Bonfire of The Unicorns?

Speaker

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

Speaker

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

Speaker

August, I've got the month of August, I've got the month of August, I've got the month of

Speaker

August, I've got the month of August, I've got the month of August, I've got the month of

Speaker

August, I've got the month of August, I've got the month of August, I've got the month

Speaker

of August, I've got the month of August, I've got the month of August, I've got the month of August, I've got the month of August, I've got the month of August, I've got the month of

Speaker

August, I've got the month of August, I've got the month of August, I've got the month of

Speaker

August, I've got the month of August, I've got the month of August, I've got the month of

Speaker

August, I've got the month of August, I've got the month of August, I've got the month of Well, if you really get into the weeds, what Hopin did wrong, and a lot of companies do wrong is just because you raise a lot of money doesn't mean you need to spend it. So if you have a balanced view of your opportunity over, let's say, 10 years, raising a lot of money doesn't need to be a problem. But it is a problem if you spend it, need to raise more. Once the valuation era has shifted to a more reasonable valuation era, and the new valuation of Hopin, by the way, where they retain a company called StreamYard that they acquired, that's not unlike the software we're using now on our show, is $400 million. So from $7.5 billion to $400 million. And I think they sold assets to RingCentral for about $300,000, $400 million as well. So let's say $800 million. So it's 10% of its former value. That wouldn't be a problem if they'd preserve capital, they had a big bank account, and they could afford to take time to grow into a valuation. But because they spent the money super fast, believing that more would come, then they're stopped because no one's going to invest on top of a $7.5 billion valuation. Oh, but the investors, if they kept all the investment money, the investors will say, why are you keeping our money? The whole point of putting it in was for you to grab market share or control the market or do this or that. Yeah, but when you raise a billion dollars, you can probably do that with $200 million and keep $800 million in the bank. You don't have to spend a billion dollars. How do you short these markets, Keith? If you saw this stuff going on, and maybe now it's happening with AI. It certainly happened a couple of years ago with crypto. Yeah. And with these other virtual online platforms during COVID. I mean, even PayPal, which is a credible company, I mean, its stock is 60 and COVID times it was up to 300. How do you short that? How can a private investor short it? Well, the way that it works in private markets is people who have the inclination to short become secondary buyers after the problems have developed. So like this week, I didn't put it in the newsletter, but last week, we talked about Michael Moritz leaving Sequoia. This week, it was announced that he's teamed up with some other people through Sequoia Heritage to create a fund for buying distressed assets. So the secondary buyers basically become people who pick up distressed assets that still have life in them. Once the valuation is... Yeah, but Moritz has a lot of money. He knows what he's doing. But for ordinary people, for myself, if I saw this ridiculous explosion of tech investments in an area which I was convinced was bound to fail, then how can you bet against it? The only way to bet against it is to bet in public markets against stocks that mirror the ecosystem, like the ARK Invest Fund that Kathy Woods runs. Invest in a lot of unicorns. Isn't this a startup idea, Keith? Couldn't you do something like this? I mean, you've always wanted to, you always talked about democratizing the market. So you've always wanted ordinary investors, retail investors, to be able to have access to startups. Why shouldn't they also have access to betting against startups? I don't have the DNA for that, Andrew. I'm not saying you do, but somebody else. Yeah, well, that already is there, really. There are various hedge funds that you can invest in. You have to be an accredited investor, but there's various hedge funds that specialize in shorting, and they short public markets because you can't short private markets. And then there's others which are what you might think of as bottom feeders. So Dave McClure, who used to be a founder of 500 startups, now has a fund that buys assets from LPs in funds, where the assets have been devalued, and he buys them for pennies on the dollar. So there is a whole ecosystem around that, and you can invest in Dave McClure's fund. So in terms of the funds, another thing you mentioned in the editorial this week is that Union Square Ventures, one of the smartest, the most successful venture funds, has marked down its assets by 26%. How well are leading venture capital firms dealing with it? They're mostly delaying the inevitable. So Union Square actually got a bit of flack for doing this. Why? It's considered too much of a write down by some people. In fact, the opposite is probably true. It's probably too little. But the fact they did it at all is a leading indicator of what's likely to come. Again, hence the Bonfire of the Unicorns title. We're beginning to see re-evaluations of the asset value of portfolios. And I think that's a process. It's going to take a year or so because audits have to happen. And when audits happen, you're going to have to justify the valuation you're holding an asset at on your books, and you won't be able to. And I would say 50% to 60% write downs are going to be more normal than 26% write downs when that happens. And what does that actually mean? So if the US marks down by 50% to 60%, it's not going to jeopardise its business. It's just jeopardising that particular fund. It's the fund. So let's say you invested $100 million in 10 companies, and none of those companies have done the next round yet. And meanwhile, the market has shifted in terms of valuation norms. That $100 million, which is your investor's money that you've put to work, may now be only worth $40 million. So you've turned $100 million into $40 million. That's quite an achievement. Right. And you have to report on that. And then it takes a long time to get it back, if you ever do. Another piece that you highlight this week is Sequoia cuts its crypto fund by 66%. That's not actually cutting the valuation. It's the actual fund, right? Yes. Sequoia has this structure, which we've discussed before on the show, which there's a top level pool of money, and they're allowed to allocate from that pool into various initiatives, one of which is crypto. So this is them deciding to allocate less of their money to crypto in the aftermath of the FTX fiasco. And the two of the partners running the crypto fund left. We talked about that last week. So they're basically... Was that a euphemism for being fired, do you think, given all their missteps? I think in venture, you don't really get fired because you have some economics in the fund, but you don't... You're not invited back, you're not... Exactly.

