Ash, thank you so much for coming on. I've known about Zetta Venture Partners for a long time. How incredibly lucky they are to have you on the team, and part of that at that firm. But I want to go back, you were super involved with AngelList, which is now a juggernaut. Maybe we start there with your background, your history, and I'd love to hear some AngelList stories from you.
There are plenty of AngelList stories. So I got involved with AngelList pretty early on. When I got there in the middle of 2012, there were about five or so engineers, couple of designers, the founders, Navin, Naval, and myself and another guy, Kevin, we were working more on figuring out how to turn AngelList from a social network into a place where you could actually do business, like complete a transaction. Not just meet someone that might have a deal going, that might be starting a company, that you might want to fund or vice versa, meet an investor.
And so it was a really remarkable time because what was quite obvious was a lot of people were meeting each other on AngelList, but then they'd still have to go offline, meet in person, do all sorts of other stuff. And fundamentally, which is not a bad thing, that's often a good part of the process, they'd have to go and deal with a lot of lawyers and accountants to put their investments together. And what was really obvious to me at the time was that we could help with that. We could get the cost down, we could make it quicker, better, faster, cheaper.
And so we worked on that for many years, and really in less than a year, we got the cost of doing a deal that involved a lot of angels down from 40 or 50 grand to less than 10 grand. And that was over the course of 10 years, the 10 years it would typically take to manage the investment until maturity, manage a startup investment until exit. So that involved lots of little things, lots of automation, lots of doing deals. Making sure we could get the volume to make it worth the time of the firm that was helping us do all of this back then, which was actually a firm from Utah. And it was a lot of fun, and it was really good because you could see you were having a lot of impact.
Startups were coming to us at the time, and they either could raise elsewhere but just didn't know how to organize the investment from a whole bunch of different angels. Or they couldn't raise elsewhere because they didn't have the right connections, they were coming from a different market, didn't have a good network and AngelList helped them meet people. And these were pretty big startups. Uber and other people met investors on AngelList. Really, really great infrastructure companies like CircleCI met people on AngelList.
There's a long, long list of companies that meet people on AngelList. And you could see that these companies were coming to you with far less options than they left with after they were using AngelList.
Yeah, AngelList, the story is just incredible. And could you ever imagine it'd be as big as it is now?
To be honest, yes.
And I don't want to pretend I have a crystal ball, but I've had this discussion with a lot of people that were early at AngelList. And we ask each other and other people ask us, "Could you have imagined?" And honestly, it was really obvious to us at the time. Because we were seeing every single little transaction. And we were seeing things get done that people couldn't do, sorry, get done, before they came to us.
And there was just this incredible feeling in the team of seeing problems and solving them really quickly. We were a really effective small team and we knew it, and that feeds on itself. When you are presented with a problem, you sit in a room with a bunch of smart people, you solve the problem and you get the solution out really quickly, you feel really good about yourself. And you feel really good about the people you're working with, and you feel really confident that the next problem that comes in you're going to be able to solve it.
And so we just had such a high rate of being able to successfully solve problems, such a high success rate and rate of doing it that there was a very palpable feeling of success there early on. And that didn't show in the numbers for a long time, and that's just the nature of startups. But it was still obvious at the time, I'm not the only one who thought that. I think all the early employees really shared that feeling.
What did you learn from Naval who's become this startup philosopher? He's taken on this incredible journey since all this too. What have you learned from him?
Countless things. Naval and I used to have breakfast, lunch and dinner together. Get a coffee together. Because we were more on the commercial side of things. And I was between his vision and leadership and ideas, and then the execution side on the engineering side. Engineering, and also business, and fundraising, and structuring, and legal and all that stuff. So I was in the middle of that and really in the flow there.
And it was the case that I just got to spend a heap of time. And as I said we were constantly solving problems related to AngelList. But of course, when you go on your coffee breaks you try to solve the world's problems and talk about all that stuff too.
So I don't even know where to start. Naval's incredibly well-read, and so he pointed me to some pretty crucial books. And all of us in the early days were pretty interesting quirky people, and all brought something different to the table. And Naval was certainly the leader of that pack in that he brought a lot of different quirky things to the table, that helped me understand a lot more about how the economy works, about how evolution works, about how important evolutionary biology is.
A core defining idea for a lot of what Naval believes, I think, and that's probably still true. Also the more prosaic stuff, but very instrumental stuff at the time, like how to really run a fund and how to really look for a great seed investment.
