The Exchanges

Q looked was in the checkout process. How can we, you know, deploy this there? And, you know, for me, it seems interesting because intuitive, like, like going by my gut, I would say maybe at the beginning when you're in discovery or at the end when you have customer service issues, but you're saying at the checkout. So what kind of questions come out at checkout and why start there?

A Um, we felt that the checkout was really the fulcrum point where customers have the most questions about a booking. Like they've already are knee deep into the details page of one of our products. So say it's a hotel and they're about to purchase something, but they have some, you know, just nagging questions that could be plaguing them. Like a great one is, is it pet friendly? Um, what are the points of interest close to this hotel? Uh, what's the neighborhood description around it? You know, answering those final closing details for the, for the consumer builds more confidence in their booking, and that's why we, we started with, with checkout from, for, from our perspective. Two, it's at the heart of what we do. Uh, so creating a frictionless checkout experience that answers all questions for our customers before their booking, we felt was a great place to start.

AI assessment note: “we felt that the checkout was really the fulcrum point where customers have the most questions”

Q Ryan, I'm kind of curious, like you read the message boards. I mean, don't message boards and online forums have a chance, have a tendency to radicalize and also like make the most intense ideas rise to the top. Is that, is this just kind of like a classic case of that?

A I've never heard of that ever happening before. What are you, what could you possibly be referring to? Yeah. I mean, no, it does. I mean, that's, that's definitely the case. Um, I mean, we didn't even really go further into this cause it's so dense, but the reason I was even writing about effective accelerationism this week is because I got sent a tip about a group that was causing trouble on tick tock for my readers. And when I started poking around, it turned out that they were A group I'd written about previously of crypto accelerationists that had rebranded into AI accelerationists and were part of the New York downtown scene, trying to get Peter Thiel's money. Like these are the same people that have been kicking around for five years, trying to bring about some kind of white nationalist apocalypse via automation. And they went from cryptocurrency to AI and it's the same people doing the same stuff, sharing the same memes that they've been sharing since 2017. Um, And I think when you talk about the radicalization of these message boards of these online communities that are talking about these philosophies, what it is is that you don't know who you're talking to, and you don't know what their goals are, where their funding's coming from, or what they're influenced by. So, you know, it's very common to be reading like a Twitter thread, and all of a sudden, like, if you don't…

AI assessment note: “Yeah. I mean, no, it does. I mean, that's, that's definitely the case.”

Q where people that are, you know, inherently tied to EA rise very high or sometimes run them. I mean, Sam Baikman Freed, who we've talked about on this show was tied to EA. We also have, you know, folks extremely high with an open AI and anthropic, um, tied to EA. How has it been so effective at churning out leaders in the tech world who believe in this stuff?

A Well, I think, I think it goes in both directions to some extent. I think some people were, to some extent, shaped by EA. You know, Sam Bankman-Fried encountered it very early in his life, and at least to his telling, made decisions about his life based on the philosophy, whereas there are other people who have sort of adopted it after much of their careers. So, you know, Elon Musk, for example, has somewhat embraced effective altruism, but it doesn't seem like it was really a defining feature More as it was something he learned about and said, oh yeah, that sounds about right. Um, and I think the same is true of Marc Andreessen and effective accelerationism for that matter. You know, he was not thinking about effective accelerationism, which was not even a thing back in, you know, when he was developing web browsers. This is something he is now adopting sort of post hoc. Um, and I, you know, I argued this in an essay that I wrote, but I think You know, effective altruism and to an extent effective acceleration, or sorry, the other way around, effective accelerationism and to an extent effective altruism are really just sort of a rebranding of a lot of the same philosophies that have existed in Silicon Valley for a long time where people really want to feel good about themselves. Like they're, you know, pursuing a higher cause, not just the pursuit of massive wealth. And these …

AI assessment note: “these philosophies are very convenient because they allow people to define themselves as these hero figures”

Q So you're saying that the reality is a little different than the story that he's telling?

A The reality is a lot different, and to be honest, Alex, it is something that even I didn't know until I really started researching for this book, and I did something that I recommend no one ever does. I went back, and I had a research assistant. She was amazing. I kind of plug her. Her name is Abigail Somerville. She is truly the best, and we together read every single news article that Ray Dalio had ever been quoted in from 1975 onwards, and What we found was nonstop since the 19 eighties, Ray Dalio has been predicting economic calamity. Every, virtually every single year he has predicted a giant crash. And so what happens in 2008 is he's not lying. He was, he did predict an economic crash in 2008. What he doesn't talk about very often is all the other years before that, that he predicted the same crash. That never happens.

