Will we reach AGI before Stripe goes public?

Here’s what Mary Ann and Alex got into: Stripe’s valuation recovers: As part of a tender offer, Stripe is now worth $65 billion. The company’s valuation has been on an up-and-down path in recent years as the company works to make its massive bulk fit into the private markets. Alex has thoughts on when it should go public, as you might imagine. Fervo Energy’s $200M+ round: What if the solution to our energy problems was not in the stars, but beneath our feet? And no, I am not talking about carbon-based fuels. No, instead, what if geothermal energy is what we’ve been waiting for? Fervo wants to make that question into a reality. Why VCs are investing in companies that shut down companies: With more startups than usual heading for closure, there’s a mountain of work ahead for founders and backers to shutter companies. Now some angels and other investors are putting their own capital into several companies that specialize in helping other firms close. Grim, but necessary. AI and the law: Microsoft’s move to invest in French AI company Mistral is not a bad way to spread its bets. Or to potentially fend of regulatory scrutiny that is building. AI in general is having a bit of a time sorting out rights — or lack thereof — that some training data may retain. Coming up this weekend we have an interview with Nubank’s CEO, and we have another great interview planned for the weekend after! Chat soon! For episode transcripts and more, head to Equity’s Simplecast website. Equity drops at 7 a.m. PT every Monday, Wednesday and Friday, so subscribe to us on Apple Podcasts, Overcast, Spotify and all the casts. TechCrunch also has a great show on crypto, a show that interviews founders and more! Credits: Equity is produced by Theresa Loconsolo with editing by Kell. Bryce Durbin is our Illustrator. We'd also like to thank the audience development team and Henry Pickavet, who manages TechCrunch audio products. Learn more about your ad choices. Visit megaphone.fm/adchoices

• 0:10Speaker 00

  Hello and welcome back to Equity, a podcast about the business of startups, where we unpack the numbers and the nuance behind the headlines. Today is March 1st. My name is Alex and I'm joined today by Senior TechCrunch Reporter on the Fintech Beat, it's Maryann Azaveto. Hey, Maryann, how are you?

• 0:27Speaker 01

  Alex, I'm good, but I think I need another one of those Nespressos that we were talking about before the show.

• 0:33Speaker 00

  Maryann and I, if you didn't know, walk around toting behind us an intravenous drip of coffee, just to keep our blood coffee levels where they need to be. Illegal in several states, but very good for podcasting. Jokes aside, Maryann, today is something of a special or important day for my history of TechCrunch, which is that I think the paywall came off of all TechCrunch Plus stuff at midnight last night. So if all goes well, according to plan, and you know, technical issues aside, we're recording this Thursday afternoon. So you know, it hasn't happened yet when I'm recording this, but it should be by the time you hear it. All of TechCrunch Plus stuff should be freely readable by everybody. So end of an era, but also an open door, I suppose.

• 1:14Speaker 01

  What a huge gift to our readers.

• 1:15Speaker 00

  Please read them, make me feel good. Anyway, on the show today, deals of the week, we have Stripe and Fervo Energy, one from the world of fintech, one from the world of ripping heat out of the earth. So a little bit of diversity on our deals. Then we are going to dive into Maryann's recent investigation into startups that have been built to shut down other startups. I have many questions about this. That's going to be a fun one. And then we're going to wrap up with an update on AI ownership and legality with a special note on Microsoft's recent investment into the French Mistral AI. But first, Maryann, Stripe is back in the news and once again, it's not going public. What's going on now?

• 1:52Speaker 01

  Right, right. So Stripe's valuation has been on a roller coaster ride over the past three years. So I reported, I think it was, is it just yesterday? That its valuation is now back up 30% to be exact, to $65 billion as the company Announced a tender offer, basically, meaning that it wanted to provide liquidity to current and former employees. It, Inc. deals with investors and it also too itself is buying some of these shares back. So the investors and Stripe together are purchasing over a billion dollars of current and former Stripe employee shares. So this is notable for a few reasons.

• 2:33Speaker 01

  One, just last year, the company was valued at $50 billion when it raised 6.5 billion. At the time, it raised that money for this very same purpose, to provide liquidity to employees and shareholders. This also means that it is most likely not going to go public this year.