Speaker

And one of the consequences, if you haven't put this piece in, because I think I must have forgotten to send you the link, but I think it's a really interesting piece in the FT this week by Ruchir Sharma, suggesting that the title is What's Wrong with Tech Giants Riding the AI Wave, not why. And the premise is that because of this collapse, or certainly crisis of early stage funding, all the innovation now is coming from big companies. And the big company news, as you know, is actually pretty good. Meta's numbers weren't bad, their stock is strong, the same with Alphabet. What do you think this is all going to mean when all the innovation comes from Meta and Facebook? And even OpenAI, which is one of the few successful AI startups, is essentially funded by Microsoft. Yeah. Well, firstly, it's accurate to say that that's what's happening. The costs of entry into AI are so high, especially the large language models that are dominating today, that unless you're a big company with a big bank account, you can't even begin to do it. OpenAI was underwritten by Microsoft, which is why they could do it. And Microsoft obviously has an ownership stake in OpenAI. There are very few others in the same league. I mean, there's four or five other large AI investments that have happened. But probably the biggest signal to how that might change is there's a company got funded this week called Together.ai. And Together.ai just did a seed round. So you heard it here first. And it's a platform for third parties to be able to carry out the kinds of training that the big companies can do. And it's all open source. It's a pay-as-you-go service, so it's not free. But it does indicate that there may be a democratization of AI capability coming bottoms up through the equivalent of SaaS companies in the AI space. So I'd say Together.ai is one to keep an eye on. And then another headline is broadly venture capital investments down 54 percent. And I assume all that investment is going into AI, into companies like Together. Not all of it, but a lot. But it tends to be going in large checks into a small number of companies. There are thousands of AI startups, literally thousands. I would think there's tens of thousands, actually. I mean, anyone who's doing a tech startup is by definition an AI startup, isn't it? Not all of them. No, there's a lot of deep science startups right now. We're going to get to deep science later in the show with your startup of the week. One of your headlines, which is interesting, which wasn't surprising again, is that you think or one of your one of the articles believes that AI will have the biggest impact in health care. That's a no brainer. And I assume that a lot of the money is going there. And a lot of money will go there. So far, not as much as he's going to. That's an Andreessen Horowitz article. And so they're selling their own book in a way when they say that. But obviously, human anatomy at the deep level is beginning to be understood. And there's no reason to believe AI won't accelerate that understanding. We already know that deep mind, for example, is able to model all the proteins in the human body, which scientists haven't been able to do. From the retail point of view, it's a no brainer. I mean, I've had lots of keen on shows about how doctors and nurses will be able to scale themselves in terms of dealing with medical issues, which currently are not very well dealt with on the retail front in terms of patients and doctors. Yeah, the real breakthrough comes when robotics meets AI, which it isn't yet. But we already have keyhole surgery. And we already have keyhole surgery, whereas surgeon can control a machine at a very high level of accuracy, down to millimeters in the body. So robotics is going to be better at that than a human once the procedures are understood enough to replicate. So I think you do end up in this robotics meets software meets human need. But the most exciting stuff, I think, is stuff I don't really understand. I'm not enough of a biologist to understand it. I thought you answered everything, Kee. Yeah, I definitely don't understand DNA. And I don't understand stem cells. And I don't understand, you know, creating viruses that attack cancer cells. But all of that human engineering is going to accelerate massively. And in terms of the broader impact of AI, as these big companies, the Metas, the NVIDIAs, the Alphabets, the Amazons, the Apples of the world control the platform, will it mean that there'll be opportunities for startup entrepreneurs in the verticals like healthcare or insurance or finance or entertainment? Yeah, the word I would use to describe all of that is unbundling. It's a cyclical thing in tech. The first thing that happens is some, you know, some mega breakthrough happens.

