Business, Innovation, and Managing Life (February 21, 2024)

Sam foreign. Hi everyone. Welcome to another episode of Q and A about business innovation and managing life. So let's see. I see a bunch of questions saved up here. One question from Perry. Can business ventures be a one man show, one person show perhaps, or is it a requirement to have a team? Depends on what you're doing. There are plenty of activities where really the business venture is using the expertise of one person. Give a, give a, A simple example is like writing a book or doing the things around writing a book, creating a, a, you know, a YouTube channel, something like this. You can do that as a, as a one person operation. Perhaps if you're writing a book and you're publishing it, you write, write the book and then there's a whole system that you know of, whether it's agents or publishers or printers, whatever, who get the book out. In the case where you're doing something like writing a book, you're not expecting that you're going to be the one running the printing press, so to speak. That's a, you know, there's the piece that you contribute and then there's the sort of downstream things that the rest of the world contributes. Same with, oh, I don't know, putting up videos, you know, you make the video, but then all of the mechanics of serving the video and, and, and all those kinds of things is not handled by you. So I think there are plenty of businesses where it's kind of a, a one person's creative efforts drives what's going on. In the case of consulting, for example, it can often be the case that you have a, essentially a one person company, a one person business, because it's the expertise of that particular person that people want. And if they say, and often the sort of cycle of these things is there'll be somebody who's a very successful, I don't know what sort of, the consultant of some kind might be a business consultant, might be an interior designer, might be all kinds of different things. And then as they scale up and they say, well, I've got too much work, I'm going to get other people to help me. The service gets worse because the thing that was the person who really had the high skill was just that individual sort of principle. Now again, in many of these things there are sort of, there are people where you need, oh, there's the person who's going to do the legwork and the help and the, you know, assisting and setting, setting up meetings and this and that and the other. In the modern world, an awful lot of these kinds of activities that are sort of the ancillary activities are outsourceable. In some ways you kind of have to make a choice. It's like if you're setting up a server system, you can either set it up yourself, which takes a fair amount of effort, or you can use some cloud provider, in which case you're probably spending more money, but you're not doing that work yourself. So I think particularly in the modern world, there are a lot of things that are pretty much a one person story. Now. It depends on what kind of a business you're doing. There are businesses where it's completely unrealistic for it to be a one person thing just because the complexity of what you're producing is too great. That's one reason. Another reason is that simply, for example, the selling of what you're doing requires human contact beyond what any one person is going to be able to do. I mean, one of the things that happens in businesses, particularly sort of creatively defined businesses that involve sort of intellectual, creative effort being spent and that's where the value is coming from. One of the unfortunate things that happens, I think it's unfortunate at least, is that you'll have a business where there's a sort of high skill person who is able to do this particular kind of thing. But the way it will end up is that that person is doing the selling of the business and then the actual work gets done by lower skill people who were sort of brought in to make sure that the work actually got done. And so you have this situation where kind of the person who really is the person most able to do the work is not actually doing the work, they're selling the work and the work's getting done by other people. I know in my company, for example, back when I started it 37 years ago now, I was concerned that that was the sort of natural thing that would happen, that I'd be the person who could explain the best what we were doing and sort of sell it the best. I decided that was just not what I wanted to do. And so I kind of made it a, a kind of a, a principle of our company that I'm not going to do that, that will have sales, business development, whatever that's doing that and they'll do what they can do and I'm not going to do that. And hopefully, and we've been successful the last 37 years making a business in which I'm not doing those things. And. But that's sort of a conscious decision and a somewhat unusual one because the more usual cases that sort of the person who knows the most, et cetera, is the person who does kind of the front line, the sort of the high value selling and so on. Well, in terms of other reasons to have a team, I mean, one is just the complexity of what you're trying to do, the number of sort of pieces of human contact that are needed and so on. Another is that you sort of find it easier to be productive if you are sort of bouncing ideas off other people. And if sort of the energy of the group of people is greater than the individual energy of the parts, so to speak, that can happen. The opposite can happen too. If three people all have a great vision for what a company should do, but it's different, that's not a very positive thing. You'd be better off with one person and just pursuing their vision and not sort of being pulled in multiple different directions by other people. So a few comments there. Anyway, let's see. Gosh. Greg is asking if I wanted to start a blog. What advice do you have for this process? How do you maintain your writings? Well, I mean, as a practical matter, you know, there are endless blogging platforms from, you know, WordPress on, where you just can sort of put the words in there more. The question is, how do you actually produce those words? What and how do you get people to care about them? I would say there are a lot of other, you know, whether it's medium, substack, I don't know, a lot of these services which kind of give you the cold sort of packaged experience of produce this thing that gets put out out there. And some of them let you give you good ways to monetize that if you want to do that. But I think more to the point is kind of how do you write things and what causes you to write things? My attitude for what I'm doing is I write when I have something to say. I have no sort of quota for myself of oh, I'm going to write a piece a month or whatever. And in fact, if one looks at the, you know, the posting dates for the things I've done, there are sometimes, you know, four or five month gaps and then there'll be multiple posts that come out within a week or two of each other. I would say that what I've done is I tend to my posts are a side effect of projects I do and a concretization of the projects I do rather than it being the case that the posts themselves are the, the end in themselves, so to speak. I'm not sure that I've optimized things in that regard. One of the things that I found is that I tend to write these. These. I, I say I'm going to do this project. I'm going to, I'm going to end up writing about this particular thing. But then it becomes very meaty. It's kind of like, well, I'm. I'm writing, you know, I'm. I'm going to write about this. And, and for me, I, I tend to write once about each subject. There are other people who will write their view about the subject now, their view about the subject in three months, in three years, et cetera. I've tended to be. I write this once, then I move on to something different. And again, it's a question of one's personal goals and workflow. But I often end up writing these really long, sometimes multiple hundred pages in the end posts. And I'm not sure that's, you know, that's, that's optimized for my objective of sort of concretizing what I've done in a project and explaining it to people. It's not optimized from the point of view of kind of easy reading for a blog or something like this. I think the, one of the things that people. It's often an issue is sort of what's this blog about? And you know, do you have a very definite kind of angle that you're trying to