Guilty

Speaker

I've got sunshine on a cloudy day When it's cold outside, I've got the month of

Speaker

November the 3rd, 2023. If it's Friday, it must be that was the week. Another show with my old friend Keith Tyr on the week in tech and two dominant stories this week both associated with guilt. The first of course is the Sam Bankman Freed verdict. It came in late last night and Keith leads. I'm not sure if this is one of your AI pieces of art, Keith, but it's unambiguous. A photo of Sam Bankman Freed, his hair looking relatively well cut and a big blue guilty. So what do you make of the trial and the verdict, Keith? Well the verdict obviously is a function of the trial. It would have been very hard for the jury to come to any other conclusion based on what I read about what happened in the courtroom. I think what happened in the courtroom is the defense strategy was flawed at the very core, especially in the era we're living in, where Bankman Freed attempted to blame his colleagues for everything that happened, claiming that he himself was unaware due to the fact that he was a, let's call it a loose leader, not a hands-on CEO but a hands-off CEO. That was the story he tried to tell. Except on the women who worked for him, he wasn't hands-off with her, was he? Exactly, that is indeed true, Andrew. We can always trust you to go to the dirty core of a thing. Well I bet that's a piece of the story as well, which I'm sure is in some way relevant. It is, and you know obviously all of those colleagues had negotiated deals to testify, so they have yet to be sentenced, but they are going to be sentenced. So you use the word Keith, and you're not a lawyer, but you use the word defense strategy. Is that a euphemism? Did they actually know what they were doing? Given that everyone, all the people around him, all his closest friends and associates and his ex-girlfriend all nabbed him, all were turned by the court, by the prosecutor, what strategy was there? I mean in the end he came out and defended himself, but was he or his strategy or his lawyers, are they just profoundly delusional? I think they may be incompetent as well as delusional, because I mean let's just talk pretending for a second that we were his defense counsel. I would have thought that the best defense would be the defense that said everything was reasonable until it out of control. That the laws were not broken, but bad behavior happened, which in the context of the industry, the crypto industry, you know is probably normal, and only when the market declines. When you say normal, what do you mean normal? You mean it's not what you're doing? No, because it's an unregulated industry, scrutiny over the uses of cash is zero. So you know we don't really know what goes on at Binance or any of the other exchanges, because they're not regulated and there's no reporting requirements. So the best defense, if he wanted to mount it, would have been that in the context he was just like everyone else, and that had the market not crashed, triggered by Binance saying it was going to sell the stablecoin, the FTX called FTT coin, that actually no customer money would have been lost. And by the way, I was just watching CNBC, it looks as if no customer money is going to be lost. It's almost all been found. So the best defense would have been there's no crime here, whereas his defense was, well there was a crime, but it was these guys that did it and I didn't know. Is there a bigger cultural story here? I mean, Michael Lewis has written about this, others, there was a good piece in the Times today, I don't think it got, it was too late to get into your newsletter by Genia Belafonte on Bankman Freed being a grown up criminal. Can't blame his parents, but he had such a weird, not a classic, but an extreme Palo Alto upbringing where he was treated as a genius. And the mother who teaches at Stanford Law School has made it clear that he never lies and he could never lie. And the parents, very prominent Stanford Law professors have been visible at the trial. Is there something sort of surreal here that these people cannot acknowledge that they do wrong? Well, I think it's in that gray area between a lie and a delusion, where the person deluded may well think that they're not lying, even though they are. Because a lie by definition has to be conscious. Otherwise it isn't a lie, but it's wrong. And I think the word lie is misused so much in our modern culture. Anything you disagree with, you know, is termed a lie. I think it's more likely delusional. And delusional in the context... Are you becoming one of those cultural relativist Keith that you love? No, no. I mean, he clearly, if you ask me, was he right or wrong? He's clearly wrong. If you ask me, did he lie? That requires a level of... If he grew up in this family where he was so, you know, the Palo Alto world of enormous privilege, this liberal aristocracy, where he's so spoiled, and so fated because of his academic intellectual skills. But he was sort of himself. And I don't know whether this is undermining, I mean, claiming he's more or less guilty, but he entered a world of delusion. And this is just a manifestation of that. So well, so you like the theme about Palo Alto. But actually, my observation about Palo Alto is this incredibly Presbyterian, and woke, and strict, like kids are not, they don't have TVs in the in the living room, because they don't want their kids. I bet he was brought up without TVs. And they don't play video games. I mean, my kids have friends who have never seen a video game. So it's incredibly like Lutheran. And this