• 2:50Speaker 00

  Yeah, I have a lot of thoughts about this. Okay, so I pulled some data before we jumped on and I want to walk people through how we got to today. Because you just did a great job encapsulating the latest stuff. But go back to 2020, Series G from Stripe was worth about six Hyundai at about a $36 billion valuation. Then 2021 comes, as we all know, things went crazy, especially fintech. The Series H was also $600 million, moving the company up to a $95 billion valuation. Not a triple, but you know, not a double either, a very impressive fundraise.

• 3:24Speaker 01

  And it made it, I think, probably the largest or one of the largest, well, highly valued private companies, that particular raise.

• 3:31Speaker 00

  I mean, also one of the largest in revenue terms. So I think you're right both ways, but yes, in valuation data, it stood out. I think maybe SpaceX is above it. This is roughly where Uber got to pre-IPO. It's kind of where you see private companies top out. I feel like, see if you were the Hyundai, you know, go public. The company didn't, and then we had Mary Anne's listed Series I from 2023. That was the $6.5 billion tender offer, I'm sorry, liquidity and tax event. That was a $50 billion valuation. And then now greater than one billion, tender offer, $65 billion valuation. 65 billion, Mary Ann. Does that feel high to you?

• 4:08Speaker 01

  You know, it's really hard to say, Alex. I mean, our own Rebecca reported earlier this year that based on secondary trading activity, the company was valued more like around $53 billion. So I did feel a little bit high based on that. What's interesting about Stripe is it, I mean, it's, it's obviously huge. They have really big customers and they've named some of them, including Microsoft, Uber, Best Buy. But at the same time, and this is a little gossip we heard from upfront, there's a lot of complaints about the company, apparently, a lot of chatter that people complain about things like it's really hard to disengage, fees are very, or cannot say the word.

• 4:47Speaker 00

  Onerous.

• 4:47Speaker 01

  Onerous. There's a lack of transparency into fees. So I feel like there is some frustration with Stripe and there are competitors out there. We've covered a few of them, including Finix, for example. So anyway, but it's still forging on. People still really care about it. And I think we all still can't wait to see if and when it goes public.

• 5:07Speaker 00

  Yeah. Dan Primack over at Axios, who's actually been on this show way back in the day, he wrote about it today and he said, or I guess it was this week, I forget if it was Thursday or Friday that this newsletter came out, but he said, you know, what if Stripe just doesn't go out? What if they just don't go public? Mostly you have to if you raise external capital. It's the game, right, Mary Ann? You raise money, grow fast, and then exit either via an IPO or an acquisition or whatever. But Stripe, you know, it's so big and so valuable and so able to kind of like control its own financial destiny, as we're seeing through these last two transactions, that maybe it doesn't and just pays dividends, I guess. It's just to not go out and have a traditional liquidity event means that it either has to repurchase all of its investor stock, which they won't want to give up for

• 5:51Speaker 00

  less than a fat premium, or it has to pay dividends, which means it has to be cash flow positive at an absolutely massive basis. But venture-backed profit companies don't do those two things. And so I'm perplexed by this. And I'm disappointed that everyone is saying that this new offer, this new tender offer is going to delay the IPO again. I mean, what are they waiting for?

• 6:09Speaker 01

  You know, Alex, it's a good question. I would be really surprised if they don't eventually go public. I think even just for public perception sake, I don't know, maybe they're waiting for the market to shift a bit. I can't really say. I would be really surprised if it doesn't happen at least sometime in the next year, year and a half or two.

• 6:28Speaker 00

  That's a 12, 18 or 24 month time. It may happen some point in time in the future. And no, but I only beef because I think you're being reasonable in your point there. I think it could be several years if they decide to go out, which is just bonkers. When did going public become the thing that everyone doesn't want to do? It was the moment of like proof of maturity and success and we built something big. And now we're going to be more public and you know, I just, I find it so strange that we're, that we're seeing this Zerp philosophy persist into a more traditional economic environment. And it feels like a disjointed approach to the modern market because this is not 2021 anymore. And you can't keep pretending that there's always going to be unlimited capital for private companies.

