Business, Innovation, and Managing Life (December 6, 2023)

Okay. Hello everyone. Welcome to another episode of Q and A about business innovation and managing life. So let's see, lots of questions here, some very long form. So let me see if I can parse them out a bit. Let's see. LC asks, how do you manage conducting deep long term innovation along with dealing with short term commercial and funding necessities needed to keep the lights on? Well, you know, one of the things that I guess I'd been lucky about is, is that we have a company that's been running for 36 years now, a little bit more than 36 years, and we've been able to maintain pretty consistent profitability primarily by making, you know, having revenues that exceed expenses. And that's achieved in part by managing expenses, which is something that is something you have to plan for in a fairly long term way. I mean, if you say in a particular week, oh my gosh, we're spending too much money, we have to cut expenses at least for the company of the scale that we have. That's a pretty unrealistic thing to do. It's something you have to plan for over a longer term. But I think one of the things that's helped us a lot is that we've had a fairly stable be base of business and that's kind of what we start from. And that allows us to think about things in a fairly long term way. And we've been fortunate not to have things like outside investors, not to be part in the public markets where we're having to impress people every quarter that we made money that quarter and so on. So we're able to take a somewhat longer term point of view. I would say that that's a feature of being a kind of closely held private business and having a fairly stable kind of market and base of customers. I'm not sure that's always easy to achieve, but that's been what we've been able to achieve. And then the question is sort of to what extent does one deal with long term types of things and to what extent is one just, just solving the problem of the day? I suppose it's been my personal sort of desire and in a sense irresponsibility to think about long term things, sometimes with a highish priority relative to short term kinds of things. I mean, for me personally, I suppose one thing that I tend to do is during the certain sort of standard pieces of the working day, I'll do meetings that relate to kind of shorter term things. And then when I'm not doing that in the evenings and so on then I personally get to work on longer term projects and the kind of research that I do and so on. I think in terms of our company, we've kind of always tried to maintain a sort of portfolio of different kinds of projects from ones that are quite short term, like there'll be some piece of software development where we're going to deliver the results in the next version that comes out in a few months or something alongside kind of much longer term things where we know this is a project that's going to take a bunch of original research and it's going to take a minimum of several years to come to fruition. And like for example, right now we're just preparing a new version of Orphan Language and Mathematica. We've had a kind of a rhythm of bringing out new versions every six months. We've maintained that for the last few years. This is kind of an integer release, so slightly bigger collection of things that are being sort of finished off for this release. But if you look at the things that we're delivering today, some of them were started only a few months ago, but the vast majority were started quite a while ago. And some of them build on long chains of development that go back a decade or more. And for example, this year is a good example of kind of what's new versus what's long term. The LLMs landed this year and there were a lot of interesting new opportunities and possibilities that were opened up by that technology which we kind of jumped on. And it was a non trivial thing to kind of take our company, which has lots of things sort of that are going on at any given time and kind of get things done with LLMs quickly. So for example, one fairly impressive example was we started in January working with OpenAI to make a plugin for ChatGPT that would connect to Orphan Language and Wolfram Alpha and that was delivered in March. And that involved lots of people both sort of core development, research, software engineering, deployment, et cetera. And I didn't count how many people were involved, but I'm going to guess there were maybe 100 people, 100 bit more than that, who were involved in that project and in pulling that together. And it was a sort of, it's a non trivial organizational capability to be able to decide we're going to do something in the next couple of months and we're going to put let's say 100 people onto it. And that is something that does not just cause complete confusion in the organization as those people work on these things. I think A number of years ago, when actually back in, what was it, 2014 maybe, when earlier period of kind of AI excitement was when image identification through deep learning became sort of possible. And at that time we built a website, Imageidentify.com that I'm sure is still up and running today. And that involved training a neural net and deploying a website and so on. And I think I counted that there were 70 something people involved in that project. But that project was done in a matter of two or three weeks. And many of those people were involved only for a couple of days. And I think it's a sign of kind of corporate capability to be able to have these things that can be done very quickly but involve the orchestration of fairly large numbers of people. It kind of reminds one of sort of human activity where it's like you do something very quickly and then you watch your kind of heart rate decrease after you finish that piece of intense activity. And it's a good sign if it goes down quickly. And similarly for an organization having something where you have a piece of sort of urgent exercise that's done and then the organization kind of goes back to what it was doing before fairly quickly. One of the challenges when you have a project and you ramp it up is to make sure that once it's finished, you can actually ramp it down again and that you don't end up with something where you put lots of people onto it and somehow it sort of keeps running even though the project is finished. I have to say that sort of forcing that in our organization in past years, I remember, or maybe 15 years ago or something now maybe even more than that. We had a project where it was a fairly intense project to get out a little product and we got it out. And then I realized that we had so adapted to this idea that when projects are finished, the people should go off and be working on different things immediately. That it came out on a Friday and by Monday wasn't clear that there was anybody paying attention to this product anymore. And that was also kind of wrong and needed sort of a change of practices to deal with. But I think I will say that for myself being able to sort of do long term things and it requires a certain degree of discipline to say I'm going to allocate some amount of time to doing this regardless of the fact that there's always another urgent thing that you could be being pulled into. And I suppose over the years I've just adapted to a certain degree of, I suppose you could call it irresponsibility that yes, there are urgent things coming in, but it's like, no, I'm going to spend the day or whatever working on this longer term thing. Even though, yes, there are sort of urgent things piling up, but I'm going to get to those the next day. And I suppose for me, kind of my primary means of kind of dealing with ongoing issues is email. I don't use texting and various kinds of messaging channels and things like this because I find that for myself at least, email is a much better rhythm for doing things. Because it's like I'm not going to look at it right at the instant it comes in. I'm going to have these blocks of time where I can systematically process. I get maybe these days, maybe five or six hundred email messages per day, and it takes a while to process those things, but I do them in blocks of time. And that's a much better way for me to concentrate on things than it would be for me to be in the middle of something and I'm writing some sentence or whatever, and in the middle of the sentence I'm like, oh my gosh, another text came in, let me deal with that one. Oh, another text came in, let me deal with that one. Now the big advantage, as far as I'm concerned about email is that it's much easier and more systematic to archive it. I now have, what is it, 33 years or something of perfect archives of my email. And it's super useful to me. It's kind of an enhanced form of memory where I get to search those things and can immediately pull up, oh, yes, I interacted with that person 25 years ago. And here's what happens. And I have sort of a vague personal memory of that. But when I actually get to see the detailed email, it really fills things in, in a useful way. So let's see a question here from Jamie. How well would you say your current understanding of business and academics is today in comparison to when I first started my career? Well, you know, it's funny because I've done different things at different times. And