I came out of AngelList with a lot of frameworks for investing, a lot of things to look for and things to strike out, when you see a team working on a problem early on. A lot of red flags, and I wrote down a lot of what Naval would use a lot of his heuristics from many years of seed investments and seed investing, and I used them in my future investing roles as well.
Were you thinking about artificial intelligence and its future back then?
Yeah, I was. I had actually helped, well, I had not helped. I was a co-founder of a company before AngelList, and that's how I met Naval actually. We used AngelList in the very, very early days around 2010 to raise a bit of money. And that company helped really big travel airlines, loyalty, and other companies manage new forms of data. So manage a lot of that social and location data that was coming into play around the time of 2009, '10, '11. And once you work in any part of the technology industry that's got to do with big data so to speak, you realize pretty quickly that the limits of human abilities around organizing and analyzing huge amounts of data are pretty obvious and very present.
So I was thinking about it then, but I was also thinking about it because as a kid I was fascinated by how the brain works. I was really into things like neuroscience and neuroanatomy and how humans learn. I remember as a kid being completely fascinated for weeks about what happens when we open a door. How do we know where to put our hand? How do we know what to do with it? Is the door even exactly there?
And then learning this idea that we're just predicting where stuff is in space and we're getting it correct enough. We don't actually ever know where anything is physically. I was obsessed with that idea for weeks and weeks when I first learned it. And so I don't know, I was always into that. I'd worked for this data company, and then at AngelList we saw thousands of companies. I had direct contact with over 20,000 companies in the couple of years I was there.
And when you just see thousands and thousands of companies working on thousands and thousands of different technologies, you start seeing trends. Trends are pretty obvious at a very large scale like that. And what was trending was the ability to use some of the modern data stack back then to start doing more powerful forms of machine learning. And then another tiny bit of trivia, the office underneath us was Kaggle and one of my really good friends at Kaggle. And so I would go down for lunch every now and again, and they would come up and come to some of our parties. Being aware of what was being done on Kaggle back then, also made it pretty obvious that this field of machine learning was taking off.
That's incredible. It's funny to think you were thinking about the existential parts of life even as a kid and how all of this even works. Were you reading philosophy like Nietzsche and things like that, or what had you thinking that?
I wouldn't say I was reading Nietzsche when I was a kid, and if I did, I think I'd be a pretty depressed kid. I wouldn't have been the happy kid that I was. I was pretty interested in a lot of that stuff. I would peek into some of those books. But no, I was really obsessed with anatomy as a kid and biology for quite a while, probably from the ages of seven to 12. Part of that was probably because my brother was into it, and I was into it, and we fed each other's curiosity a bit. Part of that was just spending a lot of time in the natural world.
I always preferred to be outside rather than inside in some ways. But honestly, part of it was just the books that were available to me. As in I had access to a lot of encyclopedias and books like that. But I didn't grow up in a world where I had the internet in my pocket all the time. So I had to seek out knowledge. And when I sought out knowledge, it was in more structured forms and I was just always bored as a kid. So if I had a set of encyclopedias, I'd just read all of them at my grandmother's house.
And I'd be at her house, my mom worked and I was there for a couple of days, or whatever had to be done. And they were the only books I had to read, so I just read all of them. And so you just go through things in a very structural way. And I think that probably forms the way you think, but also what you get interested in. And it was a combination of a lot of things why I was really interested in biology.
The weird thing is I never studied medicine as in I did all the exams, the equivalent of the MCATs and all that sort of stuff. I passed them, I got all the interviews, I got in. And then I changed my mind a couple of minutes before midnight, because the thing about medicine is you are told what to do for a very long time. And as any entrepreneur knows, it's sometimes pretty hard to be told what to do. And I was a lot more of an entrepreneur than I was a scientist.
What did you do after AngelList?
I met Mark. So I met Mark Gorenberg, a guy who's been investing for a long time in Utah and other places and someone you know. I met him while I was at AngelList. And he had started a fund and he was really focused on data and analytics, and was calling it the analytics fund, and called it Zetta and had raised part of the fund. And we just had a big mind meld over this idea that that was going to be the most important technology of the next decade. You could call it high-powered analytics, you could call it machine learning, you could call it AI.
There's lots of different parts and all these things of subsets and supersets of each other in some ways. But I met him through a friend of ours and we just got talking, talking, talking. And I was ready to concentrate on my investing at the time. I was spread very thin at AngelList as an investor, was very concentrated as a product manager and a member of the team and a manager, but spread very, very thin as an investor.