• 7:13Speaker 00

  It seems to me to be unnecessarily risky and also there's no advantage to it.

• 7:17Speaker 01

  I think it's interesting also that I mean, Plaid is on its way to going public. And you know, the, those two have become increasingly competitive as well, right? Kind of getting into each other's areas more and more. So you almost wonder if they're like, kind of playing a waiting game to see who goes public first and to see how the other does. I mean, not that of course their actions are not that dependent on what their competitors are doing necessarily. But I mean, they're both on the path to going public. Their markets or what they do overlaps. We've got Klarna in the wings also probably going public soon. Not necessarily in the same space, but fintech. So I don't know, I feel like we're just like sitting here. This is me and Alex on the edge of a cliff, biting our nails, just waiting, waiting to see what happens with all these fintechs and whether or not they'll go public and when.

• 8:03Speaker 00

  Okay. So you brought up some great stuff there that I have to touch on. So first of all, if we do see Plaid go out before Stripe, I'm going to buy Zach Parrott like a coffee because that would be hilarious to me.

• 8:16Speaker 01

  It could happen.

• 8:17Speaker 00

  Oh, it could easily happen, but I think it'd be more an indictment of Stripe than an endorsement of Plaid, which is not to take away from Plaid's success. I mean, to be clear, they've done quite well, but I mean, Lord above. To me, Stripe just seems mismanaged at this point versus being, you know, like savvy in its approach to, to not going public. Klarna though, kind of a wild card. Have you read their latest numbers? They just dropped their 2023 annual report.

• 8:42Speaker 01

  Yeah, they were pretty good still.

• 8:44Speaker 00

  Yeah, you know, I mean, Klarna is doing fine. Improving its profitability, growing faster than its core market, you know, using AI in a lot of neat ways. I guess what this says to me is there's a lot of health, it appears, in the fintech market. And sure, valuations aren't where they used to be, but that doesn't mean these aren't great companies. Like, no one, no one thinks Stripe is hiding a bunch of dead rats in its, you know, income statement. I don't think at least.

• 9:08Speaker 01

  Not at all.

• 9:08Speaker 00

  Yeah, so maybe, maybe Plaid will be the first one bold enough to actually go out and do this, but I, I, I hope they do. All right, anyways, let's move on and go from the world of high technology and finance straight into the bowels of the earth where we hunt out heat. Because, Mary Ann, if you want to do geothermal energy, you have to dig. And that brings us to Fervo Energy, about which I am fervently energetic because it just raised more than $200 million to commercialize and expand on geothermal energy. And Mary Ann, I have to say, it turns out they're doing some work down, I believe, in Texas.

• 9:41Speaker 01

  Yeah, they're based in Houston, so this was really interesting to me. I will not pretend that I really get the technology itself because it's very complex. Although Tim, the reporter who covered it, did a great job of kind of breaking it down. But it was fascinating for a few reasons. Just the technology itself, what it can do. They use directional drilling techniques that were pioneered by the oil and gas industry to extend wells that they make, quote, "far beyond their surface footprints." $221 million is huge. It's a lot of money, especially now. It's based in Houston, which I also find interesting. And that's because a lot of the techniques they're using, a lot of the talent they're using is out of the oil and gas industry, which is predominantly in the Houston area.

• 10:26Speaker 01

  Now, I lived in Houston for 14 years. The city has kind of a bad rep, I think. It's actually much more cool, much more diverse than one would expect. Restaurants are amazing. But anyway, it's kind of neat to see a company out of Houston raising this kind of money. It makes sense when you understand what they're doing and how they're doing it. And another thing that I found interesting, early angel investors include Jeff Bezos, Richard Branson, Bill Gates, and Masayoshi Son. That's kind of a high-profile list of early angels.