goes, this isn't a religious designation, because Jewish Palo Altos are the same as, as Protestant Palo Altos in this regard. And so there's a strictness, the entitlement, you don't really see it at the level of children. I mean, I mean, my kids are all in local public schools. I bet he went to a local, but so you don't think there's a broader cultural story here. You just think it's an aberration that the Bankman Freeds and Sam Bankman Freed are just this weird family that just got completely out of control. Actually, I think there is a broader cultural story, but I think it's to do with the crypto industry. And when the crypto industry, starting roughly in 2017, when he was younger, meets the Palo Alto. And let's use that term a bit loosely, because it's global. It's not just Palo Alto. But that, that kind of millennial Gen X, Gen Y mentality of ultimate freedom, from responsibility, and accountability. I think the crypto industry and that style of person came together to create an unreal world of wealth, that they really didn't understand there would be an end to. So they behaved as if it was forever. That's the delusional piece. And I think they got swept up. I mean, you really have to have lived in crypto. I got involved in crypto in 2017, when I was exec chairman at ADV in London. And as part of my, you know, the stuff I did on my own account, I did a lot of advising ICOs, and initial coin offerings. And when you advise on an ICO, the founders are raising a lot of money by selling a token that they've invented. And 10s of millions of dollars were flowing to them. And they did, they do get into this delusional world, where this thing is like, you can just print money out of out of thin air, and do whatever you want with it. Later, around 2019, these crypto plays got connected to the regular banking system. And so you cash out in dollars or euros or whatever. It went crazy in China, it went crazy in Japan, it was a global madness, that was very different to the web three narrative about transforming how software is developed, distributed, although it's crypto is a piece of web three, it's hard to separate the two. It was a speculative part of web three. And I honestly think that is more the story than Palo Alto. So crypto is the speculative, the speculative piece of web three, you behave, I like how you put it, Keith, these people behaved as if it was forever. And certainly, Sam Bankman Freed did. Is this the end of it? Is this the, is this, is this the beginning of the end, the end of the beginning? How would you define it in historical terms? I don't think it's the end, actually, because I mean, one has to be plugged into the world of crypto to understand what's happening today. But there's still, there's still a lot happening in crypto, a lot of money being raised in crypto, less than before, but still significant amounts. And crypto and AI are starting to come together as as kind of twins in some of these things. Now, the benefit of crypto is it creates a way to pay for resources on networks that doesn't require individual subscription, like web two did. And so there's something redeeming about crypto as a functional way of measuring the use of resources and paying for them. I don't think that's going away. We are speaking with Keith Teer, the CEO of SignalRank, one of the smartest AI financial investment startups, and also the author of That Was The Week, a regular feature on the show. Keith, you've jumped the gun a bit. The other big story of the week is AI. It's always a big story, but a particularly big story when it comes to guilt. You suggesting that AI, and you had an interest, a very good editorial this week, AI, in a sense, is in the world court now. I'm quoting that this week, AI was also put on trial. There were White House and Downing Street meetings for some of the biggest hitters. Sam Altman, of course, Elon Musk. How meaningful was it this week? Was it all symbolic, cosmetic, or was real stuff done in the White House and at Downing Street? I think it's mainly symbolic. You kind of hope that it's mainly symbolic, because if governments actually start requiring you to interact with them before you do anything, it's going to slow AI down terribly. I think it's mainly symbolic. I think there's some competitive element to the symbolism. Pretty much every government in the world wants to get on the stage and have a point of view about AI. They are all inviting the same very small group of people to attend, mainly to get photo time. Rishi Sunak yesterday did an onstage interview with Elon Musk, where he was the interviewer. Sunak was the interviewer. I think there's a fair amount of posturing and not a lot of substance. That said, the lobby within AI that wants it to be regulated, and that includes the leaders like OpenAI, will see this as an embracing of their legitimacy. When governments want to meet you, and you go and meet them, and you discuss things that end up with written agreements, whether vacuous or not, it's a form of legitimization. The term that's being used is regulatory capture. That is to say, by having these relationships to governments, it makes it super hard for new AI companies, of which there will be many. I don't know if you had any of these pieces in. I haven't seen them in your pieces this week. There are a lot of articles criticizing Bankman Freed and the other heavy hitters, Google's, OpenAI's, Anthropics. In trying to do a deal with the governments and being regulated, the argument is that they're sewing the market up and controlling it. What's the argument? Is there any truth to this, do you think?