communicate? Do you have a definite kind of point of view that you're trying to get across? Or is it just more of a kind of personal life travel log, so to speak? Here are different kinds of things that I've done. For me, what I write is more the second kind of thing. It's not, you know, I'm. My cause is X and I'm going to write a sequence of things in support of that cause. It's more. These are projects that I'm doing for all sorts of different reasons. Whether it's a project that is a piece of history that came up because of some anniversary, or it's an obituary, or it's a thing that came up because a product we're releasing, or it's a thing that came up because of some science I figured out how to do, or just a thing that I was curious about that I dug in deeply enough to have sort of done a project about it and written about it. There are, for me, there are lots of different sources. And again, I'm not sure that optimizes kind of sort of the, the Making a thing that will be consistent so people will know to read it and so on. I'm, I'm more, I'm more trying to sort of explain projects I've done than I am trying to sort of build the readership specifically in that way. But I think one of the things that can happen is you kind of have to develop a style for the kinds of things you're writing. And that's something I have sometimes found somewhat challenging. And I've gone through multiple different styles in my life. When I first started writing things for publication, well, not quite first, but first that was sort of mature works, let's say it was academic paper type things and those have a certain style. They have a rather dry style and they are, you know, I tried to write them with more sort of expository effort than people often put into those things. That wasn't always met with, you know, when I would publish these things in academic journals, people would say why do you have this big introduction to this thing? You know, just cut that all out and. Which was a mistake because the fact was the introduction was the part people most wanted to read because it was the part that was really doing exposition of the field that is really useful. But in any case, then there's that sort of, that style of writing. Then I wrote things like computer documentation for systems that I was building. And that's a different style of doc of writing that and then some academic papers and then books and so on and then these kind of initially blog like writings which I think are at best a super blog now so to speak, because they're much more meaty than you would typically find in a blog. And in each of these cases I was sort of developing a different style. Like in the case of things that I write now and you know, post in blog like form. The, the thing that I like about that is it's a, it's a format where I feel like I can say anything. I, I don't have to be either incredibly formal or incredibly informal. I can have kind of, I can even, I can have sort of whimsical parentheses and so on. I can, it doesn't, it's not, it's a very unconstrained format, which I like a lot of. But I think for everyone one has to sort of develop, you know, how roughly is one going to do these things? Now there are people I know where everything they write is two pages long and that's. I think I'm. Sometimes I feel like, gosh, if I could only do that because for me, it's sort of, in a sense, easier to write the 30 page or 50 page thing where I'm kind of explaining what I'm talking about, than to try and condense it all into the two pages where I worry that, wait a minute, people are not going to understand this, oh, what about these other things that I figured out? I'm never going to have a chance to say those, etc. Etc. Etc. So it's. But as I say, the, you know, I think the other thing about starting writing is to just do it. As in it's, you know, you could say, well, I'm going to write my great works and I got to figure out how to write my great work. And 10 years go by and you're still trying to figure out what's going to be the first page of the great work. One of the problems can be, and sometimes it's happened to me this way, is that you build up so much kind of focus and importance on the thing you're writing that you end up putting it off for sometimes for years. Whereas if you just said, I'm just going to write it, I'm going to dash it off, and maybe this one and maybe not very many people are going to read this, so don't worry about it. Dash off that one, you dash off another one, you develop some kind of style. And, well, it may turn out that people really like what you're writing and people discover it and you suddenly start getting a big readership. But you've kind of worked into that gradually. And so there isn't this moment of I've got my blank notebook for me, my blank document or whatever, and what am I going to do? What are the first words of my great work going to be? That to me, is often a killer. The first words of the great work are very hard to write. Unless sometimes, sometimes for me, I will have a. As I conceptualize something that I'm going to write, I'll have these first few sentences rattling around in my mind. Like these sentences have to get out. And that's an interesting place to start from. I have to say, sometimes when I have the sentences that just have to get out, I put those as the first sentences and then I keep writing from there. I keep writing from there. I realize, oh my gosh, those first sentences pointed me in the wrong direction. I really have to throw them away. And that's a difficult thing where you can have those sentences that you've been saying, I'm going to say this about this, and you realize when you really think about it more carefully, that wasn't the right place to start. You know, I think one of the things for me is sort of the starting of documents and even the starting of sections of documents. It's always like, I. I want to kind of have both. I want to kind of front the main point. I want to be pointed in the right direction, and I want to do that from the first sentence, so to speak. And that's often quite challenging to do. And I think the. But for me, the idea that just start writing that section without having defined what the direction is going to be doesn't usually work very well for me. It's usually better for me to spend the modest amount of time, maybe it's 10 minutes, half an hour, something like that, sort of noodling about what those first few sentences are going to be as a way to define the direction for the whole thing. The trick is not to noodle for a decade on what the first page of the great work is going to be. Let's see, the question here about how do I maintain my writings? I think, you know, I usually. Okay, there are two different meanings to that. I mean, one is I tend to work on a project. I begin it, I grind through it, I get it finished. For me, if there's a project where I stop it, I put it aside for a while, then I come back to it. That's super inefficient and super unpleasant for me. I mean, I find that it's. And the result is usually worse. In fact, you can kind of. You can often, you know, when I've done that, you can see the scene, so to speak, where I stopped and restarted. And for me, it's much better to just take the thing, make sure I've carved out enough time, get the thing finished, then go on to the next project. For me at least, particularly with writing projects, it's difficult to do multiple ones in parallel. Right now, for example, I happen to have wound up with three writing projects in parallel, and I'm having a hard time with that. And I'm really having to do kind of the stick slip type behavior of, you know, I'm writing one thing for a few days and then I'm switching to the other one and so on. In terms of how does one sort of keep writing things? For me, as I was saying, I keep writing things when I think I have something to say, otherwise I wouldn't write anything. And for me, I'm a little bit in a trap, which I don't necessarily recommend to people that if I have a very tiny thing to say, I'm really not going to wind up writing about it. It's just not going to happen at all because I've sort of developed a style where everything I write is fairly meaty. I mean, perhaps I might do some tweet, like post that's really short about some small thing that I want to say, but I don't really have a good intermediate scale between that and. And kind of a really, really meaty multi tens of pages type document. Miss Leisler asks, are there blogs are currently working on? Yeah, there are