so, for example, when it comes to academia, I was involved in academia early in my career. I haven't been sort of deeply involved for 35 years now. And so for me, when people tell me about sort of the detailed issues of academia and the politics of academia and the conventions of academia and so on, I'm actually kind of out of the loop by now. I think I have tried to keep some degree of awareness, but I'm not sure that I would be perfectly on key in dealing with some kinds of niceties of the academic system, which by the way varies between different fields, different countries, et cetera, et cetera, et cetera. But in terms of business, at the time when I started my career, I wasn't involved in business at all. I got involved, I started my first company, which is a long time ago now, 1981 for sure. I've gotten much more fluent in dealing with all the kinds of things that come up in business and in doing projects, research, etc. I suppose one of the things that happens is there are things where one's sort of seen it only once in one's career, but if one has a decent memory, one remembers, oh yeah, I seen something like that before. So when it comes up again, it's much faster to make a decision about that thing. And I've certainly found that. Now for the vast majority of issues that come up that are sort of strategy kinds of questions, I kind of more or less immediately know. Well, I either know that I know the answer or I know that I don't know the answer. And I don't really. It's like if it's a question of can I think about it, do I need to think about it for a few hours to figure out if I know the answer? No, I really don't. I either kind of know it cold and I kind of know what to do, or I know I don't know and I know that I need more information or whatever to be able to come up with an answer. So about those kinds of things, I become vastly more fluent and quicker as far as things like research are concerned. Same kind of story. It really helps to have a base of, oh, yeah, I've seen kind of a situation like that before. It also helps that I, I know my own ways of thinking about things and I know my own kind of limitations in thinking about things. And so I kind of am fairly clear this is something where I'm going to be able to figure it out quickly. This is something where I'm going to have to get somebody to help me. This is something where it's going to be really quite difficult and so on. So for instance, I mean, in, oh, I don't know if I'm writing something, I kind of know if I'm doing some piece of research. There's a question of do you kind of have a sort of a philosophical discussion about how things are going to work or do you actually go and look at data from the world? Well, sometimes you have to look at data from the world. Otherwise you can say absolutely nothing. But sometimes there's a whole bunch to say that's just on the basis of kind of pure thinking, so to speak, and not on the basis of detailed analysis of data. But I know that if one gets into detailed analysis of data, it's always complicated. There are lots of projects that I've done where I kind of have a rough idea of what to say and where there are interesting things to say and interesting kinds of discussions to make. But where, if I actually start looking at the data, I know it's going to be very complicated. There's going to be all sorts of issues of interpretation and so on. It's going to be hard to conclude what's going on. And so there are a number of projects where I kind of started them and then I started looking at data and then the projects kind of went off track because the detailed data just has so many other issues that aren't part of the main story of what one might have to say. I'll give you an example that actually worked out just fine, but I was sort of worried it was going to go this way. I was interested in kind of the AI future and the question of sort of what's happened with jobs in the US over the last 150 years and what's kind of the distribution of different kinds of things people have done. Well, there are issues. There are questions about that that I was curious about. There's some questions where the answer was very clear. It's like things have sort of fragmented from a time when most people worked in agriculture to times when new fields where things developed that were automated. And then on after the automation happened, you saw sort of fragmentation into lots of different kinds of subdomains and so on, and new kinds of job categories that opened up all that stuff was pretty easy to see. It was not so easy because if you look at the raw, raw data, there are all kinds of occupations that existed in 1850, and I don't even know their names anymore. And I had to do a whole sort of project aggregating kind of occupations that had had names in 1850. What are they bucketed with today? So that took a bit of effort. But then there were things I was curious about. So that all turned out to be fine and I was able to make definite conclusions and sort of weave them into a narrative. But then I was curious about other kinds of things. I was curious, I think a related project a number of years ago at this was a project where I was interested in. If you look at notable people, we have a lot of data on notable people and you look at where they were born versus where they died or where they were born versus where they mostly worked. And I was curious about the extent to which there was motion from for example, rural parts of the US to other parts of the US of people who were in some small town in the US in 1880 or something and then made it big in New York City later. And the extent to which that happened in the past and is still happening today versus that not happening so much today. And that turned out to be. I don't remember all the details of that, but that turned out to be a much more complicated story fraught with all kinds of issues about how the data was sliced and so on. And it was one of these things where I actually didn't know the answer at all why I was looking at the data. But there are cases where one has things to say about kind of what one can expect, let's say with AI and jobs, where there are things to say about that that don't require looking at the historical data. And when one does start looking at the data, one can get very confused in the case of this question about moving from rural parts of the US to other, other places and so on. Well, I've now forgotten the precise details of that study, but the main point was I never wrote anything about it and I never wrote anything about the things around that because I kind of got submerged in the actual data and couldn't really make conclusions. And I think that's so, you know, that's something I sort of learned about projects, is you have to be very careful if you're actually going to deal with real data from the world. There may be some things you'll be able to answer, but other things will take you a long time and you won't be able to come up with a definitive answer. So it's just sort of a thing in terms of the management of projects. There are probably many things to say along those lines. Let's see. Oh gosh, there's a question here. From Paul. What's my perspective about AI's omnipresence and ability to introduce universal basic income into start of society economies on Earth? What rules will be applicable? Well, you know, first thing to say is sort of automation has been coming throughout the history of technology. The story has always been there are things we humans just do for ourselves and then later on those things get automated. They either get automated because there's an actual machine that does it or kind of the whole supply chain of the world kind of segments things and makes it be the case that you no longer have to do that for yourself. It's kind of done as part of some much more larger scale activity that's not you particularly doing it. But that trend will surely continue. And surely the things that we now identify as AI, which has changed a bit over the years, we will see more and more of that kind of automation put into more and more systems in the world. Whether that's because it's cheaper to do it that way, whether because things need to be figured out fast and the computers can figure out the map faster than us humans, or whether it's just that kind of. We decide, oh well, if the computer could figure out what the things we might consider doing are, why don't we just follow the machine? Rather than having to figure it out for ourselves, we could probably read a map for ourselves. But why not let the GPS tell us where to go and why not just follow what the GPS says? Now this question of whether in fact it will be the case that there's less for humans to do when there's more automation, I was just mentioning the historical data on that does not support that idea. What seems to happen is that as some category of activities gets automated, that enables a lot more things. When agriculture got automated then there was all kinds of work to be done in food distribution and all kinds of and restaurants and this and that and the other and things that became possible because the actual individual work of going and sort of harvesting one's own food wasn't there anymore. And I think that's been sort of a continuing trend that what we understand, well, what the point of what has to be done