Speaker

It's very hard to tell. That's a little bit like the question, was Sam Bankman Freed lying or delusional? Is Sam Altman self-consciously manipulating governments to anoint OpenAI the winner? Or does he genuinely believe that OpenAI needs to be regulated due to the potential dangers? I don't know the answer. I'll give Altman the benefit of the doubt that the second is true. Honestly, I don't know. What I do know is that it won't work because AI is such a dynamic thing that new and better products appear every day. I saw one this week. I used a chat GPT for co-piloting code I write. A new one had cropped up this week that is built into the Microsoft developer software that's better. It's from a small company coming from nowhere. Google's decision this week to put $2 billion into Anthropic gives the impression that it's a fight between Anthropic and OpenAI. Actually, it isn't. There'll be new things. I think regulation is mostly impossible. Particularly in an age where government is weak. You talk about Silicon Valley power brokers wanting to meet government. I would have thought it's the wrong way around. When you see a weak, somewhat joke of a Prime Minister, Sunak in the speak, who has the biggest direction. Not that I want to be vulgar about that sort of thing. You know, what's his name? The former liberal politician who's now at Facebook. Nick. Yeah, we can't even remember his name. Exactly. Nick Clegg. Nick Clegg. Tells you that politicians want to get to know billionaires because there may be a future. Right. Much more than billionaires want to get to meet politicians. They're usually pretty mediocre and becoming even more mediocre. When you think of Sunak or Biden, these non-entity politicians, either too old or too clueless. It's hard to imagine really. It's not that hard to imagine what's really going on here. It's the politicians wanting to meet the entrepreneurs rather than the other way around and steal their thunder. Yeah, I believe that's true. If you look at the content, Musk was a lot more philosophical and thoughtful about AI than the politicians. Because he knows more about it. I mean, what does Sunak know about it? What does Biden? Biden barely knows how to dress himself in the morning. So what do these people know about? What do they even pretend to know anything about it? I'm beginning to sound, I think, a little bit too much like you, Steve. So you think that there is no emergent oligarchy of big AI companies sewing the market up, creating this regulatory regime to benefit themselves? Unspoken, but clear. Well, I would characterize it as like the early days of WebOne, when Yahoo, Excite and Infoseek were the big guys and AltaVista. And then along came Google. And everyone was blind to Google. Suddenly Google eclipsed all of them, because it had better tech. Right. That's the big question. To me, that's the, it's the trillion, it's the multi trillion dollar question is, are we, is it conceivable that there is a Google in the wings? Not only conceivable, I'd say it's inevitable. I don't know where it'll come from. And it'll come from a couple of kids, FDA, Sam Bankman Freed style geniuses in garages somewhere, or just come up with something brilliant, you think? Even though it hasn't the structure of the industry changed, is that possible today? Well, there's preconditions. So in the case of Google, AltaVista already invented the idea of an index of the whole web. And there was a piece of software called a crawler. And you could crawl the web and then index it. And in the case of AltaVista, they didn't figure out how to rank the That was the missing piece. So AltaVista's results were decent, but not great. Google just copied 99% of what was required to build a search engine, and innovated in ranking by counting links. And appropriated the Stanford University servers to download the entire internet. Yeah, so compare that to now. Large language models are now a known thing. How to train them at large scale is now a known thing. How to build interfaces to interact with them is now a known thing. The next stage is probably what you best would think of as unbundling. So unbundling is like when Craigslist, suddenly, Trulia comes along and does a way better job at real estate than Craigslist. And so Craigslist is no longer the best place for real estate, but it's still good for car sales. And then Carvana comes along and is really good at car sale. So basically unbundling is when micro or medium scale opportunities in niches can be done way better as standalone than in the general context. And suddenly, the Uber AI of everything no longer looks as good for that particular thing. And then it's death by 1000 verticals. Death by 1000 verticals. Keith, you're coming up with some brilliant headlines this week. Death by 1000 verticals. Did you just invent that or steal it from someone? I think I just made it up, but it's not... Death by 1000, I may use that, death by 1000 verticals. What about this? Elon Musk, he's always in the news, giving ex-employees one year to replace your bank. Is that death by 1000 verticals? Is X transforming itself into a finance company, a bank? I don't even know what that means. Well, so this is a great example of how good entrepreneurs should think. Clearly, banking, if you break it down into its component parts, which is you give me your money, I'll lend it to 10 other people. So if you give me $1,000, I'm going to lend $10,000. And I'll make interest payments on that. And if you ask for your money back, I can give it back to you. But you probably won't. That's banking. That's the FTX version. By the