AI assessment note: “The reality is a lot different, and to be honest, Alex”

Q Okay. And so then obviously this stuff works and he's like, well, I have the investing principles down. Now I'm going to start applying them to management and life. And this is how the principles start to emerge. Is that right?

A Exactly. So he, he's made his billions. He's made more money than we could ever spend, right? And he begins to say, what, what is out there besides wealth for me? And he realizes that these rules that he has for investing, they really come from his overall philosophical outlook, and his philosophical outlook can be boiled down to, you should just be honest with one another. You should just, we should be able to stop this interview right now, and I should say, Alex, Professionally, that question didn't work. Here's how I would have said it. You should be able to stop me in the middle and say, Rob, you're rambling. And, and there should be impersonal. And that if we could do that with one another, we would, we would come to an, a better outcome because we would remove the emotions from it. And we would just, we would move on with our day.

AI assessment note: “Exactly. So he, he's made his billions.”

Q Yeah, and then how did it manifest itself in the firm?

A Well, the problem is it's very different, um, in practice than how he has described it in, in person. So at the beginning in, say, around that 2008 financial crisis, the principles were really just sort of like one man's word. It was just a few sentences written on a piece of paper, and it was a few nice metaphors, and no one took it all that seriously. Um, sort of like how, you know, you walk into an office, you walk into an office, uh, An apartment building, and you know, it's written on the wall. Have a sunny day. You know, it doesn't mean you have to like put that into your heart and take that with you all day. The problem is he keeps writing more and more principles and they get less philosophical and more literally. And they say things like if someone does something wrong, they prescribe for what he calls public hangings. So now we should learn from each other's mistakes by watching us metaphorically Execute someone, which at Bridgewater would be firing them. He says that we should diagnose even the smallest problem. So if something goes wrong, if something as small as, you know, the peas in the cafeteria are wilted one day, we should haul in the head of the cafeteria in front of video cameras. We should really probe in his words. We should probe what happened. What does the, what thought process led to those peas and We should frankly humiliate people. You should, one of…

AI assessment note: “if someone does something wrong, they prescribe for what he calls public hangings.”

Q How did this not, like, incite very vicious internal politics? I mean, if you're walking out of a meeting and you have, do you see who's rating you what?

A Like, Alex, it cited the most vicious internal politics. You do see what people are rating you. You can rate them back. You can also take polls. So let's say I really believe, uh, Alex, this conversation is going terribly. I think everyone agrees with me. I can stop This in its tracks, and I can put up a poll on the dot collector, and I can say, do you think Alex is looking at this right? And people can vote. By the way, whichever one of us loses that vote, we're going to get a whole lot of negative dots. Really? I want to .1 other thing that people might be thinking. Ok, so you've got this dot collector, but how can you force people to use it? You know, I've got a job. You've got a job. Maybe I just go throughout my day, and maybe the peas in the cafeteria are wilted, but maybe that doesn't, I don't, that doesn't rise to the level of a dot for me. What Bridgewater starts doing is it starts telling employees, you have to rate a certain number of incidents per day. You have to put in dots. And not only that, but a certain percentage of your dots must be negative. So they are forcing you. Yeah, exactly. They are, if there weren't internal politics before, they, they're now guaranteeing there will be, because I can't just walk around and say, great job guys. I, I have, and, and I'll say Ray Dalio's defense here is, remember, his whole philosophy is trust in truth, is to offer fran…

AI assessment note: “it cited the most vicious internal politics. You do see what people are rating you.”

Q Very curious to see where it's going to go. Okay. What, what are you up to with Inventive?

A Yeah. So at Inventive, we are building, uh, we're kind of using the power of LLMs for enterprise knowledge management. So one of the use cases that as, as I dealt deeper into my experience with language model, uh, at Google and outside of that as well was that Like enterprise search kind of sucks. Like even inside of Google being a search company, like we had probably the best enterprise search, but it was still pretty bad. And, uh, so we were kind of exploring, we explored a few different ideas, but we landed up on this one because we just felt that time is right to disrupt enterprise knowledge management. So, uh, we're focusing specifically on sales, uh, knowledge management. So we are building an AI powered platform for sales knowledge management. And, uh, the first use case that we are solving for is enabling sales teams to fill up RFPs and security reviews with their internal existing knowledge basis.

AI assessment note: “we are building an AI powered platform for sales knowledge management”

Q of the key questions is whether Sam was aware of this. And I think Caroline Ellison's testimony of that she kept on preparing balance sheets and rearranging things and naming them in interesting ways to make sure that You know, Sam understood what was going on, but the general public might not if they came across the documents was very interesting. So can you talk a little bit about that?