that can be automated, but that sort of opens up a bunch of new areas where there's much more human choice and where one kind of needs humans to do those things. So kind of the concept that we can all kind of relax and not do anything because the AIs will do it all. I think there'll be a lot more things that we choose to do because certain things got automated and there'll still be things for us to do now. You know, it could be the case that we'll reach a point where we say, well enough of what we need has been is automated and is provided we don't need to do any more. We're just not interested in doing any more stuff. I don't think that's going to happen. I don't think that's the way of our species. I think if you give people more opportunity to have what might at first be classified as leisure time or something, they'll go off and they'll learn how to play video games. Well, and then they'll be streaming video games and they'll be going into esports or something. Or you give them the opportunity to do some other kind of thing that at first seems like it's just, oh, now we don't have to do the hard work. We don't have to be out in the fields collecting our own food and so on, and so we just can relax. Well, we find things that we want to do, and we end up eventually finding things that we want to do with some intensity. So that's my guess about what's going to happen. I don't think it's going to be the case. In most segments of society, the people just say, okay, we're done. Now we have what we need. We don't need to kind of push to do anymore. But, you know, and we've already seen. And this is sort of a, you know, one of the traps of these kinds of, you know, grand socioeconomic schemes where it's just like, well, you know, the state should provide for everyone type thing of the traps of that is, then, yes, there will be some segment of people who just say, well, I don't really need to do anything. And it's bad for them and bad for everybody, so to speak. Particularly bad for them. I mean, I think that's one of the features of some. We humans seem to want to actually have objectives and actually want to kind of sort of achieve things and just sort of hanging out doesn't seem to be, you know, when one says, you know, people are just hanging out somehow, even if they're not doing things which look productive, they're achieving some social purpose for themselves or whatever else, I think. Or for me, for example, I'm not a big one for playing games of any kind. And for me, that looks like an utterly unproductive activity. But yet for many people, that's something that feels like it has an inner achievement and feels worthwhile, so to speak. So one of the sort of, I suppose the patches, if it's like, well, there's nothing really that anybody needs to do is give everybody universal basic income. I have to say, I think I know there are a bunch of experiments that have been done. I know some people who've done these experiments. It's very difficult. I mean, it seems that. It just doesn't seem like a Great fit for the way we humans tend to actually be. And it seems like then once one has that, then there's all kinds of gaming of that of, oh, then let's make some complicated financial arrangement where people kind of commit their future universal basic income in terms of more now and then they'll invest that and it'll get very complicated very quickly. And I have to say I tend to feel that it's probably not very compatible with the way we are as a species. Maybe I'm short sighted in thinking that, but that tends to be my view. And I think that's a quite separate thing from issues of kind of income inequality and those types of things which I think are, I don't claim to understand enough about kind of the socioeconomic situation to know in full detail what's happening there. Although I think. It's complicated because again, there is the need, I suppose for many people for some degree of sort of I want to get ahead, which means I want to get ahead of other people. And that also the getting ahead thing is important for sort of driving economic activity and things which will eventually benefit everyone. So if you say, well, nobody can really get ahead, a lot of people will say, ah, I can't be bothered. I'm not going to put that extra effort in. I know even for myself, you know, if I look at things, I try not to pay too much attention to, you know, the details of what happens with, you know, different tax laws and things like this. But I have to say, you know, there was a recent kind of disaster in the US with the way that R and D tax credits are handled that led to all kinds of huge tax bills for us and made it, you know, it'll probably get reversed, but if that continued, it would actually be a significant disincentive to do research and development in the US in companies. And presumably that's not ultimately what's wanted. And for me, I'm sufficiently keen to do research and development that it probably won't put me off very much. But I have to say if I were to pick between let me do some big corporate research and development project versus let me do a piece of basic research that might sort of help tip the balance. And that's something where, yes, it is true that sort of, you know, provide, you know, making these kind of regulatory changes, eventually, you know, it'll maybe for some people it, it won't really have an effect because they're so keen to do R and D, they're just going to keep doing it. But there'll always be people on the margins where it's like, yeah, well, wait a minute, am I going to spend the next month doing R and D where I'm going to have to, you know, treat that as, as, as a, as a capital, capitalize that and, you know, I'm not going to be able to deduct those expenses. And so that's going to cause me all kinds of trou. Or should I do basic research where it's not nearly as short term an economic benefit for the world at large? Maybe in the long term it's a bigger economic benefit, but that might be 100 years out. And what am I going to do? And if there's some sort of government thing that tips one way or the other that might affect things. And I suppose that can be applied to many of these things. You know, I tend to think, I mean, one of the things that I personally been both frustrated about and curious about is it feels like as things change in the world of computation and so on and computational thinking and things like that, it's like we're in a very kind of open playing field in this situation, because it's not as if there's a group of people that have long built up to be great computational thinkers. It's something where there's technology ideas and things that I've been much involved in, for example, in building up that direction where it's kind of, they're just sort of getting opened up and they're open to the world. And it's not the case that people have sort of a long history of, oh, I've been doing this for five generations type thing. It's something where in a sense, it's accessible to everyone. We've worked and I've worked quite hard to make the things that we built, for example, accessible as broadly as possible. And so it's available out there. So now the question is, well, who chooses to take the initiative to take advantage of these opportunities? And that's where things get frustrating, at least for me, because I'm really curious about sort of what talent is out there in the world. I understand something about what talent is out in the kind of what one might think of as sort of the elite parts of the world of the people who live in the kind of the expensive cities and go to the fancy schools and all that kind of thing. I understand that talent, there's lots of great talent there, but I also suspect there's plenty of talent away from those places. And the question is, how does that talent get connected to the opportunities that now exist in the world. And I don't think that there are times when there are much better developed areas where there really are some real barriers. Because it's like, well, you have to have learned this and you have to have learned that. And it's not so easy to learn that in this place or that place. But I think with these areas like computational thinking, it really is pretty open. And so then the question is more initiative and desire to get involved in those things than it is, I think, sort of access to the kind of the platform that's needed to get there, so to speak. And so, for example, it's been a long running story. We run summer research programs for high school students and now a little bit for middle school students and so on. And it's been a long running issue that I've had of how do we get sort of broad distribution of folks coming to those programs. And back when we first started them, I don't know, a dozen years ago or so now it was, you know, we had kids from the top coastal schools in the US and so on, and they came and that was great. But we didn't even have sort of good geographic distribution. We started kind of advertising these programs and WolframAlpha and places like that. And then we succeeded in getting pretty good geographic distribution. We've had sort of scholarships for these programs for forever. And the frustrating thing for me is how rarely those things are taken up that we get. Although we get decent geographic distribution, we do not get particularly good socioeconomic