way, we'll spend it anyway. So if you ask for it back, I haven't got it. Well, and it's the official banking version. And it's regulated. There's rules about what those ratios need to be. What Musk is saying here is, there's no need for custodial banks. That is to say banks that hold your money. In a digital world, you can have a wallet on your phone, it can have your money in it. And you can spend it anywhere. And people will offer you services to earn interest on it, but it stays in your wallet. And banks are going to end up as dinosaurs. Now between today and that outcome, is a leapfrogging of centuries of regulatory law. And he's saying that Twitter, because it has 350 million people, can replace the need for banks, which is what is also stated by people like Revolut, which is a... Yeah, we've done some shows on that. But isn't it more than money? It's reputation as well. And if reputation is more valuable than how much cash we have in our wallet, so wouldn't X be, rather than a financial bank, a bank of everything, if he's going to pull this one off? If he's going to pull it off, it will be. Just like WeChat is in China, people spend money from their WeChat app without a bank being in the middle. They can also plug a bank in. So it can be hybrid. But the long term is non-custodial financial transactions. That means no bank, no middleman. No bank. So any bankers out there watching, Keith says no bank if Elon has his way. It's certainly an innovative... Well, the other thing, Andrew, which is worth, we maybe can come back to this another week. But the criticism here is that you wouldn't trust X with your money. Well, actually, X would not have your money. So you don't have to trust them. Yeah, we trust X, for better or worse, with our reputations. To me, this is more, as you say, Revolut, they're more of a reputation bank. And it sort of reflects the changing architecture of our socio-economy, where our value is. Our value isn't how much money we have in our wallet. Our value, we've been talking about this. Steve Gilmore is very good on this. Our value is our reputation. Yeah. It's good stuff, Keith. We are in agreement. And I have something that is really going to warm your heart. I'm going to show you because you're going to be amazed because you accused me of being cheap. And look what I got. Oh, a new iPhone. Well done. I got the, and I got the, the Max, the Pro Max, as you told me. So I'm not as cheap as you think I am. I'm not going to descend into any of your ethnic taunts. And appropriately enough, your video of the week is an Apple event shot with the iPhone that I just bought. How good is this iPhone, Keith? Well, that, I recommend everybody watches that video because it shows you that the iPhone is as good as a top-end $120,000 Alexa camera. If it, in the right hands, if you know what you're doing, you can produce the quality of endgame, which this video is, that you would need to normally have an ARRI camera, which is what most of the movies are made with. And so it is, it's exceptionally good. It's always been, we've always heard this. Is this any different than before? We've always had, well, if you put it in the hands of a great cinematographer, and people have made films with good films, nationally distributed films with iPhones, is this a significant improvement? You know, it, in the case of photography, it's little things like how much light gets to the sensor. So can you do things in low light? And it's a very, very large improvement when you're looking at those kind of measures. For normal street photography, it's probably just the same as it was in the iPhone 10. It'll be a bit better, but you won't really notice. Whereas in extreme situations where you want to be cinematic, this phone can do it and the earlier ones wouldn't have been able to. By the way, speaking of Apple and Musk's announcement about X, I'm assuming that Apple must be thinking the same thing. They're as well positioned as anyone to create a reputation financial bank, aren't they? They could, although their relationship to Goldman Sachs is, is deteriorating to the point where Goldman Sachs don't want to support the Apple card anymore. They're looking for a way to offload it. Maybe they'll give it, maybe they'll do a deal with X. Yeah, I think Apple actually is probably too much tied into old school banking, and hasn't yet made the leap to where Elon is. By the way, LinkedIn user watching from rainy London, welcome. And also, LEO, a sun sal has made three comments. So thank you guys for listening and engaging. I hope they're productive and not destructive. I hope so. I hope I pronounced the sun sal right. I think it's Portuguese, but I could have done it wrong. Well, Keith, we're off to our last two features of our regular show. First is startup of the week, but this time you've cheated. And it's startups of the week. You've got two. One is in Musk's, you love Musk, his SpaceX is Starlink. And the other is a defense tech startup Shield AI. Why did, why didn't you just give it to one of these? I gave it to Shield AI. And then I read the Starlink thing. And I was going to put the Starlink thing in the news. But actually, I think it's very significant. Starlink is now profitable. It's doing 1.4 billion in revenue. From, from, you know, I mean, when did it begin? They probably started thinking about it, I don't know, five, six years. Is this the one where you can get quick, high speed Wi Fi anywhere you want? Yeah, it's low orbit satellites all over the world, with a receiving dish, which is your modem, that gives you high enough speed Wi Fi, I think it's 20 to 