three things I'm currently writing. One is a piece about can AI solve science? Which I think I've talked about a little bit on these live streams before. I'll have more to say when I've finished this piece and have sort of fully formed my views about things, but that's sort of. A question that people keep on implicitly asking me. And so I figured I'd better write down the answer. It's something which is sort of a mixture of actual practical computer experiments of how far can you get with this or that? And sort of almost philosophical arguments about what's going on. So that's one thing I'm working on. That's the active thing of today, for example, Another thing I'm working on is kind of a study that I started about five years ago that has to do with a case where it's to do with the game of life cellular automaton, where you can. There's sort of 50 years of history, 53 years or something of history of people sort of constructing, making constructions in this kind of array of bits that represents the cellular automaton. It's about the same age as the microprocessor industry. But in the case of the cellular automaton, every innovation of people figuring out how to build, you know, things that spit out streams of localized objects or make, you know, Turing machines or print out digits of PI or something like this. All of those constructions are all just defined in collections of bits. And so what I was interested in is, is this a case where you can really quantify the kind of process of innovation. If you want to ask about innovation in the microprocessor industry, it's kind of like, well, there are all these things and there's all this kind of these inventions in physics and the discovery of this effect and so on and so on and so on, and all these questions about supply chains and all sorts of things and commercial issues and so on. And, but so it's not quite as raw to see, oh, people invented risk architectures and then this happened. One has the intuitive sense that there are certain kinds of phenomena, like a new methodology is invented and then there's very intense innovation for a short while after that, and then it levels off and then things are rediscovered and so on. One has these impressions from being in the technology industry. One has another set of impressions from looking at kind of biological organisms and biological evolution. My goal here was just to try to find in this very specific case where you really have all the data to be able to just sort of see what happened. And as is often the case with these kind of data intensive projects, the data is not as clear cut as you might have hoped. So it takes more effort to kind of, you know, extract the science from the data, so to speak. And that's what I've been working on there. And my third writing project is this book and course that I embarked on about introduction to computational thinking. Kind of the goal there is, I view sort of computational thinking as the fourth, probably sort of great kind of formalization of the world that we humans have made from starting with human language, logic, mathematics. Computation is another sort of great formalization of the world. And the question is, how do you think about the world computationally very different from how do you write code for a, a programming language, for example? It's how do you kind of think about things in the world in the sort of systematic computational kind of way so that, by the way, you can use computational language to express these things to a computer that can then help you get further. So that's a project I've embarked on and actually I will be posting some things about that because I think I'm going to post the incremental pieces as I write that, both because I'm curious about the feedback and frankly because I don't want to just disappear for a long period of time and then come out with this giant piece of output. I would prefer to kind of provide it incrementally to the world. I think it's easier for me and it's probably better for the world. So that's what I'm planning there. Okay. Aber asks the question, how do I think our company would be different if it was based in Silicon Valley? Well, I have to tell you a funny story here. So I started our company in Champagne, Illinois, because I was there, because I was. I'd started a research center at the University of Illinois and was sort of a professor there. And very quickly after I went there to start this research center. I realized that sort of the research center running activity was probably not the best way to optimize my kind of life path. And so I decided to start the company and I started it physically in, in Illinois, which turned out to be a very good headquarters place for us. It's a, it's a place where kind of the, the, the quality of talent is, is, is good. The kind of the, the general sort of approach to companies and work ethic is very different from what you find in Silicon Valley, for example. And for me, in my efforts to build kind of a long term company, it was a good fit. But sometime I think I started the company in late 1986. Sometime I think it must have been maybe late 1988 after we released our first product, Mathematica, maybe It was early 19, 1989, I was like, you know, maybe we should think about moving this company to Silicon Valley or somewhere near Silicon Valley. And because I thought, you know, that's my original theory of the company was this is a thing that I'm sort of growing for a while and then I'll take it public or whatever else and then it's kind of on to the next thing that didn't, it hasn't worked out that way. Instead of, I'm 37 years into this and it's worked out really well for me to just have a long term stable sort of entity continuing to build, continuing to sort of focus on long term things. But in any case, back in, I think it was late 1988 probably I was like, maybe we should move this company to Silicon Valley. There were all kinds of mechanical issues about doing that, but I was considering that. So I happened to be out in the Bay Area and I thought, you know, let me not, you know, actual Silicon Valley was already very expensive. It's like, let me look at a, a place that's near Silicon Valley. Let me look at Half Moon Bay, which is down on the coast over some sort of hills and so on. So I go there and the particular day I was there, it was unbelievably gloomy. And it's like, no, there's no way, you know, if we're going to go to California, there's no way if it's going to be this gloomy. This isn't for me. I don't know how gloomy it usually is there. But that was the kind of trivial thing that sort of turned me away from that idea. I think the idea might not have gotten legs anyway, but I Considered it. I think that things Silicon Valley has, in the time I've known it, which is now what, 40 something years, has had ups and downs. You know, it's, it's, it's, it's had quite a few near death experiences. You know, is the software industry going to go to Utah? Is it going to go to, you know, Seattle? Is it going to go to this place? And, and you know, there have been a series of times when it looked like Silicon Valley was kind of on its way down. Even Silicon Valley versus San Francisco, that's a, that's a different thing. But I think the, each time it's recovered and by now it has enough strength that you know, it's, it's here for the duration, so to speak. But I think particularly at times when it's sort of been on the up, you know, it becomes a very superheated place. I think that's true of a lot of towns where an industry really sort of takes hold and the industry is doing really well. And it's the car town, it's the movie town, it's the whatever else, it's the finance town, whatever. And you know, things get very superheated and I think that leads to lots of funky expectations and so on. I remember a friend of mine who was running a company in Silicon Valley. This must have been, oh, probably what the early to mid-90s telling me, you know, the day that the person who was coming in as a, as a basic receptionist, I'm not sure when he even has those anymore for his company. It's like the person wants equity in the company. This is crazy. And that was sort of a, you know, there's a certain mode of expectation for these things. I think that the thing that to me is, you know, there has been in Silicon Valley something of a, kind of a, a. There are people who are like, well, I'm at this company today and oh well, maybe I'm going to go to this other company which by the way is just on another floor of