distribution. And it's something I've been sort of, why is this happening and what can we do about it? And it's very, I've talked to so many people about this, and a common sort of thing people say is, well, you need to reach all these different sort of segments of the world and you typically end up having to reach them very locally. There's not sort of a broad way to reach all the rural schools in the US or something like this. And I've even done some field trips and things like this to try and sort of figure out what's going on. And I'm afraid the thing that one tends to find, at least I've tended to find, is that sort of the vast majority of what's happening is not. There's no, you know, to my mind, there's plenty of bright people to be found all over the place. That's certainly not the issue. The issue is what do people want to do? What do they think is worth doing? And it's different for all of us. I know I made up some surveys a number of years ago for kids particularly about what are things you might want to achieve in your life. And my guess is that by the time people are early teenagers, the answers they'll give to those things actually won't change that much as you go to later years. Although that's one of these data questions that I haven't really tested. And it's very different for different people. I mean, some people are like, oh, I'd love to have a TV show type thing. Another person might be, I'd love to be an astronaut, another person might be, I really want to just make lots of money type thing, whatever else, different kinds of objectives and they're very diverse, but sort of the kind of idea that one might go into the tech industry or something like this. I've certainly tried sort of talking about that in lots of situations. Well, a decent number of situations where it's like with kids who are not really anywhere close to the tech industry. If you go to Silicon Valley and you talk about the tech industry, everybody knows about the tech industry. You go to some rural place in the US and talk about it, nobody knows about the tech industry. They know the effects of the tech industry, but they don't know the insides of the tech industry and sort of what the career paths are for going into it and things like this. But the thing that I've noticed is that people are like, it's kind of like there's a certain amount of shrug the shoulders, but it's just not something that I do. And I would say that about some of the things that were on my list of hey, do you want to do this when you're grown up? Like, do you want to be an astronaut? No, I don't want to be an astronaut. Me personally, not my thing. Do I want to do some kind of reality TV show? No, not my thing. Although I suppose live streaming, my design review meetings might be my version of reality tv, but. But not in the usual sense. So different people have these different objectives and I think that's one of the things that can happen, I suppose is that, for example, something like the tech industry, which has been very economically successful in recent years, if people sort of attach themselves to that, they will end up being sort of economically very advantaged. But if they're like, well, the tech industry just doesn't really, for me and they don't really know anything about it, they're sort of cutting themselves out of that world. I don't think it's happening because the things like computational thinking are fundamentally inaccessible to anybody. At least I've tried to make sure that that isn't the case. I think it's more just the. Well, I don't really. It's not really a thing that I see myself doing and so it doesn't happen. How to solve that, I don't know. I'm very, very curious about how to solve that. I'm interested in suggestions about how to solve that because I've the. As I say, it's something. One of those sort of projects that I've long wanted to pursue but don't really even know how to pursue it. And I'm sort of waiting for a moment when perhaps some different sort of form of communication or something, just as things have opened up between the web and social media and things like this, where that becomes possible and then I can explore that more. Let's see. Gosh, lots of questions here. Ah, it's an interesting question from insert here. Do you think that with the disappearance of physical labor due to AI automation, it will make a comeback as a healthy hobby? You know, it's a funny thing this, because lots of things that kind of get automated away and where there's a machine that can just do it. Absolutely. People sort of choose to do them as hobbies. I mean, you know, who needs to run? You can just drive a car. But people do it as a hobby. And I think it's sort of a feature of, again, a sort of feature of our species. There are things that we're kind of built to do, and even if we don't absolutely need to do them, we will, or at least some set of people will choose to do them because it's kind of what we're built to do. And yes, so I suspect that's a thing. Let's see, RITC asks, how do you anticipate AI to AI conversations? What sort of information and insights are likely to emerge? You know, there are already plenty of AI to AI conversations that happen, whether it's sort of the bidding systems for online ads or whether it's lots of other kinds of things. I think the thing that is really the question is to what extent those conversations are tethered to human interests and human objectives. Because kind of the raw AI just doing its thing and sort of doing computational stuff for the sake of doing computational stuff, much of what will happen there won't necessarily have any connection to human objectives and may in fact be really hard for us humans to understand the point of. I mean, I've given this analogy before, but it's kind of like what happens in the. The natural world where there are all kinds of computations that are happening, where sort of natural systems are figuring out what to do next. And we look at those natural systems and for us, we don't necessarily have kind of a human narrative that explains what's going on. The system is just doing what the system does. And so it will be, and already is to some extent with AIs that they're just doing what they do. And the question of whether we can say, oh, we understand why that's happening and, and we can kind of have a human narrative about it is not clear. So I think that's a piece of the story. But in the end, a lot of what we are sort of getting AIs to do is done in the service of human goals. So while the insides may not be comprehensible insofar as we still sort of connect what we're having AIs do to human goals, and we're likely to be able to understand sort of the outer boundary of that and to be able to say, oh, that AI to AI conversation, it was just a thing that was done because these things were bidding against each other to try to figure out how to get these two humans to buy something and sell something and so on. Let's see. Well, there are funky personal questions here. Prab asks, do I ever take time off? Depends what you mean by off. I mean these live streams, for example, are an example of time that I take off from the things that I usually do. So in that sense, I absolutely take time off. Do I take time that cannot be seen as productive by anybody? I don't know about that. I tend to watch kind of consumer movies every week or two. I suppose that's one sort of time off thing that I do. Although I partly do that. Well, partly I just do that because I like to do that. And partly I do it because it does keep me slightly in touch with kind of cultural things that are going on. Although I have to say I'm not sure that that's so true anymore for movies in movie theaters, which is where I tend to like to them because I'm not a television watcher at all. And I suppose I've maintained a boundary between that causes me to say, if I'm going to go watch a movie, I'm going to go and drive to a movie theater to do it, rather than, oh, I'm just going to bring it up on my computer or Walk down to my basement or something. I think. I suppose for me, I'm a person who has a lot of things that they like to do which are probably aligned with things that one might think of as productive. But I'm not sure that many things are times off from what I normally do. But perhaps, I suppose I've had a long time hobby, long time habit of doing things as a hobby that turn into things that I do for a living, so to speak. So for example, there was a time when I was an academic for a living and I did business as a hobby. And I started doing business for a living and started doing research as a hobby. And then I'd done research pretty seriously. And then some people would think of that as a thing that I'm doing if not for a living, from a financial point of view at least as a major activity for myself. And there are other things, like education, for example, that I've sort of done as a hobby. But in a sense by aggregating things into things like summer research programs and summer schools and so on, that's something more than a pure hobby, so to speak. And I suppose there are things like, oh, I've been interested in history and history of science and so on. And for a while that was sort of a hobby type thing. And then I started writing quite a lot about it and even writing a book about it and so on. And that becomes less of a pure