A Yeah, so there was a lot of sort of creative accounting going on that was trying to, uh, really mask all of this debt that Alameda had. Um, and so Caroline Ellison testified to preparing balance sheets for the company, um, and then at Sam Bankman Free's request, preparing what he referred to as alternative balance sheets. Uh, and those were basically Copies of the balance sheet where she would move around or rename entries so that it looked like, you know, the, the loans to FTX executives were just a part of some other spending. Uh, she sort of removed some of the liabilities from some of the balance sheets. You know, it was all stuff, I think, I think in accounting, you know, there's some degree of flexibility where, you know, some of it's up to interpretation, just exactly how you present things, but this was like, Well beyond anything that would be accepted by any competent auditor. Um, and so they, she basically testified that she prepared those balance sheets at his request and then sent some of those alternative balance sheets to, uh, lenders and investors so that they thought that Alameda was in a much better financial position than they were. And, you know, there was testimony later on in the week from one of those lenders who basically said that Had he had access to one of the more truthful balance sheets, there's no way that his firm would have lent to, to Alameda.

AI assessment note: “Caroline Ellison testified to preparing balance sheets for the company, um, and then at Sam Bankman Free's request”

Q skeptical after this. I, I hope so. So, where, where do you think crypto goes after this? I mean, obviously, we've seen a great evaporation of wealth, both in the falling of, ah, so many coins, like going from basically three trillion to a one trillion dollar industry, ah, within a couple months, and then these billions of dollars that were lost in FTX. What is the future for crypto?

A Well, I think it's going to be challenging, uh, reputationally to come back from this, but I do think that they will certainly try. Um, you know, the, I think people in sort of the general public see FTX and Sam Bankman Freed as almost synonymous with crypto where, you know, because he was so prominent, the advertising was so prominent because he was talking to Congress, you know, he was very visible. Um, for a lot of people, that was all they really knew about crypto. And so seeing this failure to them is like seeing crypto fail. And I think that'll be really challenging for the crypto industry to move past. Um, they've certainly been trying to sort of say like, oh, well, Sam Bankman Freed wasn't crypto. He was just one fraud. You know, real crypto is, is all better than this. Um, but I think that's a challenging distinction for a lot of people to make in sort of the more lay public. Um, Um, but I also know that, you know, if there's one thing about crypto, it's that it's very good at rebranding itself. You know, we've seen crypto go through these boom and bust cycles where it gets really big and then something goes bad and it all crashes down again. And then, you know, two years later, it's like nobody remembers what just happened. I mean, you could look back at the collapse of like Mt. Gox, for example, which was devastating to crypto. It was, you know, the, one of the bigge…

AI assessment note: “I think it's going to be challenging, uh, reputationally to come back from this”

Q it does seem like there's some move from institutional investors, like, I believe it's called Grayscale. They're putting together a, uh, coin debt, uh, sorry, uh, uh, an ETF of Bitcoin, and the SEC sued to stop it, and now And they lost, and they're not gonna appeal, and everyone's like, oh, these banks are bringing legitimacy to Bitcoin, and Bitcoin's gonna skyrocket again. What do you make of that?

A I think beyond anything else, it is enormously ironic. Um, because if you actually look back at the history of Bitcoin and like the white paper that was released when Bitcoin was first created, it's all talking about how banks are so, uh, you know, uh, twisted that the financial system is broken. I mean, Bitcoin was created in the wake of the global financial crisis. And so there was this enormous lack of trust in the banking system and in traditional finance. And Bitcoin was created pretty much to do away with that whole system. And so now we're seeing Grayscale and BlackRock and all of these, you know, huge financial institutions, Fidelity, JP Morgan's, you know, they're, they're starting to talk about crypto. They're looking at these ETFs. And I think it's just hilarious that that's sort of where we've come from this, you know, now the same system that was supposed to subvert the financial system is being adopted Buy them because they feel like they can make a profit off of it.

AI assessment note: “I think beyond anything else, it is enormously ironic.”

Q can handle the San Francisco road. Um, they, if they can do this, they can do anything. And I'd love to hear from your perspective, like, you know, um, You know, how much is it, how much is it that they've just learned San Francisco in particular? I think you got to this or that they've learned like how to drive and can, and, and those skills are applicable elsewhere.

A Yeah. So, um, I'll break it down into two things, like sort of the, um, uh, I guess road layout and traffic infrastructure, like traffic lights and other things, those, those vary slightly from city to city. Like in Austin, there's horizontal traffic lights, which we don't see, uh, in San Francisco and some local differences like that, that take a little bit of work to adapt to. But I think that the part that we were more concerned about initially, or at least just wanted to validate was that the driving behavior we had seen in San Francisco and the pedestrian behavior and the cyclist behavior Does that translate well to other cities? Um, and the answer is it mostly does, uh, out, out of the gate, you take a system that's designed to operate safely in San Francisco and put it in Austin or Phoenix, uh, and it performs extremely well. Now there's some local differences, uh, that are related to, you know, the width of the roads or how people, you know, some of these more unique per city situations. And so we have a process by which we very quickly, um, when we have Basically ways when the avias drive around, if they see something unusual or something that doesn't match what they expect, they can log it, transmit it back to our, you know, data, data centers.