the same building and I might be there tomorrow type thing. And you know, there is not the, the kind of, the people sort of making a long term investment in one company that there's been, there's been trouble with that, let's say. Also obviously it's become a very expensive place. It's become a very expensive place to live. The expectations for salary expectations are very high. You know, in our company I think we have essentially no employees in Silicon Valley because it's just like there isn't a reason to be sort of paying the premium for what's there for a company like ours. I think that, you know, it's often the case, you know, if you say I need a such and such programmer, well, you know, I'm going to run into one at, at the coffee shop type thing and I, I'll be easily able to find one. That's, that's not a thing you're going to find in, you know, you'd find that in Silicon Valley, you won't find that elsewhere. I think that the question of whether our company would be different if it was in Silicon Valley. Yes, I'm sure it would be and I'm sure it would be the fact that the company, even existentially the fact that the company has had this sort of long term existence. I think in Silicon Valley it would be more of a, you know, can it ramp up and get to be 10 times as large and so on and sort of go to the moon or die type thing. I think it's a different, a different set of expectations. I mean, in my particular company we don't have outside investors. That's another very big difference from the typical pattern in Silicon Valley. The thing I have seen though very recently, last couple of years is an increasing number of sort of more savvy, often sort of not first time entrepreneurs saying, you know what, they come to me and they say, you know what? I want to do something like your company. I don't want to have investors who are going to sort of force me to be up or out type thing within a limited number of years. I want to run this thing as kind of a long term sort of proposition. And so they're, they're kind of looking for, you know, you can, in a sense it's not as superheated, you can't, you can't do as much as quickly. But on the other hand, it doesn't burn out. It's something you can keep consistently, consistently running. And that, that seems to be, I would say there's a, a sort of a micro trend that I've seen developing of people who are sort of doing sort of closely held private business tech companies that has been very rare in the past. And I have to say, reflecting also on the question was asked earlier, the, the fact that one needs, oh, one needs all this investment money, et cetera, et cetera, et cetera. Yeah. If you're building a, you know, a microprocessor fab or something, then then absolutely you need tons of investment money to do that. If you're if you're building physical infrastructure, something like that, or even. But, but for many of these things, which are more in the kind of, you know, creating it with sort of purely with software and so on, you don't need often a huge staff to do that and you don't therefore need sort of all of the energy of all of the kind of venture capital and so on to start that way. I feel like often, you know, companies defeat themselves by getting too much money and, and they, they end up. It ends up. Sometimes you need it in order to get over some threshold to do some particular task, but often it's, it's not actually that positive. Let's see. Let's see. RBS Comments so my moral is follow the weather. Don't establish an office where there's gloomy weather. I. I don't know that. That. I'm not sure that I can, can necessarily claim that. I mean, I'm, I'm based in the, in the Boston area and it's, it's. If I, if I always followed the weather, that probably wouldn't be where, where I was. Let's see. Go. Let's see. Christian is asking, so you think complex and. Right, complex. Instead of simplicity as a guidance, does someone with a sense of simplicity fit into the company? You know, I would like to say that if I have, you know, people sometimes say my big attribute is X and actually the attribute is not X. But my particular kind of life mode, I think is I try and get to the essence of things, which is really a drilling down to get to simplicity, but then building up from there to build the elaborations of that, that end up being some big engineering structure or some large description of things. So I think the keep it simple thing is very much my intellectual tendency in terms of sort of the core ideas. What gets built on those core ideas may be a very big thing, but the core ideas I try to make as simple as possible. That's certainly what has happened with kind of the design of our computational language with a lot of science I've done and so on. In terms of our company, for example, I would say that one of the things that's interesting about our company is we really have a very, I think, diverse collection of types of folks in sort of all possible dimensions, which includes kind of a tendency to sort of be minimalist, a tendency to be maximalist, so to speak, and so on. I think that the thing that I suppose I can often value is what is the point? Today, right before this, I was in some meeting That I was very frustrated about because it's like, what's the point? What's the point? What are we trying to. What is the simple takeaway from this? Not all of the kind of detail and so on. And so that for me is a. It's an important attribute, though it's not the only possible attribute, so to speak. Let's see. A question here about from Arie asking, saying, I gave up on the idea of incorporating science in my business income, life. Curious about your case. Do you like the process of product production, commercialization, management? So I mean, I've been sort of lucky or it isn't really. I've tried to make my own luck in this regard of being able to integrate things that I intellectually find interesting with things that make a living, so to speak, and sort of have some footprint in the world. There are things I could have done where those two things wouldn't have come together in that way. I think it's a non trivial puzzle to fit those things together. I will say one thing that often helps me is that there are things where I realize we need to do this from a product point of view. And I'm initially like, oh my gosh, this really seems kind of dull. I don't see the intellectual interest in it. But then I start div into it and I realized that in this thing that I thought was kind of dull, by the time you actually think about it with clarity, it's like, well, there's this thing to do and that thing to do. And it's actually pretty interesting. And in fact, I often learn things from things that seemed like they were done for sort of purely commercial purposes that end up being of great kind of intellectual value. One from a very long time ago, I was consultant at a company that was making massively parallel computers. This is in the mid-1980s. And they had this computer, it's called the connection machine computer. And it had these very large number of very simple processing units. And frankly, they had a hard time figuring out what on earth you could do with this thing. And I was interested in cellular autom, where you have very simple rules, a big array of different elements all sort of following simple rules. And I realized, well, this is a good fit for their computer. And then I realized, well, so what do people use computers for? And they used sort of supercomputers at the time for things like fluid dynamics computations. And so then the question was, well, I'd sort of thought a little bit about could you use things like cellular automata for fluid Mechanics. But then I kind of got, you know what, I'm just going to figure this out and figure out whether this computer can be used for this purpose. And so I did. And that turned out into a very interesting piece of basic science that actually in the end is an important building block for what's become our physics project in very recent years. But that was something that was originally, it was like, well okay, I'm supposed to be doing this, I'm trying to figure out, I'm trying to help this company let me figure out something that's of practical use. And it's like, well actually that turns into a piece of basic science that's super interesting. And that's happened to me many times. So you know, the thing that sort of, I thought was kind of a bit dull turns out when you do it well, it has interesting things inside it. You know, for me the question of