hobby, so to speak. So I suppose from that point of view, the things that I like to do tend to be things that are in some way from some definition of productive. So in that sense, I suppose I don't really take quotes time off. And in any given day I mostly am, well, spending most of my time doing things which somehow will be productive in addition to some modest amount of. I suppose I allocate a certain amount of spend a certain amount of sort of social time. Although many people would criticize me for being very poor on kind of the small talk aspect of social time. Because I'm usually with people where I'm trying to talk through things or learn things or whatever else, rather than being in the kind of small talk category. Insert comments. Would you say that a person whose job is also their hobby is a happy person or a person lacking both on the job and the hobby? You know, for me, I mean, I'm sure it's different for everybody, but for me, the alignment of kind of what I do and what I like to do is terrific. That's something that I consider to be an important aspect of my life that I can align what I like to do with what I spend my time doing. And I think this is a thing in terms of people sort of picking their jobs and so on. It's really good when one's able to do that. Sometimes people will sort of pick a job direction because of what they think that job involves, and then they'll be a little disappointed that what the job actually involves is something that isn't really what it isn't. The maybe there's a job that has moments of dramatic excitement combined with sort of 98% of the time being kind of humdrum. And one sort of looked from the outside and seen those. Those 2% pieces of excitement and says that's what that job is about. But yet one really doesn't want to live through the 98% of the humdrum time once one understands what these things actually are. And sometimes there are things which some people would consider, oh, that's kind of, oh, look, I've been doing computational language design now for 40 years. And you might say, some people might say, and it's hard work, by the way. And some people might say, how can you go on doing that for such a long time? Don't you get bored with it? And the fact is I just find it interesting and I find it very intellectually rewarding and so on. And so I keep doing it. And yes, it's productive and one's building sort of a bigger and bigger tower and one's able to do more and more the taller one builds that tower. But it also is just something that I like to do and I find fulfilling to do. And so I keep doing it. Let's see. I'm not sure I know what that question means, so I'm going to. Let's see. There's a question from David here. Is game theory useful for running a business? It has not been, in my experience, in fact, sort of formal tools and sort of formal. It's sort of disappointing because I've spent so much time on science, on figuring out things in science. Although some of what I figured out about science is that there are limitations to what one can predict and there are things where we sort of just have to let it run and see what happens. But it's sort of disappointing in a sense that the sort of. The formal methods of science I have rarely found useful in running a business. The thinking, the analytical thinking that comes from science, I find useful every day. The yes, you can figure it out. Let me look at the steps. Let me break this problem down. Let me see what's actually the important piece. Let me see what follows from what. That kind of thinking is critical every day, but the kind of. Let me go and ask my computer to do some game theory computation to figure out should I do this or that, or should I ask for this or that in some deal or something like this, I have to say I've not found that useful. I've sometimes seen if you do deals where you're negotiating some number, for example, and you kind of plot, one side asks for this number, the other side asks for this number, and you can kind of plot how the numbers come together and they do, you know, they have a kind of a definite curve. It's nothing very profound. And, you know, maybe if there were many parties involved, you could use some sort of game theoretic way of analyzing it. I don't think I've ever been involved in a situation where that would have been relevant. And so it's a little disappointing in a sense that you've got all this kind of mathematical science knowledge and yet that isn't what you can use now on a local scale. And in terms of predicting things and projecting things and so on, there are some local things one can do. Although I have to say things like, okay, how should we price this product? How should that work? We've done analyses of that kind, although usually in the end they're like, but there are things we don't know. We don't know whether people will have this reaction to that or that reaction to that, and we don't have a great model for that. Now, perhaps in a business where we had more sort of repeated kinds of things that we'd done where we could say, oh, we're bringing out a new kind of toothpaste or something, and we brought out 50 new kinds of toothpaste before we could more effectively use quantitative methods to say what was going to happen. Or given that we did this survey in this place and that survey in that place, we can conclude what's going to happen. In my experience, in the things we've been trying to do, which are mostly really doing new stuff that really hasn't been done before, certainly by us and often by the world at large, it's really hard to figure those things out. I mean, I'm quite certain that if we'd done a survey of whether anybody would care about WolframAlpha before it came out, we would have said, ah, this is hopeless, nobody's going to care. But then once the thing actually exists and People can see what it is and it gets some momentum, then it's a different story. And so I tend to think that these kinds of, at least for me, for a non repeatable business, so to speak, that isn't sort of doing the same stuff over and over again, I think that. But as rather doing sort of new things all the time, I think it's pretty hard to use kind of these, you know, something like game theory or these various quantitative methods. So it has not been useful to me. As I say, I think I certainly am aware of cases where people at least say it's been used. Whether it's for working out some complicated multi party agreement like you know, the law of the sea or something between all different countries or whether it's about sort of figuring out the auctions for frequency allocations for cell phones and other kinds of things. These are places where they're both multiple parties involved and there's perhaps more repeated activity and where these things can be much more useful. But I think for kind of the innovative company, maybe it's not the most useful case. Des asks not necessarily business, but fun. Have you ever been to a magic show? Yes, I have been to a few. I, I have to say I always find them a bit frustrating because it's kind of like I'm, I'm very much of a, a person who wants to understand things and such a show is specifically intended for, you know, it's in the, you are not expected to understand this type category and so it's not my favorite kind of activity. Although the showmanship of the magicians is often fun to see and kind of, it's always, particularly when you see somebody doing basically anything. Well, that's always interesting. Let's see. Questions here. Gala asks how have interview processes changed since you began your career? Are there ways to improve application, screening and potential candidates? Could VR and AR environments be a way to test candidates in the work environment? Well, let's see for sure. I'm not sure the extent to which there have been changes in sort of the landscape of interviewing. For me personally there have been changes just because the kinds of interviews that I would personally do have changed as the organization I've been running has gotten larger and things like this. I mean I always find interviews interesting. I find people interesting and sort of the challenge of an interview is there's a person, you have a role, or maybe you don't even have a role that you've picked out. But the question is how does this person fit with the role you have or Some role that might be worth creating and how do you understand enough about that person to see whether there's such a fit? And for me, I suppose I've tended to adopt a pretty. My objective is just to understand the person well enough that I can answer that question. And for example, if I get to the end of the interview and I really don't understand the person and can't answer for myself even sort of basic questions about how the person would react to different kinds of roles and so on, then it's probably not a person that I should hire. Sometimes made mistakes about that because it's like what you don't know. In my experience, what you don't know is something that is a risk, so to speak. And so the more you can kind of feel like you really got to know the person well enough to sort of understand what fits are possible is a good thing. And I suppose my tendency has been to just talk to people about the things that they know about. And different people have very different interviewing approaches. I mean, there have been