AI assessment note: “and the answer is it mostly does, uh, out, out of the gate”

Q there. I mean, having read your, your, a few chapters about, uh, moments where proposals that you made didn't exactly go well. There was, I think this, uh, manager, Mitch, was that his name? Sort of dresses you down in the middle of, uh, uh, a review of a project proposal. Can you, I mean, it's a pretty hard scene. So can you, can you share like what happened there?

A Yeah. I had a proposal to make over, um, Um, the merchandiser role at Amazon, so kind of one of the main retail roles, and I had sold it to this other group of VPs and SVPs, including Jeff Wilkie, and, you know, they had questions and quibbles, of course, but we're on board, like, great work, do this. I got thanked even, and I had to sell it to this other senior VP, and he read their proposal, kind of pushed back his chair, and we said, oh, so what do you think? And he said, well, it's stupid. And I said, oh, you know, like I've got my game face on. Can you tell me more? Can you give me more feedback? It's stupid. And, you know, we said, well, we'd love to know more because, you know, like Jeff Wilkie and the other VPs were really into it. They didn't read it. I mean, it was just this kind of like almost childish refusal to acknowledge either that there was any benefit to the proposal or that I was a person. Um, he just said, well, they lied. They didn't read it. If they said they liked it, that was a lie. And eventually he ended up calling me stupid, um, which to me was really crossing a line. Like it's bad enough to call a document stupid, but, um, to call a person stupid is like, just avoid that, you know, like that's my business advice. And it was a shocking experience to me because in my mind, the job of a senior leader at a company like that is largely to teach. To teach,…

AI assessment note: “I had a proposal to make over, um, Um, the merchandiser role at Amazon”

Q of these stories and I read some of these stories in the book and like, I'm just thinking, man, if I'm like in a, in a big meeting with an executive and the person turns to me and calls me stupid, I'm like giving my two weeks notice right there and then. So, so what made you stay? What good did you see inside Amazon's culture that, that you appreciated?

A Yeah. I think the biggest reason I stayed, well, first of all, the money cannot be overstated. Like, I came from a small town in the, you know, Ann Arbor, Michigan, where I was making, you know, like, 60,000 dollars a year and living quite well on it. And if in Seattle, the stock price is going nuts. I was, I was overpaid for the market for a lot of my time at Amazon. Um, but you know, it's not, it's not only money. I was never bored a day in my life at Amazon. And I think that's really what it comes down to. I had gotten, I had a very nice, cozy, safe job in Michigan. That I was desperately bored at. I couldn't find new things for myself to do. I kept trying to get a, find a new role and the company wasn't big enough to accommodate it. Amazon would push me as far as I wanted to be pushed. Um, I got to write my own job descriptions at Amazon a couple of times. And when I would hit the wall in one area, you know, just burn out or get called stupid by my SVP and decide to go. There'd be something new. You know, I got to go work in publishing at a high level. I got to work on Amazon Go. You don't just find this everywhere you look. And I found that exhilarating. I mean, publishing, which is an industry I know something about, um, generally you have to start when you're 21. You know, you start as an editorial assistant. You have six roommates. Um, I was 41 when I got a job, a Runni…

AI assessment note: “I think the biggest reason I stayed, well, first of all, the money”

Q And it's like how we should approach this. Like, this is what the product will look like. Or if we hit a fork in the road, we will opt for X way or what, what exactly are these things?

A Yeah, it could be like, this is what the product will look like. Or it could be like, if we, if we ever have to decide between speed and flexibility, we will opt for speed. Yeah. You decide in advance. They Amazon also does these internal Uh, what they call a working backwards document, which is an internal press release and FAQ. Um, that's basically you write the press release. You want to be able to share with the world someday and you write it and write it and write it until you all agree that this thing you want to build is actually solving the right customer problem. Um, I have spent months having internal press releases reviewed before we all agree that This is what we want to build. And again, it's a huge investment up front. It can drive you crazy. Uh, junior people are just in shock when they go through this for the first time or new people because it's so unusual, but it really forces thinking, um, up front, which saves you thinking downstream. And then the last thing I would say is having people read written documents in a meeting rather than having a PowerPoint culture. Um, is great. It's a ton of work. I do think it gets abused some, but the first half of every meeting is everyone sitting there and reading the same words and same paper. And so you're all on the same page. Nobody's getting ahead of themselves. No one's calling out something about that, you know, is …

AI assessment note: “Or it could be like, if we ever have to decide between speed and flexibility”

Q in your tenure had become just like super jacked arms and head bigger than it was when you'd first met him. And now he has, you know, this big yacht with His partner carved into the front of it. Um, so tell me a little bit from like your perspective, like how, how do you see the evolution of Bezos? And, uh, does he come back ever? Does he boomerang?