kind of how can I take things I'm doing and kind of productize them and so on. I mean not everything I've been doing done is I know how to productize it. I mean things about the fundamental theory of physics, I thought that was not really productizable. It's turned out that there are a bunch of implications, particularly in distributed computing and so on that are much more near term than I ever imagined. I thought this was going to be a multi hundred year story to get to anything that was applicable. But that's not the case in fact. And I think the things, when it comes to kind of taking ideas, you know, I, I, I suppose another point is that when I'm sort of, it is for me very helpful to have this kind of matrix into which I am putting ideas. You know, raw free floating ideas don't tend to have a very happy life in my experience. It's when you have a matrix like oh, we're building this big product and it has many small ideas that can be sort of used innovatively or even when you have something you're consistently writing, you say I've got another idea, I'm going to write something about it that provides some kind of matrix for what you do with the idea. If it's an idea, that's not an idea which you would write something about. It's an idea that has to do with, I don't know, some thing that you go out in the world and do or something and you don't have a matrix for it. To me it ends up being a free floating idea that it's very hard to make it go anywhere. And I Think this question of when it comes to. So once you've built one of these sort of matrices for making ideas sort of take flight, that then the question is sort of to feed that and you put in ideas and you kind of have to think about, well, how do I turn this idea into something real in the world? For me, that's just one of these kind of analytical thinking exercises of the same kind that I'm doing in science. It isn't really that different as far as I'm concerned. It's kind of. Yes, well, you're thinking about the world and about sort of commercial positioning and maybe it's pricing and maybe it's some other thing about deployment and so on. But these are all things that are sort of analytical things. It's just like, just, you know, you know, facts about, I don't know, something about physics, for example, about how black holes work. Well, I also know facts about how server configurations are set up and so on. These are just facts one knows. And if one's trying to figure out how to do something with some, I don't know, thing that one's trying to develop, it's sort of wheeling in. The fact that one knows about how one should set up virtual machines to do this or whatever is to me very similar to knowing how one should kind of set up some metric for some black hole thing. These are both just sort of fact like things that one knows that one has to kind of wheel in when one needs them for some particular kind of thinking one's doing. So for me, I don't make a huge distinction between these different kinds of areas. I think people can sometimes get very. The kind of the thinking apparatus is only turned on in one direction or another. It's like, well, I can think about black holes, but I can't think about server configurations. But really at some fundamental level, kind of the thought processes that are going on have certain similarities. And I don't know whether I'm not sure that the precise black hole versus server. And I'm now starting to think about causal graphs of information flow in server networks and so on. And maybe there's even a closer analogy than I was at first imagining. But, but, and even the fact that one can imagine such an analogy is telling one something about the value of having those two sort of fact bases to work from. But you know, so I think, you know, I have to say in terms of kind of the life planning thing, you know, there are people who say I have this hobby on the one side and I have this Work on the other side. And that can work well for people. It can be, you know, it can be almost relaxing to have these different kinds of things that one does. I know for me, there are some different kinds of things I do, like these live streams which have nothing to do with anything I do in the rest of my life, so to speak. And I like this. I find it a good break from other kinds of things that I'm doing. But so, you know, there's one is the kind of the work and the hobby type thing separated. Another is, well, let me make my work the thing that I really like to do. And that's a thing that sort of requires a whole sort of puzzle solving thing. And I've been fairly happy with the way that I've been able to do that in my life. But it's a complicated, specific thing about what is it exactly that you like versus what is it exactly that the world sort of cares about and has on offer, so to speak? Let's see. Christian asks. Innovative companies need a culture of dealing with failure. Quantify failure for measuring innovative rate. You know, in our company, we don't have a lot of outright failure. And you know, what tends to happen is a project. First of all, maybe we're spoiled because I have a decent amount of experience and the number of projects we start that are just crazy is not large. Most projects we start, even though those projects can be quite ambitious, they're not crazy. They don't go in. They don't just end up in a total waste of time, so to speak. But the other thing that happens is a project that sort of gets a certain distance and then you realize, whoops, there's a block. We can't get past that block. The trick is to change direction of that project. Yes. If you said, we got a plan, we're going to follow the plan. The plan takes us off a cliff. Okay, we're dead then that's one thing. But when you realize there's the edge of the cliff right there, and you see the edge of the cliff and you're like, wait a minute, okay, now let's go in this other direction, you can often end up in a very interesting place. And that's happened many times to us. And I suppose one of the things that's probably true is that a certain agility of management is needed so that you don't just say to people, okay, you're marching this direction. Here are the plans. Just go in this direction. And they say, oops, there's a cliff here. And you say, well, this was our plan, so go march off the cliff. And instead, there has to be some mechanism for sort of getting that information back and realizing, whoops, we have to make a change of direction here. I have to say I'm not sure that even in all these years that we perfectly solved that problem. The way that I mostly try to solve that problem is by having enough reviews of projects that are going on and enough kinds of ways that I have some visibility into what projects we're doing that I can see, oops, this thing is heading for a cliff, or people are kind of making the noises that sound like we're seeing the edge of the cliff type thing. And then one's able to kind of figure out how to divert it in a different way. I mean, there are projects like, for example, in machine learning, it's a quintessential case where this happened to us a bunch of times, where it's like, can we make this work with machine learning? Okay, let's give it a try. Oh, oops, it doesn't work. That's a place where you could end up with, you really don't know how hard to push. You don't know, oh, if we pushed harder, we spent more money on training, we were cleverer about how we prepared the data, and so on, then we'll have a breakthrough. It's sort of weird in that respect, actually. I'm not sure that. I mean, to me, I feel like I have some intuition for what it feels like when it's going to work if you push it hard enough in that particular case, in many types of projects. I feel like I have good intuition for this question of did it not get there because it wasn't pushed hard enough, or did it not get there because you can't get there from here? That's an important question for projects. And how do you see the signs of it wasn't pushed hard enough versus you just can't get there from here. I think my heuristic for this tends to be, you know, I'm talking to the people who are at the front lines and, you know, sometimes I'll be talking to people and it's like, yeah, they're really pretty smart in what they're doing. And, you know, whenever I come up with an idea, I said, did you try this? They say, yeah, we tried that. And this is how it gets blocked, as opposed to. I talk to them and it's like they say, well, we're blocked. And I say, well, what about this? Say, oh, that's