40. So we could still do this show on Starlink. We could do it anywhere in the world from from a desert. We can all go and live in the countryside, Keith, we don't get stuck in San Francisco or Palo Alto. Exactly. So, so, you know, imagine Musk has Starlink, Tesla, X, you start, you start to put together, and of course, SpaceX. And I mean, has the world ever seen such a big thinking creator able to execute and deliver profits? Right. And I think with Twitter, actually, or X is what I don't really understand is he clearly is a genius. Why he everyone when 95% of people think of him, they think of him as the owner of X. And it reflects him at his worst. I just don't know why. Well, that's, that's, that's, I would say, up to now, because if he if he really executes this, X is the everything app vision. Twitter, the functionality that we remember from Twitter will be part of X, obviously, forever. But X won't be thought of as being equal to Twitter, it'll be something beyond Twitter. And then you'll start to think, wow, he, he knew what he was doing all the way through. I wonder whether he'll just he should just fold X into SpaceX or fold some of these things. We will see lots of lots more Musk stories in the future. And then what about Shield AI, Keith? Why is that interesting? Well, I put it in because of defense tech. So I have a I have an odd reason for putting this in because the company itself is, it's highly valued and like a lot of AI companies, and it, it does. It got just raised $200 million at a $2.7 billion valuation, making it a classic unicorn in a kind of post unicorn age. Yeah, and it focuses on security. And it's a good example of this unbundling we talked about earlier, AI for specific purposes. Then by 1000 verticals. Correct. But the reason I put it in is because of the defense tech bit, which is, which is massively growing, and probably should be cause for concern, that the whole world is starting to build software for defense. Or, you know, obviously, defense can also mean offense. And the, and, you know, that debate about the nuclear bomb, which is, should nuclear fission have been used to build a bomb or to build power stations? And the answer ends up being both. And scientists get into this dilemma of, do they want to support governments who are seeking to be better at doing war than other governments? And that is a growing debate. And shield AI is in the middle of it along with Is Teal involved in this? He seems to just as Musk is controlling most of the world, Teal seems to be controlling most of the defense and military. I don't know the answer. I didn't go and do the research on who's behind it. So we can look at that afterwards. And finally, and you've been very good today, Keith, you haven't brought up VCs or investment or startup money, but you do slip something in at the end. The X of the week is by Alex Iscold. I thought he played for the Arsenal, but clearly not. A $3 million seed out of the gate or even after pre-seed with some traction is going to be very far and is going to be very far in between in this market for some time. The new reality is $500,000 to $750,000 pre-seed followed by 1.5 million to 2 million seeds. So things are going back to normal. Do you agree with Alex? I don't actually. The truth is that the valuations and amount raised at pre-seed and seed have held pretty steady and even trended upwards. And Harry Stebbings, by the way, replies to him in that thread saying, in your dreams. And I kind of agree with Harry Stebbings. Although it is an interesting, both Alex and Harry are super plugged in angel investors. So they, you know, they're not bystanders. So their point of view is backed up by their experience. So Alex may be right, but I suspect the best early stage companies can still attract larger amounts of higher valuations. What's interesting, and if we were to discuss it, is that is problematic because valuations are a function of what happens next. You know, whenever you invest in a company, the question you should have in your mind is what happens next? And if the answer to what happens next isn't, someone else will come in and write a bigger check at a higher valuation than you did, then your money is going to be, you know, stranded. And you won't make money from it. And I think the biggest story of venture is those larger later stage checks have dried up. But still, people feel that there's a couple of kids that wander into their office, probably looking a little bit like Sam Bankman Freed before he went to jail. They could be the next Google, which is still driving this early stage stuff. Is that fair? Yeah, that's what you look for. The word is being used a lot now. We had a headline a few weeks ago about the power law. And the power law is you're looking for the outsized winners. It's a little bit like on American Idol when Simon Cowell is looking for the next superstar. And, you know, there's lots of very good performers, but they're not the next superstar. So you don't focus on them. You're only looking for the next superstar. And that's true in venture. And so venture is a superstar business by nature. It's a superstar business, not necessarily to individualize it down to the founder. The companies become superstars. And I think that that's not well understood. You know, to play venture, you need one winner in a fund. The SignalRank version is different. You need 300 investments, each of them small, and you'll do better than someone looking for a superstar, like an index approach. That is a debate. But I think it's a super interesting time for venture because it's going to transform.