times when people and times when people are very keen on let's test people, let's give people some very explicit kind of set of test questions and this and that and the other. For most, well, things that I've interviewed people for, that's not really been a relevant thing. I mean, I suppose occasionally, yeah, I mean, I suppose for. Yeah, there are plenty of positions actually where there are scenarios and you won't say what would you do in this case? And I have to say for me, I'm much more interested in seeing how the person thinks about it than I am in knowing whether they happen to know the answer and nail it immediately. Again, that's a bit variable because there are cases where it's like the more you know, the better you're going to be able to do at this or that thing. But I would say that this whole question of whether as sort of a formal, you know, a test process for most kinds of roles, I have not seen that work very well. I have to say there are some somewhat more entry level roles that we have at our company where testing is relevant partly because we're doing things where people have no idea whether they'd be good at it. So for example, something like data curation, that's something that requires certain kinds of turns of mind and skills and so on. And it's like nobody really. Well, very few people kind of go to school for data curation. And the people we would hire would not be of that type anyway. And so it's really a question of is this a thing that somebody can do? And really what you're doing there is. You're kind of. It's a trial of well, this is what the actual job consists of. Can you do this? Well, do you like doing this? And so on. And that's a. So that's a place where I suppose that explicit kind of testing process works. I can't see that working for things like software development or lots of other kinds of activities like that. I would tend to think that for many of those. Well now, in terms of things like doing interviews by video and so on. You know, it's funny thing, I realized that I've been using screen sharing sort of web conference screen sharing or something since the early 1990s. And I think that the first time. And I had done a fair number of video conference talks, but the first time I think that I did a back and forth one on one discussion with somebody was an interview that I did must have been. When was that? Mid-1990s, I think. And it was kind of funny because the person, this person who worked for our company for a while is now off as an academic, but was sort of an interesting experience. The sort of. The first video interview and I didn't do another one for several years after that, quite a few years after that. You know, to what extent do I find that, for example, video interviews work as well as in person ones? I think they're fine these days. I think that I haven't. Can I. Can I really describe. I'm not even sure. You know, one of the things I realized for myself is that I'll run into people and I'll be very confused about whether I've met the person in person or only through video at this point. So I think that's, you know, that distinction has sort of reduced itself in terms of, of, you know, so. So is it going to be important to have the kind of. Oh, by the way, I've done in the past particularly I did many, many, many interviews just by phone without any sort of video or meeting or anything like that. Partly I did that because, oh, I was doing interviews at times when I was. Had a piece of time available because I was driving from here to there, there and it was a convenient thing to do. And sometimes because I actually found in some ways, at least in the past with, with not such good video quality and things like this, I found the video aspects of interviews confusing and I found it actually easier to just do it by pure sort of audio. I found That a kind of a clearer way to concentrate on what was actually being said. And I suppose for me, in interviews and things, one of the challenges always, at least for me, is always to listen enough and really listen to what the person is saying and not sort of assume that you know what they're saying, because what they actually say is relevant to what they're actually like, so to speak. I also find with interviews, one of the things that for me is a very negative sign in an interview is if I realize that at the end of the interview I did more talking than the person I was interviewing did. That's a really bad sign because for me, I'm always like, I'll end up feeling like I need to help the person out and I'll be doing lots of the talking. And that's a no. That's really not way it should be for the kinds of things I'm interested in. So I suppose other things that I suppose over the time of my career, like 40 years or something, there are probably changes in HR rules of things you can ask and can't ask in interviews. I don't think I was ever asking the things that you're not supposed to ask today back in the past. But I suppose for some people, that's been a change. I mean, one of the things that I think is always the case is that what one is interested in, in hiring people for things, is can they do what they need to do in the context of whatever organization one has? Their independent life and times are their independent life and times. And I take that point of view with respect to things we do at the company. It's like people. I have no idea what all the different views and causes and sort of things people are energetic about. In almost no cases do I know that about people unless they happen to tell me for some reason or another. But it's something where, from my point of view, it's like there are things that one is trying to do at the company or an organization, and those are things that one hopes people can come together and be effective at, whether they go off and they're sort of doing activism in direction one or direction two. And maybe there are two people in a team who do it in the exact opposite directions. And it's like, I don't think. I think the goal is to provide an environment where that's stuff that's happening outside, and it's not something that affects what's happening in the actual activities one's doing in the organization. And so similarly, it doesn't really, you know, it's not relevant in, you know, there are plenty of things that are relevant to whether somebody can work effectively in some particular role, but those sort of outside sort of quotes, hobby activities, which might be very serious activities for all one one knows, aren't, you know, aren't relevant if, if, if somebody is able to do the thing that one is interested in having them do. All right, let's see. Well, Ann is asking, should there be an AI system which does computational language design? The computational language could then be used to tackle problems of any kind and feedback to the language design AI. Yeah, well, actually, it's funny because one of the concrete uses that I've made of LLMs in the last several months has been that LLMs write Wolfram language code. And they do a sometimes good job, sometimes not so good job, but one of the things they do is hallucinate names of functions that they kind of think or expect will be there, that we try to tell them, read the documentation, so you know what's actually there. But sometimes they'll make up the names of functions. And actually they're not bad at doing that because they kind of know the vernacular of how we've named functions. They also know kind of what's out there in the world. They kind of have seen code and Wolfram language and every other, you know, in every sort of lower level programming language, et cetera, et cetera, et cetera. And they've, and they've also kind of see what people talk about and so on. And so sometimes they can say more or less, well, this is a thing that is sort of a lump of computational work that people identify and talk about and they can suggest names of functions for our language. And so that's really a use for these things. I think that the big sort of the take what's out there right now and find pieces that. Sort of are new lumps of computational work that might be relevant for the language. I think that is a potentially useful direction. Another thing that I think is potentially useful is there are all kinds of libraries that people have produced for all kinds of purposes out there in the world. And there's sort of a question of can we connect those in a meaningful way to our language and can we sort of take the design principles of Wolfram language and apply them to these other things? Because if we can, then we can give people tools where they're used to the framework we've made, the very coherent framework I like to think that we've made in Wolfram language and can they pull in this other random library, for example, that would be quite difficult to use in its native environment, so to speak. But if it's been glued into Wolfram language with the right sort of design framework around it, then it becomes easy for people to use. And there's a question of whether LLMs can help us to make that connection, can help us to take, you know, oh, here's this thing, and it's this library and it's got all this documentation, and you really have to dive deep into some complicated set of data structures and other things to be able to use this library in its native form. But, well, now it can kind