Interesting. What about that? Oh, we didn't try that. And you know, where when you poke at it, you find that there's sort of. They didn't try a lot of things, then that feels like a project where you say, well, you know, let's just think through what you could actually try. Oh, you get to a thing where they say, oh, that's an interesting idea, I guess we should go and try that. And they come back and say, yeah, we solved it, and off you go. But it's pointless to just tell people, just push harder without. I suppose this is a point, at least for my style of management is the just push harder and solve the problem is rarely a useful thing to say. I mean, in a sense you really have to work with people and say, what is the problem? What's actually going on? And if they say babble, babble, babble, some complicated jargon thing and you just say, oh, I don't understand that, you know, just push harder, that's kind of a bad sign. That's a kind of a, a, you know, that's a, a bad dynamic, so to speak. I mean, usually I, I take the point of view, it's kind of perhaps an arrogant one that, you know, I can understand anything if I try, if I put effort into it and there won't be anything where people say, babble, babble, babble, you know, and I'll just not be able to understand it. It's like, okay, can you explain that to me? And another thing that's important and can you explain that to me is you have to have very little. You know, if I don't understand something, I'll just say, I don't understand this and I don't care that somebody says, but that means, you know, how could you not understand this? That's in, you know, a first year such and such class. You know, you must be an idiot if you don't understand this. I think I feel like one feature of, I don't know, some kind of confidence is I just don't care. I mean, it's like I'm more interested in understanding this. I don't worry about whether the person that I'm asking, tell me about this elementary thing says, then you must be absolutely clueless. I mean, I suppose it helps in, in a case like mine that I know a bunch of stuff about a bunch of areas. And so it, it, you know, while there will be holes in what I know, it's, it's something where people, you know, you, you can clearly, you clearly can make Progress fairly quickly. Even if there's some hole in what you know, there's other stuff that you, you, you already do know. But so I think this, this idea of, you know, when projects seem to be floundering and people say they're impossible, you kind of have to listen to what the actual questions are and see whether you can help. And I found if I can't help, then I know why it's blocked. And then I have to say, well, let's redirect and do something slightly different. But we really don't get to the situation where it's like we're marching in this direction and oops, we got to the cliff and we marched off it. And so the project quotes failed. It's a very rare thing for us. I think it's much more common in companies where, for example, the management doesn't really understand what the people or doesn't choose to understand, doesn't make the effort to understand what the people kind of on the ground are going through, so to speak. That can be a much more of a, we have this plan. Go execute on the plan. Oops, we couldn't do that. So it's a failure. I think that's, you know, and I, I have to say I'm, I'm, I'm not a huge believer in this kind of set things up, you know, push very hard, and then, oops, it failed. Splat. I'm much more of a believer in, you know, if we put effort into it, let's have it, you know, get to some success, even if that success wasn't the original thing we were going for. I mean, I think it's also a question, perhaps it's a consequence of dealing with a company that doesn't have kind of infinite resources to try all possible things. Perhaps if you have enough investment money and so on, it's a better strategy to just sort of shoot off these missiles in different directions. And some of them may go splat. But if you have one that hits, then you're in good shape. And also, sometimes it's a question of what you're trying to do is not for the sake of having a success at a product level. You're trying to have a success at the level of convincing your investors to put in another round of investment or whatever else which has a different kind of purpose to it. Let's see, Tugboat asks, is that opportunity to be a CEO of a company that employs AI? You mean that employs it as a, as an alternative to employing humans? You know, I have to Say in the building of my company, the main thing that I suppose at a meta level that we've done is we build automation. And so, you know, we probably do, you know, we have maybe 800 people. We probably do. I would say in terms of sort of productivity, we might be more like a 5,000 person company or something, maybe more. Why is that? Well, it's because we built successive layers of automation, which by the way is key to the products we sell and lets other people automate as well as us. But this idea, I mean, in a sense we've been using the AI workers idea for, for three decades because we're progressively building automation that lets us kind of do things not with people, but with automated systems. So that's, you know, that's been a thing for, at least for a company like mine for a long time. Not always the case people, you know, for some companies it's like, you know, do you use robots in a factory or do you use humans? Sometimes it's cheaper to use humans. And depending on sort of the dynamics of your company and whether you're fed with investment, investment money and things like this, it may be cheaper to use humans than to try and build those automated systems. But you know, I think in, in terms of applying AI, gosh, I mean, there's, there's an interesting question, is you're CEOing a company that uses AI for something. How much do you have to understand about AI? I would argue quite a lot. I would argue that the CEO who runs a company whose operations they don't understand, that's not a good situation. In fact, I would say there are companies that have been decently successful in that regard where the company is sort of taken over by the, as people say, rather unkindly, the bean counters, so to speak. And usually what's happened there is that the company really has a pretty good thing going. It has some kind of cash cow type thing. And then you can kind of cut out some of the things that would be building for the future. But for the next 10 years, 20 years, whatever it is, you can kind of slim those things down. Then you have a much more profitable operation where you can just keep doing what you have been doing and the people in charge don't really have to understand what it is that's going on. It's just the feet keep walking, even though the head has been, you know, is off in the clouds somewhere. I think the. So I think my strong belief is, well, there's a question, do you need to be able to write code or do you? And the answer to that is no. Although to fundamentally say I don't write code as a matter of principle is probably a mistake. But I have to say I've seen many examples of people where the point is, do you understand the key ideas? And can you have a conversation with an engineer or something that gets at the key ideas? Even if when they say, well, what is the type declaration for this low level language for this thing? You don't have the slightest idea. That isn't the important part. The important part is there's a big picture of what can be achieved and being able to get into that and not just saying, hey, we want to make a product go off product people and make the product and don't and say just invent the product. I don't know, I don't, you know, I don't care how it works. I just want a product that sells for X amount of money or something. It's shappy is asking how did you achieve product market fit? You know, I'm very bad on these business jargon terms and somebody was just telling me about things about how, oh, you know, this or that aspect of what we have but doesn't achieve product market fit. And I'm like, what, what are you actually talking about? And it turned into some story about the details about the way that we do group licensing and so on, which is to me hardly a, you know, it's, it's, it's an, it's an important issue because it's a question, you know, can people buy our stuff in a configuration that is convenient for them? And that's an important thing to get right. But I wouldn't consider it a sort of global issue in that