries recently where you could get access, the university libraries where you could get access to the stacks. And that's really cool because you see all these just weird books that you never, you know, you kind of didn't imagine, oh, that was published in the 1950s and oh, that's pretty interesting. It. So. Yes, I noticed Foxholt's comments. There's already an inundation of LLM written peer reviewed papers. Yeah, yeah. I mean look, it's. The whole thing is kind of. I don't know how that's going to end because I think LLMs are usable as reviewers as well. And what's an LLM going to do? An LLM is trained on existing knowledge. So as a reviewer, an LLM is going to say, oh my gosh, this paper isn't based on existing knowledge. Shocking, you know, reject it. So you're kind of, if the LLMs, if you automate reviewing with LLMs, then you have just reverted all of kind of knowledge development to the mean, so to speak. So I think the. Yes, it's, it's a, you know, I have to say it is a shame to me that the kind of the, it's an interesting experience to go into university libraries, for example, particularly university libraries if you're interested in kind of high end academic type stuff and really being able to browse things. I think people didn't perhaps recognize the value of that as much as maybe they should have done. And at universities things got fragmented into, oh, there isn't just one main library, there's 15 different libraries. You have to go specifically say, I'm going to go to this library, that library. Maybe there weren't as many people like me who were interested in a generality of things than one might have hoped and people were just narrowly interested in, oh, I just read these two particular journals and I never look at anything else. So. But I think it's an interesting question what the kind of role and mission of libraries should be. And I think, you know, access to online resources and sort of licensing of those things has been at various times a mission of libraries. Although I'm not sure it is so much these days, but it's something that could be, could be valuable. Another thing at universities and so on is being the data repository for material coming from the university. That's another thing that there was interest in that a decade or so ago. I'm not sure that really panned out. That to me would be a good, a really worthwhile sort of mission for that kind of group. All right. Oh, I see somebody's commenting that in Sweden there's a book bus that drives around with books to suburban areas. That's cool. I mean, there are these little cool mechanisms. Like every so often you'll see in some towns, you'll see these little boxes that people have out on a street that say, you know, here are some books you can borrow, put them back here, or put your own books into this box. You know, you see that in some communities that seems like a cool thing. It's, you know, books remain in my opinion, a very useful unit, a unit of knowledge in effect. I mean, there's a certain set of things like I can go to a website, I look at one web page, I'll learn a certain amount. But a book has this feature that it's a lump of knowledge. You kind of know how big it is. It kind of has a certain set of expectations. I mean, I've been surprised that the online versions of books are in many ways less, perhaps viscerally useful to me often than physical versions of books. So one might not have expected that. Anyway. All right. Well, thanks very much for joining me today and look forward to chatting with you another time. Bye for now.