of. The LLM can kind of figure out how does that connect with things that are part of the sort of usual Wolfram language framework for designing things. If it can make that connection and maybe even write candidate code, then it becomes a thing that one can just build, and then one can have something that is available, for example, from our function repository, where it can, underneath, it can call that library. I do think, by the way, that that's one of the important use cases for LLMs and AI in general is kind of the offline use case. I mean, there are plenty of online use cases which are things like, you know, pick which websites I might go to based on the query I just made, or do write some essay for something that I now have to write right now. But there's a different use, which is I'm trying to do something which I'm going to burn into a system that's going to be repeatedly used. But I design that thing using an AI, and where there's typically a human in the loop, where the AI is suggesting something and it's like, the human is like, yes, okay. And the human might go to some effort because it's going to be something that's repeatedly used, but still they're assisted by the AI. And I think that's a thing, whether it's in writing code. I mean, Wolfram language happens to be a sort of uniquely suitable case for that, because the language is one which is intended to be read by humans as well as written. So you can kind of have the AI produce Wolfram language. You as a human can look at it and say, is this what I want? Yes, it is. Then it becomes a thing that you can use as kind of a brick to be part of something that you write. I have to say, I don't have the impression that we've had as much use for that as one might think, partly because in Wolfram Language. Our whole goal has been to automate the process of programming as much as possible. So you use some low level programming language, you're going to have a big slab of boilerplate code just in order to read that image in from a file or something, or just sort of the throat clearing before you can even write anything. And that's something that the AI can easily trot out for you. We've spent a lot of effort to make it be the case that the things you write in Wolfram language are as directly kind of the way you think about it and minimally communicated to a computer as possible. Let's see, I should probably wrap up soon here, but Des asks, do you ever worry you'll end up like Wittgenstein, solving philosophy and the boundaries of science with an innovative math related system, then a couple of years later decide you're wrong? There are many ways. You know, one of the things about historical figures is, you know, any one of us humans has a whole spectrum of different characteristics. When you said end up like Wittgenstein, I was thinking about Wittgenstein's time living in a small house in a fjord in Norway. But that wasn't what you were talking about. This kind of we go in one direction and then we find out that we're wrong and we go in another direction. I have to say that I go in a direction and then I realize it was just the wrong direction. Has not happened to me very often in my life. And maybe that's luck, maybe it's lack of risk taking on my part that it should have happened more often because, you know, it's a sign one isn't taking enough risks if one doesn't have more disasters, so to speak. But I would say that, well, one general thing that's probably true is that anything, in my view, anything one does well enough, there is stuff that gets done in that process that is worth having. Even if some of the direction by which one does it, the sort of original reason for which one did it may not make any sense. I mean, for some reason I'm thinking of Kepler here where Kepler had a belief, sort of theologically induced belief about kind of the perfection of the planets. And he thought that, oh, there was this kind of collection of, of polyhedra that were contained one in the next and so on that would sort of nested to produce the particular orbits and so on of planets about that he was completely wrong. But nevertheless, he did a good enough job of the math of studying planetary orbits and things like this that he came up with Kepler's laws and he came up with a lot of very valid celestial mechanics math, so to speak. So that's a good example of even though the overall objective is kind of crazy, the doing of it is done well enough that there's a lot of very good content that comes out just as a result of the doing of the thing. And I think that's probably, I'm trying to think of an example in my life where that's happened. I think another general thing is that, well, there's the question of whether you. There's a question of whether you have a strategy that allows you to get to something. Then there's the question of where you say, I know what the answer is going to be. Let me find a path to that answer. When one's doing science, that's kind of a deadly point of view to say, I kind of know what the answer is, let me find a path to get to it. That's the way that one ends up making terrible mistakes in science. It's kind of like you have to actually do the science and see where it leads and kind of see where the chips fall. And it's an important psychological thing to not get so invested in the. And I know what the answer is going to be that if the chips fall in a different direction, you're always kind of pull the chips over and get them to the direction you want. On the other hand, you have to have enough idea where things might go that you actually go in any direction at all. Because if you're like, well, I have, you know, I can't imagine, you know, I figure this field has been worked on so much, there's not going to be anything interesting to say. Then you're not going to do anything. So you have to have some kind of, I know what direction I'm what, I know where to start, so to speak. Even if I don't know where it's going to end up, that's an important thing. And I suppose for myself, I've been more concerned in many cases with the this is how to start than in knowing where it's going to end up and guessing where it might end up and being wrong about it. Like for example, I might have said about our physics project, I did not think the physics project would work nearly as well as it did. I thought that we wouldn't be at the place where we are today for, I don't know, 50 years, let's say. And so some of my kind of longer term thinking about what one might do in the project was based on that assumption. But nevertheless I kind of knew what direction to start going in, so just started going in that direction. And it turned out we found out we could get a lot further than we expected. And so that's a thing where sort of the, in a sense, I suppose some wrong planning was done a little bit, but not much in terms of, well, okay, let me give some examples here. So there have been things where I have somewhat changed my mind about how things work, things about the kind of foundations of mathematics and what mathematics really is. And the things where I was very confused for a long time, for example, about why human mathematicians don't run into undecidability and other kinds of computational issues more often than they do. And it took me a long time to realize that sort of the fundamental reason was something to do with kind of how humans pick the mathematics that they do. And I have different ideas about how that might work that I don't think are right anymore. I suppose even in physics I had things where I was sort of prejudiced against the kind of multi way approach to thinking about quantum mechanics. And it took me a long time to kind of overcome those prejudices and really take seriously the possibility that could be how things would work. And then to realize actually it turned out to be much better than I'd expected. I think trying to think about other types of things on technology. Yeah, there are bets one makes. I mean like for example, one early 1990s, I had seen early virtual reality systems and I was like, oh, this is going to be great. Well, I didn't know it was going to be great. I thought it was going to be interesting and I thought we should explore it because it kind of was a, was a worthwhile potential direction. Turned out to be a bust at that time. And the people who worked at the company at that time who still work at the company where I'd been pushing let's do VR, they're like, eh, it didn't work in the early 1990s, why do you think it's going to work now? But that's a case where I would say that had we developed some things that were sort of high enough, were sort of where we had gone deeply enough into what was going on of coordinate systems for different, for images and things like this, that would have been worthwhile even if VR had been a bust back in the early 1990s in that particular case we didn't get far enough into it to, to get to that. But in terms of, I don't know the history of Wittgenstein well enough actually to know what particular directions he took. I mean, I think there are people, different sort of paths for careers. There are people who do these big arcs of projects. There are people who just sort of who more are keen on taking the next hill and doing the next kind of next particular thing. And I suppose for myself I've been much more of a large arcs of projects