Speaker

Large language models is a good example. But if you go way back in time, the internet itself and the idea of a portal was, you know, there was only there was one portal, it was called Yahoo. And then there was two, Netscape. And then there were three, Excite, and Infoseek and others. And then there was Google, and AltaVista. And before that, it was AOL. Before that, AOL. Now, if you roll the clock forward, what happened? Today, we have specialist sites for very specific interests. They go to YouTube, we don't have a video site anymore, we have millions of video sites, each one dedicated to niche subjects. So unbundling is the process of technology being applied more and more specifically to human needs or wants. And I think with AI, you're going to see that in every domain, anything that humans demand, someone will build AI to supply the demand. And that's unbundling. So the big opportunities for startups is to pick a space, let's say, real estate, real estate sales, homes, you know, what will be the AI impact on Redfin or Zillow, somebody's going to build something that's way better at valuation, at selection, at access to purchases, and it's going to revolutionize something that already exists by making it better. That's unbundling, and I think that is inevitable. Is that what you, what one of your articles this week describes as historic futurism? You, by Kyle Harrison, subtitled Using Science Fiction to Invent the Future. Can we learn about what will happen from the past? Is it almost inevitable, Keith, with your Marxist background? Is this a kind of historical determinism working itself out? Well, there's no inevitability in history, that's for sure. So in that sense, the historical determinists are wrong. But technology does open up the possibility for human investigation and decision to focus on something. And historical futurism is my first, I can't remember if I made it the first essay, but it's one of my first essays. It's really good. And what it really says is that you can choose the future you want based on the available tools in the present. You can already imagine and create what comes next. So that to me is true. I've always believed that the entrepreneur really is a believer in something that doesn't exist, but can. The science fiction is the believer in something that doesn't exist and can't. So the difference between science fiction and entrepreneurialism is whether it can exist. Uber didn't exist, but it already could. And Travis Kalanick figured that out. Same with Airbnb. It didn't exist, but it could. And so historical futurism is reminding yourself that the present is only a canvas. And it's your job to pick the tools from the present to create something that doesn't exist, but can. And if it exists, people want. Yeah, you got to be audacious. I mean, at the time when Airbnb and Uber launched, they both seemed entirely absurd, and now they seem completely natural. Another, your broadcast of the week is from the great Jeanne Tyr. Your better half, Keith. What's Jeanne been talking about? Assessing venture capital. I assume that all your wisdom about this collapse of VC is taken from Jeanne, because she's a crunch base and she's smarter than you anyway. Yeah. Well, I think Kate Clark also gets a call out, but this podcast was done by Standard & Poor's. We all have heard of the S&P 500. That's Standard & Poor's. And they asked her to talk about what's going on in venture capital. And she does a great job. It's had a lot of on X, formerly known as Twitter. And she pointed it out to me only this morning, by the way, so I included it at the last minute. And confession, I actually haven't listened to it. Oh, my God. Did you tell Jeanne that? Well, she knows because she only just told me about it. You'll have to do that this afternoon. A couple of other pieces of news before we get to our startup of the week. Interesting piece on VW investing in