And I wanted to concentrate my investing skills and put a lot of what I'd learned into practice, and take on a lot more responsibility. So I did and I joined Mark, and we worked on Zetta together for many years until the end of 2021 when I left Zetta, and started doing more of my own stuff.
Yeah. Sorry, at the beginning of this conversation, I was acting as though you were still there, but what had you leave Zetta?
Oh, so many different things. Overall life is long, there's so much to do in life. And I just had a lot of other ideas, a lot of other things to do. And I'd also written a book about AI called The AI-First Company. You can look up my name, you can look up that and find it anywhere in the world. And writing a book is a really amazing thing because when people read it, they get ideas, that's the benefit of sharing knowledge. You're giving someone a bit of knowledge and they can have it too.
You can both have that piece of knowledge. It's not giving them a physical object where either one or the other of you has it. And people contact you, and people contact you with different ideas about companies they want to start of course, but also research they want to do. And also policies they think that the government should enact and all sorts of different things.
And so I was getting all sorts of people contacting me with all these ideas. And at Zetta I had to be extremely focused on our job, which was to invest the money of the charities and the pension funds, and others that we invested for. And I had to say no to all of those things, and it just got a bit overwhelming at some point. So life is long, there's lots of things to do. And we had a really good team, et cetera, and I wanted to go out into the world again and do other things.
Speaking of your book, what I really want to talk to you about is the future of AI and where you think AI is right now. Obviously ChatGPT, in particular ChatGPT-4 has really blown people's minds and woken a lot of us up. Even the president's are having meetings about this now, how it's regulated and things like that. So it's really kind of woken up the world. Where do you think it's at, and where do you think it's going? And I know that's an enormous question.
Yeah, it is, and I'm glad you recognize that. So where it's at is a very good point when you consider the... In a sense we've been building language models since the very beginning of AI. A lot of the earliest forms of AI were programmers writing programs to write more code to save them time. And that's a language model in a sense, but it's a language model that spits out language understood by computers, not by humans.
Then we were building language models in early translation systems. IBM in the '60s was building statistical machine translation systems. They were models, they were things that understood language or generated language. And so we've been working on this stuff for a long time, but we've been able to build very large language models. We've been able to process the data, get into those models. We've been able to have access to distributed computing to run those models, all that sort of stuff.
So where is it at? It's in a great place. It's in a place where so much good research, so much good thinking, so much good work on core bits of infrastructure has culminated in these products that are incredible. And I think the breakthroughs have really been around some of the finer tuning around data collection and model training that was done by OpenAI and just excellent product development work.
ChatGPT itself is a really good, easy to use product. And I don't think it's controversial to say that a lot of the core technology behind ChatGPT was not developed or is proprietary to, or is a completely breakthrough thing that was done only by OpenAI. But a lot of the product work and the data management work was. And so we're at a really good place because they did all of that work, and it was a huge amount of work by some incredible people.
So it's in a really good place in a way. I think it's in a good place in another way because people are really quite a lot more concerned, and I would stay awake to what this technology is capable of. And I say this in a very positive way because really for years I've been battling to try and convince people that they can use this stuff to do things, and save time, and money, and effort, and energy and resources.
I've been trying to convince people of this for a long time and once they see ChatGPT work they're like, "Oh, I can use this for this or that, I'm planning this and automating that." So being such an easy to use product is more accessible by a lot more people with far less technical skills. And I think that's a good thing because then people that are from one domain, but have no experience with AI products can see how AI products can help them in their domain. And so I think we're in a good place for that reason, because more people are more aware of it.
I think we're maybe in a slightly negative place because people are freaked out by it, and I get that. But in a sense I think we're freaked out by it, because we can see it understands us a little bit. And I was just thinking about this today. It's like when you see an animal just walking through a piece of land, you're not really freaked out by that animal. You're like, "Oh, that's just doing its own thing." It's looking for another animal. It's looking for its partner. It's looking for a piece of fruit, whatever it's looking for. When you start getting the impression that that animal is tracking you, as in trying to predict your next movement. And when you start getting the impression that the animal knows where you're going, you start getting freaked out.
And the thing is we've had AIs that have been doing things in the background all the time, but we haven't cared about it because we haven't had the impression that it has any understanding of us, or has anything to do with us. But now we're getting the impression that it understands us, and actually nothing's changed. A lot's changed, but also nothing's changed. It doesn't actually understand us. We just get this weird feeling that it understands us because the predictions are more in line with our own predictions.