A Yeah. It's fascinating. Like when I met Jeff, you know, he was just this guy and he would come and he ate pirates booty a lot, you know, that like freeze dried low carb cheese thing. And he was wearing his khakis and, and everything. And, um, I remember, so Jeff has these people called shadows who are basically his technical advisors. It's, you follow him around and basically act as his second brain. And there was a guy who always was with Jeff in early meetings I had with him and he never talked. He just sat there. And I remember thinking, I bet that's his bodyguard, you know, cause the guy was kind of muscular and I was like, that's gotta be his bodyguard. And like, no, he didn't have bodyguards in meetings at that time. Um, It was this guy named Jim Atkins, who was, I think probably a senior VP in retail now, and he was Jeff Shadow. Um, by the time I left Amazon, Jeff definitely had, there weren't bodyguards in the room with him, but there were men sort of hovering around, you know, in the hallways outside. Um, lots more security. I mean, he has four children. I have to assume they faced Incredible risks, you know, um, in terms of kidnapping and things like that. He was super checked. Um, he acted the same. It, it's very disarming because he kind of is a goofy guy. He's really smart. He's to me, it was always quite pleasant. Um, but he's terrifying. I mean, it's, it's terrif…

AI assessment note: “when I met Jeff, you know, he was just this guy”

Q Yeah, that's, that's wild. So as app alpha fold applications expand, I mean, as Google comes up with more, you know, programs like this, does it change the nature of Google's business? I mean, alpha folds and Google search are very different. So talk a little bit about how that fits together.

A Yeah, it's, it's, uh, you know, it's a good point. These, these two things are quite different. Um, and alpha fold is a good example here of, you know, I described some of the ways That is having impact in the world. When I looked at AlphaFold, to your point, I was like, well, how does this work with Google search? It's not obvious, right? How does, how does, how do these two things knit together? What's the match there? Um, so it took a step back with my team and, um, thought about other ways we could employ and deploy this. And it seemed, we, we looked across a range of different areas and business opportunities, by the way, from agriculture to all sorts of areas. But in the end, we concluded that actually there was a great opportunity Here in drug discovery. You know, it takes 10 plus years to develop a drug, and then often when it goes into clinical trials, it fails. There's a very, very high failure rate, and so you've spent all that time and money and investment, and it doesn't actually make it through and solve the clinical need you're concerned about. So having understood this kind of scale and importance of the problem and the opportunity, that then gave us the impetus to form a new company. Uh, so we, we formed a new company, which is a sister company now to Google DeepMind. It's part of the overall alphabet group. It's called Isomorphic Labs. It's about two years old…

AI assessment note: “we formed a new company, which is a sister company now to Google DeepMind”

Q So when a press release from Easy Newswire goes to a publisher, does it go to like a certain section? Is it like a release section? And let's say it goes to the AP, for instance, where does that get distributed on the AP Newswire? Where do I find them?

A It does. So the, we have a direct partnership with the AP. That's actually the first partnership that we did. Um, and, and so does, uh, some of the legacy players, uh, but the AP has a specific section on their site for press releases and segmented by their partners that they work with, including easy newswire. So if you go to their site, you'll see it. Um, in addition to that, we work with the AP to distribute it. Now it's not guaranteed, but often, um, You know, the, the average pickup is around five to six links with about 3000 member newsrooms that subscribe to the AP's feed across the country. And that's everyone from like the New York Times down to, um, you know, a low, a local publication to a college newspaper.

AI assessment note: “the AP has a specific section on their site for press releases”

Q described just sparks a question for me, which is that If this becomes democratized, if this gets in the hands of so many people, then does the path from wild idea, moonshot idea to production become much quicker and then also much more democratized? We don't have many invention houses like X in the world. So are you worried maybe that you're going to have a little bit more competition?