kind of person. Although I suppose to some extent I have managed to find that the problem with doing that is you can end up with projects where, oh, this one's going to take 25 years. And it's really hard from a personal point of view and also an organizational point of view to say I'm going to take 25 years to do this. There's a tremendous tendency by the time you're a few years into that that you've really lost track of what the point is and you're not being well anchored to any kind of sort of real strategic objectives. And so I would say that for myself, one of the things I've used I suppose is I write all these pieces. I suppose they're sort of generalized blogs and I'll write one about one thing, one about another thing. For me, those provide useful waypoints in potentially rather large arcs of general activity. And so I think that's sort of my attempt to kind of bridge the gap between sort of these vast multi decade arcs of projects and things that are much more short term. And I think there are plenty of people I know who've often done lots of interesting things where they will just produce a paper every month or two and it's about different things at different times and it's very statistical and sometimes they'll have one of those papers will be like a fantastic hit, a career making hit and others will just be completely forgotten. And it's kind of like you just. It's sort of a shotgun type approach. I suppose in the early part of my career I did that a little bit in particle physics and so on at a time when that was a very hot field. And yeah, you get some hits and some misses and it's very hard to predict which will be which. And it's even not clear if you look at the kind of long timescale, the things, there might be some paper one wrote or something which at the time when it came out nobody cared about it. Twenty years later you find out lots of people have discovered it and found that it's interesting. So I suppose those are things that can Happen. But I think this question of sort of people who write one paper about this, one paper about that, one paper about that, or something like this, and they're all very different, that they have less of this, oh, I went in the wrong direction. Let me kind of go in a completely different direction, I suppose. Yeah, it hasn't been a thing so far that's happened to me that the sort of some big direction has turned out to be the wrong direction. I'll give you an example of one that is. Yeah, I'll give you a couple of examples here. So one is, do you have, as a business direction, do you have a private company that doesn't have investors, you don't take it public, you have a sort of simple business model where you make stuff and you sell it to people. Or do you have something that's more kind of, in some ways I see it as more speculative of, you know, you've got investors, you're not really making money, but you're kind of getting lots of people to use your stuff and you kind of hope that one day you'll have ads for other things that will pay for that and you'll have this whole kind of complicated business structure. For me, I have decided to do something very, in a sense, conservative on the business structure side, if very innovative on the content side. And maybe that's a mistake, I don't know. I'm constantly wondering whether there are different things one could think about in terms of software distribution and so on. There are all kinds of questions about how do you license things, what's free, what's not. Could you, at a time when there were NFTs and blockchain was bigger, was a question of could you set up tokens where you reward people for contributing to the ecosystem and have that be a way that kind of people get software and so on. I didn't really ever manage to figure that out to the point that I was satisfied with it. But that's the thing where maybe one day I'll say no, it was the wrong direction. I should have had a company where we brought in a gazillion dollars of investment money and we weren't kind of concentrating on just make stuff where we can make money from it type thing that might be a thing. I've wondered about this many, many, many times whether sort of there's a different path every time. I kind of poke a little bit at that path. When we're doing spin off companies and things like this, I'm usually like, it's kind of like, no, not for me. There are issues about doing it that are not a good fit. We're talking about kind of interviewing and when people are good fits for things, I think there are aspects of that where I'm just not a good fit for running a business that kind of way. I don't want the. I sort of don't want the complexity and stress on the business side of doing those kinds of things. I rather prefer to have sort of the thing where I'm being innovative and so on. Be on the content side. I suppose that's one example. Another, there are other ones which are more kind of decisions about how one does things. I suppose another one would be, I've not been in the business of, for example, writing academic papers for a long time. And realistically, I'll write something that may be quite technical, but I really like the fact that when I write something, I can write it at a level of I just want to explain what's going on. I don't want to be hung up on the formalities of it's got to sound like an academic paper. Because there may be things that I want to say that are kind of. That are very useful for somebody understanding what I'm talking about, but they're kind of informal types of things to say, and they're not things I could kind of couch in the sort of formal language of an academic paper. I also think that, realistically, the distribution channels that I have for things I write are much better than typical academic paper distribution channels. And 100 times more people will read the things I write in the venues that I have, then will read it in your average academic journal and so on. I also think that. But. So that's been a decision I've made. And I haven't written a sort of ordinary academic paper since 1986. A bunch of things I've written get published after the fact in academic journals, but that's a different thing. But I haven't sort of written four academic journals since 1986. Is that the right decision? Is that the wrong decision? I think it's the right decision. I could be. It's sort of funny because lots of people will run into me who are kind of academically oriented and they assume that I've been publishing papers all these years. And it's like, nope, not a single one. And it's funny on the other side because they're people that I know where I kind of know some things they've been doing. And I'm kind of. I don't know Whether these people have been publishing papers or not, I have no idea whether they've been more doing science communication or more doing some business they're running, and whether they've been had some mill that's producing papers on the side, I just don't know. And not knowing is kind of a sign that maybe the value of producing those things is not as high as it might have been. And that's a large part of the decision that I made not to do this. I mean, the other part is that, you know, the world of academic papers, unless you game it, which is not so hard to do, but is a world of sort of incremental kind of things, not a world of big new ideas. And it's kind of almost ironic when I used to do that, that anything that I would write that actually had a serious new idea, the sort of. The peer review system would say, oh, no, no, no, no, it can't possibly be right. And you take some, you know, by pushing a bit and gaming it a bit, it's like, okay, fine, whatever, you publish the paper. But it's still just as a matter of the kind of. The generic response is very. It's just not a system that's geared towards two things. That's geared towards optimizing exposition and that's geared towards optimized innovation. And since both of those things are things that I'm very interested in, it's not a great fit. But again, it's one of these things where I could decide at some point, oh, my God, I made a terrible mistake all these years in doing that. And I suppose there are other things. What other structural things are there? I mean, there are all kinds of things. I mean, there are things like one can decide, oh, I don't know, you know, should I play professor somewhere? You know, I do it in kind of my. Through our summer schools and things like that. I have this intense period of a few weeks per year where I'm doing that. But, you know, should I be signed up with some university and going and giving a lecture once a week there? You know, I've not chosen to do that. I do live streams as some alternative. I suppose I could be giving a lecture to some class at some university. Instead, I've chosen not to do that Again, it's one of these things where one makes these choices and maybe one decides in retrospect that that was a mistake anyway. Okay? And I'm being reminded that I need to go back to my day job because I am trying to get some things done for our upcoming version of Wolfram Language, which hopefully you all will have a chance to partake in. All right. Better go. Thank you very much for lots of interesting questions. They've gotten me to think about lots of things. Always a good reason, at least for me to do these live streams. So thanks and bye for now.