As in the predictions about what a chunk of text means, are more in line with our own predictions or conclusions about what a chunk of text means. And so I think we're in a bad place because we're getting freaked out about it. But it's a very, I don't know, it's not a very rational feeling. I don't think. We're in a good place in many ways, but not in so many other ways. So that's my commentary on AI as a product and the technology, there's a lot more about where we're at as an industry in terms of economically, and from a regulatory sense and whatever else, but I guess there are other discussions.
What is the fear? You have Elon and other folks who are like, "Hey, this is more dangerous than nuclear weapons if not treated the right way." And I'm sure he's right, I don't know, but it's never fully articulated how. Why is that true? What is the fear?
It's not. So I'm not going to guess as to what other people's fears are because I'll be wrong. I'm not them. I can tell you what my fears are. My fears are that AI is used as a tool to get intellectual and emotional leverage on propaganda, and then in weapon systems. I'm fearful of a lot of different things around AI, but they're the two main ones.
And what I mean by that with respect to the first is quite linked to a lot of these language models, as in it's quite obvious that AI can be used to generate text that makes a lot of sense to us. Can be used to generate an agent that can talk to a human in a way where it thinks it's another human, so it feels some connection to it.
It's emotionally manipulated by it. It can be used to target people. It can be used to understand people in a way where you can get a message that's going to get their attention really quickly. And where I'm going with all of this is propaganda has been used by governments in all sorts of ways for many, many years to achieve their goals.
And governments have goals around propaganda and sometimes they're good, they get people to do things that we need to have collective action around, otherwise there'll be big negative events that eventuate.
But it's also used to do things that are less positive. And using AI to generate, and target, and distribute propaganda is something that is evidently very accessible to governments and is potentially extremely powerful. I hesitate to even use the word potentially because I know it's being used in a very powerful way today, so I'm very, very scared of that. And then with weapon systems, again, it's another discussion, but it's just the physical manifestation of targeting.
And so once we can do super simple examples, very accurate, very large scale face recognition, you can target weapons on a one-to-one basis very, very easily.
Oh, interesting. I hadn't even thought about that, that's super interesting.
There's another one.
On the propaganda front, how worried are you, the United States has a big presidential election coming up, but you know what I mean, all countries have elections. How is this going to affect it? What do you see happening for—
It's probably going to run the whole thing. I'm hyperbolic. How's it going to affect it? I think the question is how's it not going to affect it? There are so many news channels now. There are so many little sub-communities, there's so many ways to get to them. A lot of these channels are open.
And look, I do not doubt for a second the power of the various agencies in the US to control a lot, incredibly competent, incredibly good, and whatever else. And I won't speak too much about that. But at the end of the day, people get their information from lots of sources and it's hard. It's really hard to control every source of information that someone gets. So they're going to be a lot of people getting a lot of information about a lot of things.
What do you think the future of media is now that AI is in the mix? We're seeing BuzzFeed, everybody's just going layoffs and losing market cap, and all these types of things. What happens here? This has been a discussion for a long time, even pre-AI being at the forefront of all of our minds with the internet, Facebook, and social media taking a lot of their revenue away. But what do they do now? I don't even know.
This is such an interesting question. I'll start the answer by saying what do I know? I'm not in the media industry, and I'm frankly very detached from the media industry. I don't consume much media at all, especially modern stuff. I consume old stuff, offline and a lot of that. So I will start by saying I'm pretty out of touch in some ways. In other ways I'm pretty in touch, because I've worked with a lot of the tools that are used in media. So a bunch of random points. I think overall what I'm very curious to see, and I'd love to take a positive view on this, but I can see myself taking a very negative view on this, is with a lot of computer—by that I mean AI generated content.
Do humans eventually just get really sick of really boring AI generated stuff? Or do they get sucked in by that and go down to that level. As in do we go up to the level of only consuming highly creative, very human-centric, very emotionally engaging content that only another human could generate, or human with the help of some tools, AI based tools. Or do we go down to the level of only consuming content that is mass-produced, mass-generated by AI.
And I'm not saying it's good or bad, I'm just saying it's not necessarily creative in the way that humans can be creative, because it's not that creative in many ways. And I mean this in terms of visual, sonic or what a trio should say, and verbal content. I'm not really sure where we're going to end up there. I can be skeptical or I can be generous in my predictions, but it doesn't really matter. The point is that we could go either way.
I think more pragmatically, what is very exciting is that a lot of the very tedious stuff that you have to do, for example, editing sound files, editing movies, even editing articles, there's some extremely tedious stuff. And I've been a sound engineer at various points in my life in an amateur semi-professional sense. And then I've also edited movies, and I've also edited photographs, and spin into these things in life. And there's some really boring stuff you have to do, I'll bet.