• 10:59Speaker 00

  Absolutely. And what do Jeffy, Richie, Billy, and Masa have in common? Well, two things. One, lots of money, and also a general bias towards things that might make humanity slightly better off. The Gates Foundation is the kind of er example of that, I think, right now amongst tech philanthropy. But if we can get geothermal energy to work at scale, and has a couple of major advantages. Mary Ann, you touched on the location of it and the talent. We already have a lot of people that can do horizontal drilling from centralized wells, thanks to the fracking industry. And we also know that the planet's very hot. And so if you can get down there, you can create a temperature differential that will yield steam, which is what basically everything does in the world of power generation, dispense turbines and then create electricity. Hydroelectric is an exception to that, but that's slightly

• 11:43Speaker 00

  different. So I asked Tim how this works, because much like yourself, Mary Ann, I'm not a scientist. And it turns out that it gets really hot, much closer to the surface than I thought. And all you need to make steam is 100 degrees Celsius. And so some of these wells are getting up to like 200. So plenty of energy down there that you can use to essentially create steam, run these turbines. I love this. I don't know exactly how it works. That part's lost on me. But what this yields is something that can do baseload power. And if you're asking yourself, Alex, what's baseload power? That's what I asked Tim. And Tim said, it means that it is a constant source of energy. Wind power, solar power, both awesome, but episodic when you have wind or when you have sun. So you

• 12:27Speaker 00

  don't need batteries to make geothermal energy work, which means that it's just more consistent and therefore cheaper probably to operate. And we're just taking some of the Earth's heat, you know, like a little mosquito sucking out the energy. And I think this is just super duper cool.

• 12:41Speaker 01

  It's fascinating. It's really fascinating, especially to me, because you're taking these, you know, fracking is kind of a controversial practice. You're taking these techniques and actually turning it into something that can be very, very positive, right, for the Earth, the environment. I love that. And as the oil and gas industry sort of shifts, you know, giving workers another place they can go. There's a lot of, I mean, there's just so many amazing, cool things about this.

• 13:05Speaker 00

  Mary Ann, why, why is fracking controversial?

• 13:09Speaker 01

  I don't think we have time to get into that right now, but.

• 13:11Speaker 00

  We can list a couple of things that happens. For example, waves of microquakes.

• 13:16Speaker 01

  Waves.

• 13:16Speaker 00

  Right? Like it can lead to lots more localized seismic activity, for example. So fracking is not a solution to all of our problems, but in this case, it is the application of an oil and gas technique to something that is much better. I don't think we're ever going to have a technology power solution that is perfect, but I do think that this is one that does have a lot of potential that is quite exciting. And it does also give me a chance to bring up my new pet topic, which is a video game named Dyson Sphere Program. I love kind of like city management, factory automation games because they tickle my brain. And I find it really fun to kind of think about this sort of company after having my head stuck in that world because it's all about power generation and resource management and all that stuff.

• 13:58Speaker 00

  And there's also geothermal power in that world. And so when I saw it come up, I'm like, we have to bring up Fervo because I want to say the words Dyson Sphere Program on the podcast at least twice. And now I've got it.

• 14:08Speaker 01

  You did it. Yay.

• 14:10Speaker 00

  All right. When we come back from a very short break, we are going to talk about how VCs are investing and taking apart their own investments. It'll be a lot of fun. We're right back after this.

• 14:23Speaker 01

  So this week I wrote about this phenomenon that seems to be happening in the startup world. We had two companies this week that raised in this space. I have to share the backstory about this really quickly. Okay. So about a month ago, I was like, you know, I've noticed there's more startups that have been created to help other startups shut down. Let me reach out to them. I'd like to write a feature on this topic. Coincidentally, both those startups were like, oh, glad you touched base. We just raised some money. So I was like, perfect timing. So I decided to, to write this feature and the two startups that raised money in this space are Sunset and Simple Closure. Okay. So Sunset raised 1.45 million dollars in

• 15:04Speaker 01

  seed funding, mostly from angel investors. I'll get into that more later. Simple Closure raised another 4 million. This was less than six months after it raised one and a half million in pre-seed money, which I did cover back in the day, which was only like six months ago. And then also, Carta announced a similar offering in February called Carta Conclusions. So to summarize, basically, what they're all trying to do is help these startups, which, by the way, the statistics show that about 90% of startups do end up failing.

• 15:38Speaker 00

  Right.