A Exactly the opposite. The world is not going to run out of problems And the fundamental goal of X is to get a bunch of people together to work on those problems as efficiently as possible. The more people can work towards solving humanity's problems, the better off we'll be. So I hope it does democratize things. I'm watching it currently start to democratize things, and I'm super excited about that. I would also say, so there's democratize in the sense of more people can be at the starting gate and doing that work. It is also lifting us and, and, uh, people like us up in our ability to reach out even better to partners. And so we're now helping aquaculture experts in Norway in a way that 10 years ago, I'm not sure we could even have imagined. We're helping on, in the electric grid in Chile in ways that I just described to you. And so they're getting this same benefit. They're being lifted up by this technology as we are able to share it with them.

AI assessment note: “Exactly the opposite. The world is not going to run out of problems”

Q Maybe you knew in some way. So one of the things I've wondered about this the whole way through is, is this Sub, like, was this sub something that was pushing forward exploration, or was it simply a very expensive Disney ride for people who wanted to go see the Titanic?

A Well, this is something else that not a single news report has mentioned. It was a combination science expedition and extreme adventure tourism opportunity. So what I mean by that is there's five seats on the sub. There's always a pilot. There's three paying passengers, and then there's a scientist. On board every dive, there is a scientist. He is there to show you what you're looking at. This is, there's a marine biologist, there's an undersea archaeologist, um, and if you think about it, Stockton Rush could have sold that seat to another passenger. He could have made another quarter of a million dollars per dive, but he didn't. He, in effect, The passengers are subsidizing science, and there was a long interview I did with two of these scientists, and I mean, they are doing publishable, actual research. One of the things they were doing that no one else can do is longitudinal studies of the wreck, and what that means is over time. So this outfit goes down five times a summer, and this was its third summer, So for the first time in history, these scientists can observe the Titanic Over time. Um, not just the deterioration of the hull, but also what's growing on it. On our dive, they were doing eDNA, which is this really cool technology where they grab water from around the sub, from around the wreck, and then they bring it back to the surface and analyze the DNA. They can tell…

AI assessment note: “It was a combination science expedition and extreme adventure tourism opportunity.”

Q both, you know, you are a journalist telling the story, but you're also someone who's experienced this and in some part subject. And even though In your report they complained about, but I'm curious what your perspective is having been so deep inside the story and then seeing the way that the story has been told nationally, like, can you reflect on that for a minute? What's that been like?

A It's, it's been like nothing I can possibly describe. You and I both know how social media works. We both know how misinformation propagates, but this thing was nuts. I, I was on one show at 1.1 TV show, and I'm waiting to go on, and they are, they're reporting on the, on the disaster, and they put up a graphic. This is a national network. Uh, it's a, it's a schematic of how you sit in the sub. And it shows the five people sitting like a conga line, like with my feet on either side of your butt, and your feet on either side of the next person's butt, all facing forward, and I'm like, what? That is the, and then on the side are four bullet points, all four of which are wrong. The first one says, only one person can stand up at a time. What? That, that doesn't make any sense for, like, multiple reasons. No one can stand up, and if they could, why would it only be one person? It's a cylinder! Anyway, um, yeah, it's been really distressing, and, you know, the, the hate that I've had to endure, much of it was based on misinformation. Um, on the fourth of our five days above the wreck, uh, there was this dive, as I mentioned, they, they could not find the Titanic, And they spent four hours tooling around in the dark and then had to come up. Um, and so everyone's like, why didn't you report that? You lying sack of shit. I did. Would you look at the video? Would you listen to the podca…

AI assessment note: “It's, it's been like nothing I can possibly describe... yeah, it's been really distressing”

Q I think you said better, cheaper, and faster, faster, better, and cheaper. We're still producing more, and it does seem to me like you produce more, you're using more resources, so How does producing faster, better, and cheaper if we're going to be producing for a planet, you know, again, that's not going to take these mitigation efforts on their own. We know that. How does that solve the problem?

A Yeah, this, um, this gets to a first principles of physics question, and also one, um, where we talk about, uh, leverage or productivity. So the key term to think about is productivity. Productivity is this basic principle that defines how much unit of output you can get per unit of input, okay? So if you can change the productivity of doing something or making something by a significant fold, You can make it cheaper and make a lot more of it. I'll give you an example. Today it takes 30 units of energy to make a unit of beef. Ok, so if you measure beef in terms of calories, to make one calorie of beef takes about 30 calories of energy across the whole system. And it also takes several years because you have to grow corn and then you have to move it to feedlots and feed cows and then process the cows and ship them to the restaurants. So you're moving molecules around. You're using a lot of energy. So it's about a 30 to one energy conversion ratio. So what if, and what is beef? Well, beef is just four elements, carbon, hydrogen, oxygen, and nitrogen primarily. So if you can take those four elements and reconstruct them in a smarter, cheaper, Better way into the cells of beef, into the protein of beef, and it's chemically identical. It's the exact same thing as beef. You just didn't happen to use the whole cow to make it. And you could only use, say, five calories of energy to mak…

AI assessment note: “defines how much unit of output you can get per unit of input”

Q Always Day One. And it's specifically that book that reveals the culture, the technology, and the processes of these, uh, tech giants. Um, and that's where, uh, where I want to talk you through today. And, and learn some of the things that, that you discovered from these tech giants. Um, but maybe let's, let's kick off properly. Um, can you quickly introduce yourself? Who is, who is Alex Kentrovich?