And I'm really excited for AI to do a lot of that. It's a company that I invested in a very long time ago as an angel, Canva. Right now you can just take a photo, they've got this magic feature where you take a photo of yourself. I take a photo of myself like this and just say I want it to be me wearing a suit, because I'm using it for LinkedIn or whatever. And you just click a button and it's done. You can do that in Photoshop if you've got a spare day.
It takes forever in Photoshop.
But who wants to do that? It's boring.
It takes way too long.
Some people find it therapeutic, but I find that quite boring. So I'm really excited in a very pragmatic, practical way for how we can get rid of all of that and just do more fun, exciting stuff.
What do you look for when you look to invest in an AI first company, and how do you know if that's real? Or now literally every tech company is an AI first company, or that's how they're pitching it. How do you know when it's real and when it's not, and what do you look for?
Broadly this is my whole job, so I've spent 15 to 20 years answering that question and building systems, checklists, and processes and whatever else to come to that conclusion. Very broadly, is it special and will it capture value? Is it special? That's a lot of technical analysis, which is, have they actually invented something? Is it actually different? Is it something that's very defensible? Can you patent it, or is it just impossible to copy? Do they have a data set that no one else has that's been used to train it? Do they have a piece of infrastructure that no one or very few people can access to, it's super-fast, whatever it is.
There's a whole bunch I write about in my book about how to determine whether something is special in some ways, particularly on the data side. On the core models and infrastructure side, I don't talk about that in the book because it's too technical for the audience, for that particular audience I should say. So is it special and then does it capture value? This is something I do talk about a lot in the book because it involves using frameworks of competitive advantage. And the whole reason I wrote the book is to go back a sec.
There are lots of ways to figure out if someone's got a competitive advantage. Michael Porter wrote a bunch of books about this, a whole bunch of people go to business school and study all this stuff. And I just didn't think any of those frameworks were very relevant in the world of AI. So I wrote a book that tried to develop a new framework for figuring out if someone has a competitive advantage in the world we are in today, in the world of AI.
The second area with the big set of questions that I ask about is can it capture value in a way that nothing else can capture value, or no one else with the same technology could capture value. And going back to your previous question. I really haven't seen many if any people articulate where a lot of the value's going to be captured in the fields that are getting a lot of traction right now. So fields like generative AI, large language models and whatnot. We know who's building some really good products like OpenAI and others.
We know who's providing a lot of the infrastructure, NVIDIA and Google and others. And we know who is integrating a lot of the technology into their products, Microsoft and whatnot. But we also know there are thousands of other startups purporting to have a way to use a lot of this technology to capture value and build a company around it, and a sustainable company.
I'm sure that a lot of people are being very clear about why they think that is the case, hopefully some people will read the book and figure out how to do that. Hopefully, I can help a lot of people figure out how to do that. And hopefully other people will come up with new ways to figure out how to do that, to capture that value from what is fundamentally a pretty open, not totally open, but a pretty openly available technology at this point.
Yeah, it is interesting the future of large language models and how much further that can be taken. But something like Midjourney is interesting, I'm sure you know. Creating the images and things like that. Don't they only have six or eight employees and it's game changing what they've done. And the whole company’s run through Discord, it's wild.
Do you see more companies being like that?
Yeah, for sure. I've seen companies be like this for a decade. All of these little bits of running a company, creating software, building infrastructure like that have been coming into the world for a long time now. So you've had different ways to do distributed work that have been better and better over time. You've had different ways just taking your example obviously. Different ways to run AI models on different computers that have been getting better and better for a while now.
And as I said at the beginning, it's been a combination of lots and lots of different things, and you see these amazing things happen. So I'm really excited to see more of it happen. You will see more of it happen, you'll see more of a few mainstream and hopefully gradually these new ways of doing things. Not just building companies and remote work and all that sort of stuff. I don't find that very interesting, in terms of that's not necessarily game-changing.
But building different models and integrating these AIs into products, that's going to be fascinating and hopefully they'll replace old ways of doing things that at this point it's just wasting a lot of people's time and money.
Well, beyond should you be working in person, should you be working at home, hybrid, all that type of stuff. I agree with you, I feel like that conversation's been had, but what is the future of work beyond that?
I hope that the future of work is more self-defined. I think work for the longest time has been very much defined by a very small number of people in society. In feudal times it was defined by the people that owned the land. They decided what work everyone else did. They had to plant these seeds, sow and reap and blah, blah, blah, do this labor to get value from the land. And then the work that was done by people in society was decided by those that had access to factors of production like oil extraction, or manufacturing certain things, or loom, or whatever it was. And again, there were very few people in society that had access to various things, or were just super great entrepreneurs and inventors.