• 15:39Speaker 01

  That's a huge, total addressable market, right? I mean, you know, you've got a lot of potential customers out there. And what they're doing is trying to make it easier, faster, just more streamlined, cheaper, supposedly, for these companies to wind down operations. And that includes things like helping them dispose of their assets, their IP, tax questions, all of it.

• 16:01Speaker 00

  So, I really like this, but I want to start with the names. So Sunset, that's what you do when you kill off a product, it gets sunsetted. Simple Closure, though, sounds more like a therapy technique for people who just got dumped. But, okay, I'll allow it. And then Carta Conclusions, that just sounds like a poorly named, like, data blog. Like, here are Carta's conclusions versus Carta taking you to the point of conclusion. I, euphemistic names drive me bonkers, but I guess these are all relatively twee and therefore acceptable by the marketplace. First question, Mary Ann, 1.45 million for Sunset, not a lot of money. Is this more of a services business than a software business?

• 16:41Speaker 01

  Well, that's a good question. I feel like it's a little bit of both. It's not a ton of money, but I think they're just, they're trying to be careful. That's a good point that you raised because one of the co-founders that I talked to, Brendan Mahoney, told me that after talking to a lot of these startups, part of their issue, the reason why they were winding down is that even though they had good businesses, they had decent ARR, they had raised venture capital. And even though they were doing well, they weren't building what you would call, you know, a venture scale business. And so they were like, okay, we're just gonna cut our losses and move on.

• 17:16Speaker 00

  So why does that happen? If you're listening to this and you are less deep in the world of venture, let me try to help here a little bit. Venture capital invests in a company comes with expectations. You're essentially on a timer or a treadmill, if you will. VCs invest capital ahead of your ability to generate revenue and gross profit to fund your operations. So you can hire more, build more, invest more in sales and grow more quickly. The trade off is you take on capital, you're expected to grow like mad. And a VC will evaluate business that is at $5 million in annual recurring revenue, and that's growing at 1% per year at $0.

• 17:57Speaker 00

  The same VC may value a company at 5 million ARR, growing at 100% each year at a gajillion dollars, because growth is what startups do. So if you end up flatlining at a single digit ARR number, you can effectively lose all your venture momentum. And then you have a cost basis that you can't support. And then you essentially have to decide what to do with it because you're not going to raise more money for a non-growth business. So revenue is an important thing to keep track of, but for startups, growth really is the name of the game. And I'm not shocked ever seeing Mary Ann and companies with ARR hit a wall because right now it's harder to raise money.

• 18:31Speaker 01

  It's harder to raise money. And so that's what these co-founders realized. And they said, you know what? We don't, we don't want to be caught in that. We don't want to raise a lot of capital and then ourselves be in the same boat as many of our customers. So they chose to go through this, this non-traditional way of fundraising. Most of their investment dollars came from angel investors, including hustle funds Eric Bond, weekend funds Ryan Hoover, layoffs dot FYI creator Roger Lee, who is also a founder himself. And so another interesting thing about this is they're offering their investors an incentive. This is something I wanted to delve into, but kind of ran out of time and may revisit. But basically they're saying, hey, if you refer us customers, your stake goes up.

• 19:14Speaker 00

  Now, Mary Ann, I gotta say, watching the guy who does the layoffs website invest in the company that helps other companies die feels a little bit like the newspaper reporter in charge of local obituaries putting capital into the funeral parlor, but we will have a chance to talk to him about this very soon.

• 19:30Speaker 01

  Yeah, I'll be interviewing him next week for an equity episode that will air a week from this Saturday. So definitely keep your ears and eyes open for that.

• 19:40Speaker 00

  Yes. And if you don't know what we're talking about, just go to layoffs.fyi. It's a fantastic website. I've used their data time and time again since it launched to help understand what's going on out there in the world of tech layoffs.

• 19:51Speaker 01

  Also, a simple closure side, Infinity Ventures led their raise, which also included, quote, "strong participation" from Anthem's group, Fox Capital, other existing backers. They claim that they've seen a ton of growth, that they've crossed seven figures in annualized revenue already.