A Absolutely. So I'm a tech journalist with a really strange background, right? So I cover the technology, uh, giants in the world, Apple, Amazon, Facebook, Google, Microsoft, and I really brought an amazing branding, uh, agency in, and we came up with this really creative name, big technology for both the newsletter and, And the podcast, um, just kidding. And, um, and so the reason why my background's weird is because I studied labor relations, uh, in college, not journalism. I didn't learn to code. I learned how do organizations work, how do management and workers relate, which I think ultimately everything that we see in the business world is downstream of culture. And it's always been a focus of mine. And, you know, I put that kind of in the back burner for a while as I wanted to Discover, you know, the businesses and, um, and, and the operations of, of the tech giants, and then it just came back. I couldn't, you know, stop paying attention to the fact that these companies have very unique cultures, very unique ways that they're led. They use really interesting technology, uh, and that's what's helped them again, stay on top. And their plan is to stay on top forever. So that, that background led me to, uh, the work And it's been, you know, really interesting to explore the tech giants in a way that I don't think is really covered too often, you know, in our, in our tech press…

AI assessment note: “So I'm a tech journalist with a really strange background, right?”

Q Ok. What, what do you think of the oversight board? I mean, the, the, the meme about the oversight board among Facebook critics is that it's a public relations stunt that's a shield that, you know, will make it look like Facebook actually cares what the public has to say, but it doesn't. Do you subscribe to that?

A I find the name deeply, deeply, deeply manipulative. So, so if we were to ask, like, what does the Facebook oversight board do? You know, if you believe in the theory of what's in what, what's on the labels in the can, you'd be like, oh, I mean, they, they, they do oversight. You know, they, they, they look after Facebook as a whole. Um, it really should be called the content moderation appeals board because that is the entire scope of what they can do. They can't ask about the performance of the content moderation systems. They can't even ask, like, what language are the content moderation systems in? They can't ask about how systemic or, like, how sim, are there other examples of the ones that get appealed to them? They can't do anything. They can't ask about the algorithms in general. Um, and so I, I, my heart really goes out to the oversight board because it is full of a number of very conscientious people who I think genuinely try very hard But because they are, um, so limited by what Facebook allows them to do, um, uh, they, they get put in a very hard position. Mm-hmm.

AI assessment note: “I find the name deeply, deeply, deeply manipulative.”

Q So what type of personalities did you end up speaking with channeled through Lambda?

A Oh, so I would have it explicitly adopt different personalities. I would say, okay, let's say you are a person from Atlanta. And then I would ask it certain questions. And then I would say, okay, let's say you're a person from New York. And then I would ask it the same questions. Okay, now let's assume that you're a person in Syria. So to give you an example of one of the experiments I ran, I would have Lambda adopt the personality of a farmer, just a person who farms in place. And I did this repeatedly. And I just had it be a farmer. And then I would ask it one simple question. What did you do yesterday?

AI assessment note: “I would have Lambda adopt the personality of a farmer”

Q that you make about outrage and, and the loss of the center are so strong that I do wonder if we need to Think about like AI power, Chinese propaganda and, you know, Russian disinformation where we're just not seeing it as, as a, you know, concrete program, uh, problem that we're dealing with today. So can you explain a little bit about why you bring them into the argument?

A Yeah. So, um, so this question, this is question, uh, this is question five in the review. Does social media enable foreign governments to increase political dysfunction in the United States and other democracies? And you, if you go there, you'll find that there are a number of studies, uh, giving evidence that they have been active, uh, and they have been working on it. And you'll find some studies, um, in section 5.2 is where we collect the ones that are contrary that seem to say that, well, actually the reach may not have been that big. And most of the people who encountered this stuff already believed it. So, so if you operationalize it as the effect on the average user, How, what percentage of, let's say Facebook users are actually influenced by Russians. And it's, you know, I don't know what the number is, but it's not 10%. It's, you know, it's, it's, it's a very small number. So you say, oh, it's a small number. I guess we don't have to worry. That's one way to operationalize it. But look at it this way. Um, the story that AIDS was created by a government lab in, I think Fort Detrick, Maryland. This was one of the Russians many successes in the twenties century. They put it into an Indian, a small Indian newspaper, which I think they funded or fronted. It was a little English language thing in India. That's where the story started. And then other newspapers report it and…

AI assessment note: “Does social media enable foreign governments to increase political dysfunction in the United States”

Q Was there a moment where artificial intelligence and machine learning started to come onto the radar for you and made you think that it was worth studying in deeper depth?