regard. I think my attitude about products, kind of, maybe a crazy one, is I build stuff I would find useful and I do find useful. And that's sort of an overarching. At least there's a market size of one and it, you know, when people say I'm going to build this product for this market that I absolutely am not part of, I'm always a bit skeptical. It's always like, you know, there's a certain extra thing one gets from being in that market oneself and being, you know, I feel like one understands a whole bunch more about it. And I think, you know, by the time you're out surveying people and so on, it's, you've, I think you've typically lost the game. I think that it is, you have to have some sort of mental model of what this market Is you have to imagine, you know, start imagining people in that market and would they understand this? Would they understand this? And so on. Those are important things. I, I, I'm not sure, you know, for me, I'm building products that I know I want. That is a useful thing to a useful kind of guiding light, I suppose the, but you know, and, and then you have to be sensible about, well, how does it, how do other people get to the place where they can find this useful as well? But I, I think, yeah, so maybe that's a sign that. That's probably all I have to say about that. Christian is asking again about failure. Is there a place for a consulting firm which helps with dealing with failure as an instrument of innovation? I don't know. I think, you know, consulting is always a weird thing because it's, you know, the dynamics of it can get so strange when it comes to, for example, strategy consulting. You know, if you're like a high priced strategy consultant, you're coming in and you're sort of, you know, in the end you're going to be selling strategy consulting. And at least depending on how you do it, I mean, okay, for example, in my case, I kind of agree to be an advisor to various companies. Usually when I think that what the company is doing is interesting, I like the CEO. I think they'll take my advice, those kinds of things. So I sign up for a bunch of these kinds of things that has the feature that I'm not in there saying I'm going to bill you a bunch of hours. It's not how it works at all. It's not like I'm making my living from providing strategy consulting. It's like if I have something useful to say, I'll say it. If the CEO wants to ask something, I'll try and respond to it. But it's not kind of, I'm not in a dynamic where I'm sort of forcing myself on the company because that's the way I make a living, so to speak. And I think that dynamic of kind of the way you make a living is by giving more advice type thing can have some pathological features. I think when it comes to sort of one, one of the things that's very commonly happens with consultants is the company is going along, it's beavering along, it's kind of, it's down in the trenches, nobody's looking out of the trench and anything else that's going on, consultant comes in, parachutes in and the consultant can see a little bit further out and maybe has seen a bunch of other companies and says, hey, why didn't you think about this? Why didn't you think about that? And that's a useful function, I think, when it comes to. It's an interesting question whether if you look at companies, okay, how does innovation happen in companies? If you have a company that's beavering away, doing something successfully, it's often pretty hard to make innovation happen because what happens, people say, I've got a day job, I've got the job I'm doing. And by the way, we don't need to do anything else. You know, it's not worth even thinking about that unless we're going to make a billion dollars a year from that. So if you've got this thing that's just a little thing, that's just a little seedling that's just starting to grow, it's like, we don't care about that. That might make $2 million this year. We just don't care. We only care about the ones that are going to make a billion dollars. Well, the seedling is not going to get to that point for quite a few years, and it certainly won't get to that point if it's never allowed to grow. But it's hard for many companies. It's hard to say. There is these little things, the little pieces of innovation that happen aside from the mainstream. So what happens in practice? In practice, a lot of innovation gets done in small companies that then get eaten up by larger companies because the innovation is happening when there's five people who are madly trying to make something work. Maybe they don't manage to make it quite work, or they don't have the right distribution channels for it or something. Then they get gobbled up by a big company. And if the big company is lucky, that sort of seed of innovation will actually take root and grow in the larger company. Sadly, a large fraction of the time that little seed of innovation was, gets kind of sort of strangled when it gets into the larger company just dies. And the larger company, you know, spent the money to buy the smaller company. And it really kind of, maybe it got a couple of useful hires out of that. Maybe the very act of buying the smaller company kind of kicked the bigger company into paying attention to this or that thing and that was a positive in its own right. Or maybe the buying of that smaller company was cool and the stock market responded to it and that was good for the larger company and so on. But that's a very common model for innovation in lots of industries. I Think the other thing one could try to do, which I've certainly done, is to have sort of a special projects group that does the kind of the funky projects and is not part of the mainstream kind of flow of the company. And usually that's CEO pet project kind of kind of domain. And that's the way I've done it. And you know the funny thing about special projects groups as in all these years I've never really had a long term special projects group because my special projects groups do things and the things they do are successful enough that they become unspecial general projects and that the structure that got built to do that kind of out of the mainstream, it becomes the mainstream. And so then you kind of have to build another sort of out of the mainstream special projects thing to do the next special project kind of thing. But that's a mechanism that I found. I suppose it's like doing internal startups. That's a mechanism that is usually has to be pretty CEO related. It's I think hard to get that to work if it's sort of, sort of somewhere lower down in the organization because it tends to get squashed by other forces. It only really works if it's a sort of CEO pet project kind of thing. But I think that the dynamics of, I mean an interesting question. If you came into a company and it had a project that was failing and, or that had failed, could you as an outside consultant do something to help? I mean there's certainly people who for companies as a whole are sort of turnaround specialists who come in and say this company is dying or died and what can we do to put life back into it? I'm not sure that the life ever really lasts that well. I think it, you know, you can get the thing to have a little kind of, you know, to be, to kind of have a burst of energy again and possibly financial success. But I'm not sure that the, that the full organism ever really recovers from that kind of thing. And I don't know what, in terms of projects I'm, you know, this kind of redirection of a project from inside the organization. I've seen that, seen that absolutely work many times. The idea of you've got a project and it's, it failed, for example, what do you do with it? I, I'm, I'm not sure. I think that. I mean the one thing I will say about something that we do as a company is we try to keep pretty good records. And when a project is being hibernated and that sometimes happens. We try and sort of hibernate it well, so that in five years, when we realize we need this project again, it isn't a question of, hey, I can't find anything to do with this project. This happened just recently, just in the last few days. There was some project that was hibernated long ago that was being brought back to life after maybe a decade or so. Trying to remember what that project was, I don't remember. But anyway, you know, it turned out we had reasonably good records, which will allow us to kind of start again from where we left off rather than having to start from scratch. Let's see. Eli asks, can you remember a time in your life when you