• 20:07Speaker 00

  Yeah. Okay, so services or software is an important question to ask because in the world of high growth startups, if you're doing services, you'll never have great gross margins and you only scale with headcount, which is inefficient. So everyone wants to build software, very high gross margins, easy to scale, et cetera. These businesses, though, are a little bit more hands-on. It feels, Mary Ann, because when I read your story, we're talking about IP, we're talking about incorporation, a lot of things that could have edge cases, nuance to them, legal ramifications. And so it probably is not that easy. So the annualized number versus an annual recurring revenue figure makes a lot of sense to me. And I'm not shocked that they're seeing that much demand. I just wonder, how hard is it to go from single-digit millions in annualized revenue for these companies to double-digit millions in annualized revenue?

• 20:53Speaker 01

  One other thing I do want to point out, while this is kind of new in the startup world in terms of companies doing this, it's not new at all. I mean, obviously, businesses have been shutting down for years and years. I did talk to Marty Pinchinson from Sherwood Partners, who's been doing this for like 25 years since the dot com bust. So I just wanted to point that out. So, you know, even though startups are kind of now really getting into it, there have been other institutions doing it for some time now. But I'm curious to see how this plays out, especially as markets change. I don't know. It's kind of fascinating to me. I'm going to definitely be paying attention.

• 21:28Speaker 00

  Yeah, I remember, I had this like memory from the back of my head. I'm on Caltrain in the Bay Area. I'm probably like... Seventeen, eighteen, and I'm reading Red Herring in print.

• 21:42Speaker 01

  Yes, yes.

• 21:43Speaker 00

  And I remember reading about a company that did this. And they were talking about what do you do with 500 office chairs and printers and stuff? Clearly ended a different era of the tech, you know, industry. But it just goes to show to your point that shutting down companies that have a non-traditional asset base, IP, some hardware, et cetera, is a challenge and one that is unique. It is a little sad to see a startup cluster emerge around helping startups die. But it's better to have tools and services to do this than to not, because then it's just a mess. So.

• 22:19Speaker 01

  Yeah, I think I'd like to end this on a quote from Infinity Ventures co-founder Jeremy Jonker, who's firm led that financing into simple closure. He said, "We are seeing a meaningful increase in startups facing challenges. Historically, these startups raised at robust valuations and have not grown into the metrics necessary to raise an up round. As such, they are facing the question of raising at a meaningful down round, selling the company and/or potentially shutting down the business." Therefore, he said, "He thinks the addressable opportunity," for his investment, of course, "is sizable and increasing every day." Wow. So I don't know if that's good or bad, but yes, definitely truth to that, I think.

• 22:58Speaker 00

  I have many more analogies, but I'm not sure which ones would get us in trouble. So I'm going to go ahead and not make any more to avoid that and instead move us on to AI, the legality thereof, and this recent Microsoft investment into Mistral AI. Mary Ann, should I give people an overview or just dive right in?

• 23:14Speaker 01

  Maybe a very quick one, Alex.

• 23:16Speaker 00

  Okay, so OpenAI comes out and goes, "Woohoo, look at this, look at what we built." Everyone goes, "Oh shit, we need to catch up in the world of AI. We can't have all the foundation model companies be American. Enter France." And Mistral AI, which quickly raised a lot of money and has put together some very impressive large language models that I keep seeing people talk about on Twitter and apparently are seeing wide adoption amongst certain startups. So Mistral is kind of like the European Union's shot at having a leading company in the current moment of AI. Now, will it succeed? I don't know. Certainly has raised less capital than certain companies, your Anthropic's, your OpenAI's, but it does have a good foothold. And then Microsoft decided to put 15 million

• 23:58Speaker 00

  euro into the business. But Mary Anne, Microsoft owns 49% of OpenAI's commercial division, which led to you asking a question for us in our prep call.

• 24:08Speaker 01

  Yeah, I was kind of confused. I was like, wait a minute. If Microsoft is such a big shareholder in OpenAI, is this not a conflict of interest? Like, I mean, is it because it's a different continent that it feels like it's not? Or what? Because I don't know, it just, it feels a little, little weird to me.