And then we can go into the intellectual era, sort of highlight these eras in the book. But the work that most people in the world did was determined by the very few people that had access to the core factor of production at that time. Now the core factors of production are just more distributed. I'm not saying they're totally evenly distributed, they're far from it. And I'm not saying they're well distributed even in the hands of the right people. And I'm not saying lots of things. I'm just saying they're more openly, evenly well distributed than they were in the past. And that means more people can choose, I hope, what I see now, but I hope this is more the case in the future, what work they do.
And interestingly, this is my personal mission, just as a side note, and this is why I worked at AngelList, and I only bring this up because you wanted to talk about AngelList at the beginning. My dream in working at AngelList was that I have a very small impact on helping more people work for themselves. So if they had access to capital, if they had access to talent, and they had access to distribution of their idea and whatever else, these are all things you can do on AngelList. Or AngelList, Wellfound and Product Hunt and all the constituent companies now, then they would be able to work for themselves as in choose the work that they want to do.
And I think that is a very humanist perspective but it's all philosophy, but it's mine. And I work in AI for the same reason, which is I think it can help more people choose the work that they want to do. And whether that is just because they have to do less work because AI's doing more of the work for them, or whether that is because they can get AI to do lots of the work that they don't want to do, and they can focus on doing a lot of a certain type of work, it's besides the point. But I think the future of work is more people choosing what they can do.
Now of course I can go on and on about all the different technologies that AI enables that help you get work done faster. But I would say I'm most excited for the prosumer category, which is the individual small business owner that may have zero or a small amount of employees. And then being able to use AI to do things like manage their books, produce content for social media, make a flier, make whatever. Manage their marketing campaign, create a budget for them, and do all these things that are just generic business things that have nothing to do with the actual product they're building.
They might be machining a lens or building a microscope—I was just thinking about microscopes and telescopes over the weekend, so that's my example. But they might be building something that is actually very unique, artisanal, technical, whatever it is. All this other generic business stuff, which is just a thing you have to do to get your product out there in the economy. It's just not central to what they want to do and it's not central to them, the value they're adding to the world, the unique value they're bringing into the world. I hope AI can help people do a lot of that stuff.
What are your thoughts on universal basic income? Something that Sam Altman and others have been talking about for a while.
I was really interested in it a while ago, but frankly I'm out of the debate now, so I don't really have a good answer for you there. I am not a super political person. And I've opted out of living in a society where that's even really a question. I live in Italy, and I live in a society where you have less of a need for policies like that because there's a little bit more of a socialized system, or a system which takes care of people in a different way. So it's not a debate I'm part of anymore and not well-informed about.
I'm interested in how you live your life. We've been doing this about 40 minutes now and that seems like a recurring thing. It seems like you have a set of values and principles and ways that you go about living your own life, which is beautiful. The fact that you didn't even entertain that question is fascinating to me. Tell me how you live your life? What's a day look like? I'm super fascinated.
It's very nice of you to describe it as beautiful. I think some people would. I think a lot of people would describe it as rigid, and not very beautiful. So I think you keyed into a few things there. It's amazing how perceptive you are, and that you've noticed these things just in these 40 minutes or so.
So a lot of what you said is roughly right, which is I do sit down and write down every couple of years the principles I want to live by. I want to be honest and have a certain degree of integrity. I want to live according to this principle or that principle, and I score myself against them. Did all the things that I do this year align with that or not? And so it is very principled and I have a set. And it doesn't change much over the years I would say, this is the core set of principles, which I guess you'd say are quite moral in category, they're quite categorically moral.
They don't change much, but I do have that, that's the multi-year view. The intra-year view is that I do set goals that are very clear. I set them with a small group of people, there's four of us. We check in with each other every week, every quarter, and every year. We check in with each other every week in a Google Doc. Every quarter with a meeting and a quarterly goals doc, and every year with an offsite, it's called the elephants. And there's a couple of articles on the internet about how these groups work. But we started a group together a while ago, and so intra-year I live my life very much centered on those goals and whatnot. And then daily, it's actually about what I don't do as what I do do, so there's a lot that I've cut out of life.