China, 5% in Xpeng. Has the China crisis passed? I mean, the headlines in the New York Times today are Biden and the Chinese are talking to one another again. Have we got over that hysteria that was earlier this summer about TikTok and the Chinese and all the rest of it? It's quietened down quite a bit, hasn't it? And diplomacy... Which is good. Yeah. And diplomacy suggests that's the trend. It did. The economist announced this week that China is now the world's largest exporter of cars, which it wasn't up until now. But in the same copy of The Economist, they also wrote that it's going to take at least two more decades before China is the largest economy in the world, eclipsing the US. So there's balance and nuance there. Volkswagen taking a 5% stake. There's a couple of parts to that. The first is the Chinese electric vehicle industry is super innovative and the cars are very high end luxury cars. The Chinese buyers demand luxury. So they don't buy Teslas that much because the internal Tesla is too Spartan for them. So VW is doing... Well, I would, as a Tesla owner, in fact, we're a two Tesla family, I would describe it as minimalist rather than Spartan. It's minimalist. And a Chinese buyer wants soft, comfy chairs and they want luxury. The biggest English word used in China is VIP. And so there's this kind of desire for higher end. And so that was always a challenge to Tesla. One of my friends is Veronica Wu. She used to run Tesla in China. And she used to tell me that her hardest job was selling to high net worth individuals there because they wanted something that was more like a Mercedes or an Audi or a BMW level of luxury. And the fact that Tesla was so far ahead technologically in EVs wasn't enough. Well, now XPeng and NIO and Li and BYD, there's four huge ones in China, are all super competitive. And VW wants a piece of the Chinese market. It also wants to take technology from China into the VW and Audi franchise. They're the same company outside of China. I mean, that's doubly good news. Good news on any investments, real investments in EVs plus the global network is a good thing. And the pushback against economic nationalism or autarky finally, or second finally, our startup of the week key. I have to admit, not only have I not heard of this company, I don't even understand what this industry is. What is LK99, the DIY race to replicate LK99? So LK99 is a code word for a South Korean effort to create what's known as a superconductor that can operate in normal temperature. As if it was room temperature. So what is a superconductor? When electricity runs through wires, there's friction. Copper is usually the chosen wire to run through. Silver is a better conductor than copper, but copper is cheap enough to use. But it creates friction, which is why you get heat in systems that use copper, because friction creates heat. A superconductor is something that allows electricity to pass with no friction, literally zero friction. So there's no heat and no loss. And it's been, and by the way, it can pass through the air. It doesn't need a wire. It can pass through thin air. So electricity can be moved around in space. And this has been something everyone's been chasing for a long, long time. I remember there was a tech crunch disrupt company that claimed to have done it, and then it turned out they hadn't. So last week, two papers were published by the South Korean PhDs saying they'd figured it out. And I think it was two or three weeks ago. And now everyone's very preoccupied with firstly, doubting whether they really did do it. But if they did it, trying to replicate it, they published the papers so anyone can replicate it. If it's real, it means things like nuclear fusion, and close to free power becomes possible. And much cheaper, smaller components become possible, like transistors give off heat. If superconducting was real, it's as profound as the transistor in terms of