• 24:26Speaker 00

  Yes, it does seem a little weird. But is it actually a point of conflict or merely a conflict in point? And I think the issue here is Microsoft wants to spread its bets around just because that's not a bad way to hedge. For example, what if OpenAI suddenly slows down? And Mistral makes better programs. Well, then it has a small piece of it. But I don't think Microsoft really expects OpenAI to not be one of the leading companies in the foundation model business in the next 10 years. So why this deal? And why does OpenAI probably not care that they've done this? I think it's to cover their butt legally. We have seen governments around the world start to say, you know, this Microsoft OpenAI deal seems a little odd, perhaps too cute by half, if you want to use the English idiom. And by that, I mean, Microsoft has managed to

• 25:10Speaker 00

  find a way to get a lot of visibility and ownership of OpenAI's work, but not bring it into its actual auspices, and therefore avoid antitrust scrutiny and so forth.

• 25:22Speaker 01

  Very interesting theory I hadn't thought of. I mean, I'll just say one more thing on this topic, because I know we need to move on to other, other things related to OpenAI. But I had a little bit of a flashback to Sequoia backing out of its investment in Finex, because it was a big investor in Stripe. So I know Microsoft is not a venture capital firm, per se. But anyway, it's just really interesting to me how there are different ways of these sort of things being handled, depending on the players and the industry.

• 25:53Speaker 00

  Another theory, because now that you've raised that up, it's actually a very salient example. The thing that I would throw in is, what's the chance that OpenAI buys Mistral?

• 26:03Speaker 01

  Hmm, yeah.

• 26:04Speaker 00

  Microsoft would then on both sides of that particular coin. So heads or tails, they win, right? I don't know. Microsoft is savvy and they have been moving really fast to get AI integrated into their products. I just saw them release a new co-pilot for finance this week or something. So they're, they're busy with it, but there is some kind of clouds on the horizon, if you will. And you may recall that the New York Times sued OpenAI essentially for ingesting its material as part of at least through GPT-2's model creation, Maryanne. And there's been a little bit of pushback there.

• 26:37Speaker 01

  Yeah, I mean, OpenAI is defending itself and they're saying that the publication hacked into its product and basically repeated something like thousands of times to try to get the result it wanted to claim that OpenAI was using its material like verbatim. So there's a lot of back and forth. I have like mixed feelings about this because as a writer, if OpenAI takes my articles and then is just spouting it off verbatim, I don't know how I feel about that. And do they credit the publications from which they get this material? I don't think so. Is that not plagiarism? I don't know. There's a lot of blurriness here.

• 27:16Speaker 00

  Edge case though, right? Like, I don't think that most of the time you type in, you know, tell me why Sequoia backed out of Phoenix and it's not going to say, well, you know, here's Maryanne's story. I'm not going to credit her. I do think that's possible, but I think OpenAI has made a reasonable case that that is sufficiently fringe as to be more indicative of the data source versus then the data's use. That's been my read of it. But I don't know, I spent a lot of time writing stuff and caring about it and doing the work. And it does seem odd to me that it can be taken, ingested, smoothied, and then re-served on the other end for profit without my work being compensated. Because

• 28:01Speaker 00

  if we hadn't done this, then they wouldn't exist.

• 28:05Speaker 01

  Yeah, that's, I mean, really, that's what I mean. I have mixed feelings about it. So they can claim all these things about, oh, it's public, it's all this and it's all that, but that doesn't make it right.

• 28:14Speaker 00

  No. Also, the, the value exchange that we have in the world of search, I know some people are thinking, well, what about search? Yeah, but spidering and linking is different than ingesting wholesale and regurgitating in part or whole.

• 28:29Speaker 01

  Right.

• 28:31Speaker 00

  And I think that the AI companies kind of agree with this because they're signing deals with with different companies. Like, for example, Reddit's data is often used in certain data sets that are fed into LLMs and they are making several hundred million dollars a year now, according to their IPO filing, for that work. You know, I've written thousands of things for TechCrunch. And, you know, I don't own any of them. I guess, I don't know. The money behind private equity funds that invested into Apollo, that own Yahoo, that own TechCrunch, that own me own it. But, you know, if I'm generating value outside the business, I think the business should get paid for it. It feels weird to be saying that that's controversial, because a lot of the folks who are saying that this stuff

• 29:12Speaker 00

  should be ingested for free because it's important, are also the ones who are capitalists in their own right. They're just, in this case, saying that it's more important for the tech to move faster than for the capital to flow where it should, which I think is self-serving. I'm not shocked. It's business, but it does seem slightly gross to me.