Haven't looked at serialized television or anything on a screen, except a feature length movie at a movie theater since 2006. Haven't read news content since 2006 until the start of the war in Ukraine, when I started reading the news a lot again. But I didn't have a TIME subscription, never read the newspaper and anything daily. I'd only read things that were produced on a quarterly or annual basis, so there's just a lot I don't do. I don't watch sports. I've never watched sports. So I don't do a lot of things, but I do do a lot of things.
There's three things that happen every single day, which is I do a lot of sport. I do call a lot of my friends and family, and I do a lot of work in a very specific area. And even things with work, I don't invest in anything consumer, I don't invest in anything except AI really. Now AI covers a lot, but again, it's a lot about what you don't do. And that is strategy. Strategy is about picking what you don't do.
Michael Porter said that, a lot of people said that. And I think people think about strategy a lot at work, but remarkably, people don't think about strategy a lot in their own personal life. And that's fine, you can roll through things, that's totally fine. But I've just got a lot of joy out of setting a strategy and really having an impact in very, very small ways, so I'm continuing to do that.
What is your thinking behind not reading news? I'm fascinated by that. I think I understand the TV one, definitely understand the sports one. You're either all in on that at our age or you're not. But what about the news that's fascinating to me, that would be hard for me to give up.
A lot of reasons. So it started with an economic analysis, very basic, and I don't want to make it sound like I did some extensive research project on this. It became clear to me in the early 2000s and 2010, that the way journalists were getting incentivized was changing. So they were getting paid, not every news outlet of course, but a lot of the news outlets by the amount of clicks they got basically. And so they'd write clickbait is what the word we use for it now. But generically, they'd write articles that get people's attention for just low enough to click on it. And then the rest of the article didn't even have to be that good, because the journalists were paid at that point and they didn't care. So just the incentives changed in the industry, and that changed what content was getting produced and I thought it was worse, so that's one thing.
Secondly, I just started thinking about, "Okay, I can do a million things in life, but where will I actually have any impact?" And the reality is you can only really have an impact very, very locally that is on your friends, the family, your family, the people you work with and that's about it. Trying to impact something on the other side of the planet, that has nothing to do with any of your expertise, anything you know about, it's just not going to happen. You're not going to really have any positive impact. At worst you're going to just pontificate, and at best you're going to be a pest.
If you try to get involved in some problem on the other side of the planet that you know nothing about solving, or nothing relevant to solving it. Or you don't have the resources to solve it or whatever. So I just focused ultra locally, and that means not reading about things that don't affect you. The third thing is there's a bit of a mental health aspect to this. There are varying degrees of responsibility around various news outlets and journalists and what they write about. Censorship has changed a lot over the years.
Information flows used to be controlled in very different ways that were arguably better for the aggregate mental health of society and even the median mental health. There's a lot of horrible stuff out there that is sometimes important to know, but communicated in a way where it can be very distressing. And in other times it's just not important for anyone to know, no one needs to know about it because it propagates stuff that really shouldn't be propagated. So I won't go into detail, it's sort of obvious. There's a bit of a mental health aspect to this.
And then finally, it's selfish, but you've just got to be a bit brutal with your time sometimes. And look, it might be good to know a little bit more about what's going on, but if you're going to have an impact, the reality is you just have to focus on something for longer than anyone else focuses on that. And you can't do that if you're reading every release by every agency and every company and whatever else.
And then the last thing is it's the volume thing, it goes back to the data problem. It used to be the case that there were only so many news outlets, and you could consume pretty much all the news over breakfast, all the news that there was. All the news that you needed to consume to be as informed as anyone else. You could pretty much consume that over breakfast. Now, to be as informed as a lot of other people, you basically read the news all day long, that's it. There's a lot of people that do that. I think the expectations are very unrealistic at this point. And so I gave up on that, trying to be as informed as a lot of other people because it's really hard with the volume that we have.
For sure. I want to be super respectful of your time. I have three more questions. I'm fascinated. One is pretty simple. What do you read? Do you read fiction? What are you reading day to day?
So I try to chunk my reading in areas, and I'll read lots of different things in those areas. So I'm interested in intelligent systems, AI, investing, the craft of investing and understanding how to manage risk and make your decisions. And then Italian culture, it's alliterative, three Is, so I only read things in those areas. Now they can be pretty broad. For example, I consider a neuro philosophy book or a neuroscience book to fit under intelligence systems. And I'd also consider some computer science stuff to fit under that. Italian culture can encompass a lot of different stuff, language history and all sorts of stuff. And investing can encompass economics, economic history. So look, in a sense, it seems narrow, but it's actually fairly broad.