realized learning was fun? I always thought it was fun, I think, gosh, I suppose I just like, liked understanding stuff about the world. I mean, I, I, I would say that the, the part when I was sort of in school, early years of that and so on, you know, I would learn whatever it is, Latin and Greek and things like this. And it's like, what the heck is the point of this? But somehow I guess I got, I liked the dynamic of kind of learning the thing well, even independent of the fact that I didn't quite know what the point of the thing was. So even the satisfaction of doing it well, learning it well, I think was a helpful driver for me, even if the content wouldn't necessarily have been the thing I would have picked at that time. Although I'm very glad I learned those things, you know, 50 years later or something, because they've actually been useful and helpful. Walrus asks, do you look for practical or monetary value when choosing a project? Do you have advice for folks who aren't as financially independent, you know, for, for There are projects that I do that are independent of, of sort of financial considerations up to a certain scale. Beyond that scale for me at least, you know, I made a decent amount of money in my life. I have a reasonably large company and things like this. So, you know, there's a, but there's always a scale beyond which I can't do that without this thing making some kind of financial sense. And that financial sense might be, oh, you know, it's going to make us money, or, oh, it's going to be a thing that we're going to get donations for or whatever else it is. I would say that things that I do myself just on my own, I don't really have much of a calculus of those things. You know, when I'm sitting writing something or something, I Don't really have much of a is this going to make money or not? I suppose that's, you know, it's really, is this something I want to do kind of thing. When it comes to things where I'm paying other people to do stuff, then that becomes much more of a, this has to be a viable, you know, this has to have some sort of viable business objective. And, you know, sometimes, I don't know. Are there cases where there are things where I say I'd really like to do this, but I can't see a way to make it financially sensible? I suppose there are some. But beyond some scale, I always try to make sure there's some kind of business model. Sometimes it's an aspirational business model. Sometimes it takes a while to turn into reality, so to speak. But I don't tend to. I think it's always a helpful thing to say, how does this fit into the world? I mean, you can do something and it's a bit unsatisfying. If you do a thing and you say, I think this thing is really cool, but the world just absolutely doesn't care. It's something where I think it is, at least for me. I'm not doing the things usually for the sake of kind of, you know, for the sake of kind of the world's reaction type thing, but I am. It is something where it feels more meaningful to me if I do a thing and a bunch of people read it or a bunch of people use it or whatever else. And as soon as you have that sort of driver, then those are also typically things that you can figure out how to monetize. And sometimes it's challenging. I mean, one of the biggest challenges is can you monetize basic science? I haven't figured out how to do that other than essentially through philanthropy. This question of kind of, if you have a piece of basic science, you're figuring out some cool things, but it's going to have applications in 50 years, maybe, maybe it's 100, maybe you're off by another factor of five and then. But you're doing it now, today. And, oh, by the way, you want to hire five people to help you with it. Ten people, whatever it is, that then it's kind of a. How do you figure out how to monetize that? I don't know the answer to that. I think there probably is a way. I have a few ideas about it, but for me, it's sort of an unsolved problem. And so those kinds of things, I suppose it it limits the scale at which I am prepared to do those things. But I don't think that's a terrible. I think that if you let these things grow too much without any kind of, kind of feedback from the world, I kind of feel like it's those things, they don't grow as well as if you have something where you're actually kind of engaging with the world and doing things which have an obvious utility there. Let's see. I think I have to go on back to my day job in a moment, but just maybe one more question here. Faraday asks, what's your advice for building standard operating procedures? Oh, gosh. I mean, we try to make sure that. Okay, first point is any procedure that is very rarely used, everybody will forget. It's, you know, there is, within our company, for example, I know, though have almost never read, you know, vast amounts of material about our sort of operating procedures for all sorts of different kinds of things. But. My impression is that people's kind of searching for, wait a minute, what do I do in this case? Oh, whatever. That, that isn't terribly successful. That what's successful is there's an operating procedure, there's a piece of automation that does this particular thing and we keep running that same, you know, that, that same, that same sort of program. In effect, then people know what it is and then they follow it. And, and the trick is, you know. Okay, so I'll give you a story from long ago in the history of our company when business cards were more of a thing and when it became clear that There was a 14 step procedure that for getting business cards for new employees. And it's like, you know, as soon as you know that, you know it's crazy and you say, well, why is this? Well, let's see, what was there. And there are many crazy steps. And then you say, well, let's just simplify it. There's, you know, as part of some onboarding thing, there's one checkbox that causes this to happen and so on, and, or when somebody's title changes, which is known to the HR group, then there's this mechanism that lets you do this. I mean, in a sense, it's a design for how. It's like designing a piece of software. It's like designing a program. How does the program of the company work, so to speak? And you know, insofar as you can make things which are sort of. My impression is insofar as you can make things where you're sort of repeatedly doing these things and where sort of Maybe the exception cases are handled more by automation than they are by a human having to know that it's on page 47 of the Operating procedures manual, then you're in better shape. I would say that for. Yeah, so, you know, this automation thing I think is a, is big thing. Plus the kind of the, the concept, I mean, okay, so, so the kind of the, the concept of using sort of software methodology to design these things I think is really worthwhile. Now sometimes that's driven by, by, by kind of practical considerations of, of, of, of trying to, of, of. Of actually setting things up. So for example, we have a, you know, you have an ERP system, what is it? Enterprise resource planning, transaction processing and so on system. And you know, we've just been building such a system for the last many years using our own technology stack. A very lovely thing. And in the process of building it, we have realized how we can automate many things, how we can represent many kinds of policies, procedures and so on as essentially pieces of, as programs, as pieces of computational language code. And that is to me a more, a more fertile approach than saying I'm going to write it down in some manual. And then there's always the question of when people kind of join the company, how do they get to know how to actually operate within the procedures that exist? And this is a sign of, I don't know whether it's successful or not, delegation. I'm not actually sure how that works in many parts of our company. I know different groups have different kind of onboarding procedures and so on and different kinds of training programs for people to get them up to speed with the systems that they're using. But yeah, I think, yeah, so those are some very incoherent thoughts about standard operating procedures. All right. With that, I should probably. Part of my standard operating procedure is that I have some meeting which I don't even know what it is actually right now that. Ah, yes, okay. It's about something completely different. So I should, should, should wrap up here for now and thanks for joining me and see you again another time. Bye for now.