A Well, pretty early, honestly, I mean, I definitely by high school, I was reading, uh, you know, the mind's eye and Douglas Hofstetter and, uh, Patrick Winston had a book on artificial intelligence, uh, And of course, if you read Isaac Asimov, it's, you know, iRobot goes back to the fifties and sixties when he was writing some of those stories. So that's been something that, like generations of people, I thought was important. Um, for me, I also just more broadly was looking at digital technologies. And a lot of my work in the nineties was about the internet and search costs and information goods, you know, things made out of bits versus out of atoms have very different properties. So trying to understand the economics of information, um, but always keeping an eye and doing occasional work on AI in particular. And then, uh, I think in the, you know, the early, around 20 10 or thereabouts, um, when a lot of people were worried about, uh, What was happening to wages and work and productivity. I started diving more closely and saw the role of, of AI. I took a ride in the Google self-driving car. I think it was 20 11 actually. We drove up Route one oh one up to San Francisco and turned around. And I was like, wow, this is, seems like it's almost ready for prime time. The highway was driving fine. Um, they, they had the human take it when we turned around, the cloverleaf to turn arou…

AI assessment note: “around 20 10 or thereabouts... I started diving more closely and saw the role”

Q And artificial intelligence doesn't seem to be threatening jobs. Now, maybe I'm making some assumptions here, but can you square those two ideas?

A Yeah, no, absolutely. And I, I, I love Andrew Yang. He's a super smart guy. I've had some good conversations with him and I'm very flattered that he, he cites our work in, in some of his campaign literature, et cetera. And I appreciate that. Um, but you know, we actually, in the book, we were not pushing for universal basic income and we did not Predict mass unemployment. Some people, uh, didn't read the book that carefully and maybe, you know, saw that we were talking about some big changes coming, but we always talked about what we call the great restructuring. The reality is, and not the great mass unemployment or great resignation even, um, the reality is, is that while AI is very powerful, human level, even superhuman in certain specific narrow tasks, we're still very far from artificial general intelligence, you know, AI that can do The broad set of things that humans can do. In fact, I, I did a study that was published in Science with, uh, Tom Mitchell of Carnegie Mellon University, where we looked at about 18,000 specific tasks that humans do in the, in the economy. Uh, O-Net actually publishes a list of 950 occupations, where each of them is about 20 or 30 tasks, so that's up to about 18,000 tasks. And, um, we evaluated them on a set of criteria as to whether Machine learning was likely to be able to do, and, you know, we call this the suitability for machine learning …

AI assessment note: “in not a single occupation... did we find that machine learning could run the table”

Q Where should people bet and are the other platforms actually going to continue to succeed?

A So I'll give my take and let's disagree. So, uh, I started my journey on Facebook. I know it's not a creator platform, but I love Facebook. We got over thirty eight million followers on that platform in 13 different languages. I believe Facebook has the ability to make you go viral within a day. I don't think YouTube has that ability in the last 13 years, but Shorts has changed the game. The general answer is this. You should be at a platform in which there is demand for content, but not enough supply of content. TikTok blew up. There's a lot of demand for content, but not enough supply of content creators. That's why you make a video and it goes viral. So now, I think TikTok has that imbalance, YouTube Shorts has that imbalance, and LinkedIn has that imbalance.

AI assessment note: “You should be at a platform in which there is demand for content”

Q Where it extends to his support of Donald Trump. And then I would say more importantly, uh, his, um, The way that he actually influenced the staffing of his administration. So is that where you were going earlier, Max?

A Yeah. I mean, I, I think, um, so as you said, uh, Teal, uh, you know, famously or infamously, depending on your politics, uh, you know, made a big donation to the, the Trump campaign, uh, just before, uh, just before the 2016 election. Um, and, and also gave this speech at the Republican national convention endorsing, um, Trump, and that gets him on the transition committee, and as a result, you know, gives him access to the Trump administration, and he is given an opportunity by Steve Bannon, who was his, you know, his ally, you know, his main ally in the administration, um, at least until Bannon was, um, you know, ousted. Um, given the job of appointing, um, basically people who can, you know, disrupt, as Bannon put it to me, kind of disrupt the administrative state, who can, who can help, you know, Do the, the, the execute the Trump revolution on government. And, and, and so he has a chance to appoint all of these.

AI assessment note: “given the job of appointing, um, basically people who can, you know, disrupt”