What do I read? I try to pretty much just read books and research papers. That's it. Full length books and research papers. There are huge downsides to that. There are a lot of books that are way longer than they need to be, and that aren't very well written. And keeping up with research is very hard, you have to be a very discerning consumer. So look, it's not the best way to consume. They're not the best two media types to pick, but they're just what I pick.Sure, I listen to the odd podcast, but really very, very few.
Maybe one a week or something like that. Areas only really two media types, and then I try to chunk it. So I try to within those areas, read two or three things, maybe a book, a research paper, and then make friends with someone who's an expert in an area at the same time. So I'll read all about a particular type of the brain, or a particular neurological disorder all at once in a one-month period. Or I'll read all about a particular period of economic history, or a particular market phenomenon all at once. I'll pick two good books and read one good essay about it, and then call someone I know who's an expert in that area about it all in the same month. So I try to chunk little subtopics together.
Within your answer to that question, you touched on my second question, which is why Italy?
Lots of reasons. The first is I really appreciate the sense of charity and the care for the weak here, whether that's people with physical or mental disabilities, whether it's old, even children. The attitude to treating people in those categories here is very, very different to other parts of the world in some ways, in terms of government support and social support. And I like living in a society where the weaker are taken care of, so that's one big reason.
The second is, of all the countries in the world I could have moved to that weren't my home country, Australia or the country that I made my home for a very long time, America, Italy is the one where I'm just more accepted because my family's originally from here a few generations ago. So I'm able to get a passport and people get that I fit here.
That wouldn't be true if I moved to Poland or somewhere else like that because I'm not from there, and so there's a certain cultural acceptance. It's a really wonderful place for the things I like to do. I do ski touring, and cycling, and trail running, and things like that. And the mountains here are phenomenal. And it's really world-class in a lot of areas that are pretty important to me.
So areas like neurosurgery, ophthalmology, and things related to how the brain works, they're world-class, various areas of mathematics and engineering, machine learning, world-class. Even some other up and coming areas like quantum computing, they're pretty good. So there's a lot of opportunity to talk to some of the best people in the world.
And then it's just central, you can get to anywhere in Europe super quickly. And if there's anything you really want to do, you can probably find it in Europe. So it's really a central place. And the other thing is there are a couple of really important things in life to be a functioning human being, eat well, have good healthcare and move well. And those things are very high quality here, but very low cost. And I think that's a good thing to optimize for.
Yeah, for sure. My final question is we believe chances are just as important when they're given as when they're taken. We talk a lot as entrepreneurs and investors in the business and entrepreneurial world, or the tech world about chances that we've taken. But then there's this whole other aspect of this, which is chances that have been given to us, or chances that we have the opportunity to give.
And I wonder from your perspective, who stands out when you think of chances that you were given and who that was and what that meant to your life and your career?
Because it is the case that we're constantly giving and taking chances, making, taking chances. And it's not the case that you're just a chance taker, someone has to be on the other end of that a lot of the time. I will though give the really, really obvious answer, which is my mom and dad in that they keep taking chances on me. They keep giving me advice, I ignore it, I make a mistake and they go, "That's okay. We'll just give you more advice." They keep watching me do stupid things and then go, "It's okay, you'll pick yourself up, and we'll give you another chance to get through your life without doing something that kills you." So the obvious answer is my mom and dad, they've been giving me chances since I was a kid. And they just keep doing it even though I'm now a developed adult. So it's a funny one, but it's an obvious one.
Then there are the various people in my career. I think the whole team at AngelList, not just Navin and Naval, but the whole team at AngelList took a chance on me. I wasn't contributing to the product from day one. I eventually did, but I was contributing in other ways. And I eventually helped transform the whole business in many ways.
Penguin Random House who took a chance on me with my book, they thought, "All right, this is an interesting topic. You have some expertise in it," but I wasn't a published author at that point. I'd published lots of articles, but I hadn't published a book and they took a chance for me, and I'm really, really grateful for that. My partner Mark at Zetta, he took a chance for me. I was a proven investor in some ways, but not in others.
Obviously, we made a very sensible, rational business decision to work together in lots of ways. But he took a real chance on me, and we were taking a chance on each other. We were very, very complimentary, but very different. We got something off the ground, so he did. There are lots of people that I have in my life, and I really like this question because it's making me think who can I take a chance on tomorrow? So I'm going to think about it, so thanks for asking.
I love that. Thank you Ash so much for taking the time. I highly recommend everybody read your book. We'll put a link to it. We'll put it on CEO.com. I think it's a must read in 2023. Ash, thanks so much for coming on.
Thank you very much, Clint.