• 29:29Speaker 01

  It does.

• 29:30Speaker 00

  Summing all that up, it's gonna be fun to see where the courts go on this.

• 29:33Speaker 01

  Yeah, especially because it becomes really complicated. Because how, I mean, I don't even see how they could, if they wanted to compensate the companies or publications or whatever. I mean, you know, that would become extremely complicated, probably slow them down greatly. And I just don't see how that could happen realistically.

• 29:50Speaker 00

  The American model in technology and governance is to let business run ahead and do lots of things. And then we come with a dustpan and a broom afterwards and try to tidy it up.

• 30:02Speaker 01

  Crypto.

• 30:03Speaker 00

  Okay, yes, super valid, although also unpopular. They're amongst the same venture capital set. But in this case, we have let OpenAI build a multi-billion dollar a year business off the back of stuff they didn't make. And I think there should be a way to get compensated for that, because the companies that are backing them are famously litigious, right? Do you think Microsoft would give you the benefit of the doubt so you could build tech off of its data?

• 30:25Speaker 01

  No.

• 30:25Speaker 00

  No. So I would just like some consistency from the tech players here. You can't, you can't be worth $3 trillion and dash around money in the tens of billions and then tell me when you use my shit that it's, it's hard. It's capitalism for me and not for thee, is how it feels, which I just think is ridiculous. I'm a capitalist. I made a thing. You're using it. Let's have the value flow both ways.

• 30:49Speaker 01

  That would be the right thing to do.

• 30:50Speaker 00

  Last thing, Mary Ann, just on the startup front, I will say that Mistral continues to win, which is good for the EU, which is good for the local startup scenes in France. And France does seem to be, I think, right now, the most interesting startup market in Europe, taking out Germany, I think, as lead in that role.

• 31:07Speaker 01

  Very interesting. Yeah, it announced earlier this week it was building a new large language model that could really rival ChatGPT-4.

• 31:15Speaker 00

  Yeah, which is impressive because GPT-4 is kind of still, I think, state of the art. Although I would love to see GPT-5, you know, sooner the better. Here's a fun game. Will we get to AGI before Stripe goes public? We may never know because we might die before both. All right, that is our show for this fine Friday. Equity, of course, is back on Monday, Wednesday and Friday of next week. And we do have an interview coming out this weekend. Mary Ann, tell us what's coming up.

• 31:41Speaker 01

  I talked to Nubank CEO David Velez. The company just recently crossed, I think it was a $50 billion market cap. And yeah, we had an interesting conversation about that and just LatAm in general. So definitely don't miss it. He's a great interview.

• 31:57Speaker 00

  Yeah, he really is. And of course, we have moved our Wednesday show to more startup focused topics. It's a deep dive on funding rounds and that sort of thing. And so we've scooted our interviews into the weekend to ensure that we can stay as focused as we can on our core topics while bringing in other voices. All right, that's all the time we have for today. We are Equity Pod over on X and Threads. We are TechCrunch Pods on TikTok. The TechCrunch Podcast Network has two sister shows to this one. Chain Reaction with Jackie Melanick on all things crypto, and then Found with Becca Scootac and Dominic Midori Davis talking to founders about how they built what they did. We'll see you all at a later stage. Goodbye.

• 32:31Speaker 01

  Bye.

• 32:34Speaker 00

  Equity is hosted by myself, Alex Wilhelm, and TechCrunch Senior Reporter, Mary Ann Azvedo. We are produced by Teresa Locan Solo with editing by Kel. Bryce Durbin is our illustrator. And a big thank you to the audience development team and Henry Pickavet, who manages TechCrunch Audio Products. Thank you so much for listening and we'll talk to you next time.