Netflix buys WB + why Jason should run Disney | E2219

24:17Like, can they take over a TV, major TV news network? They should have been able to. I said this on a previous episode, man, them having a 24-hour news channel globally on Netflix would have been... It would have been a game changer, but I think international regulatory approval of that kind of a deal was going to be like probably a non-starter for them. On that note, though, Lon, the EU is expected to pass this through. Everything that I read said that the EU is going to be fine. So it's going to be interesting to see if it's the American regulators that actually are the sticking point here. I don't think so.

24:42I think they're going to let it fly. I'll be honest. We're going into the golden age of M &A. We've talked about it here. We're seeing a lot of ticky tacky singles and doubles and acquisitions. I told you I'm predicting it now. There's going to be a$10 billion to$50 billion acquisition in 2026. And by 2027, there'll be a$100 billion company gets acquired. My thought though is that in the wake of the Meta FTC deal that we said was going to be a starting gun because it said that meta didn't have a monopoly in personal social networking because of competition and changing technology. It's hard to prove.

25:09I think that applies directly here, Jason. How can you say that Netflix is going to be anti-competitive here when they still have Disney Plus on the side and they're up against YouTube and they're up against... Peacock. I just signed up for Peacock. I had no choice. There are certainly in the US market, there are certainly other players and remaining options. And also I think the fact that CNN is not part of this deal and is going to get spun off and could still be acquired by a more Trump friendly voice means there really isn't as big a win in shutting this one down. I feel like there's no clear win for the Trump FCC in putting a stop to this.

25:44It's not like they're letting a news network go into the hands of hated enemies or anything. I don't feel like it may raise that much scrutiny as people are thinking. I wager, wager to think that Polymarket is in the news. They are indeed in the news. And they're in the news because someone is particularly good at making bets over on Polymarket. There's one user, went by the name Alpha Raccoon, now rebranded to 0AXFEE, that made about a million dollars, Jason, in 24 hours, making very granular bets on Google happenings. Now, people are calling this person an insider. Forbes is covering it. It's viral all over Reddit.

26:26And I got to say, it looks pretty damn suspicious. I presume you've already, you've seen this in the news. I mean, I saw the headline. I'll be honest. I've been pretty busy here. I'm on the road. I'm in San Francisco for the All-In Holiday Spectacular on Saturday night and my demo day, which I'm going to go right to after this, Friday afternoon. So I have not been following the details of the story. But I do know that Alpha Raccoon has been betting on Google and some of their markets, correct? Am I correct? Yes. So there's been a couple of major things that they've done. First off, they made a series of yes and no bets about when Gemini 3, Google's recently released AI model, would come out.

27:02They correctly predicted the day and bet against all the other days. This led them to making a$424 ,000 gain, betting that Gemini would not be released by November 15th. They bet no that it wouldn't come out by October 31st,$114 ,000. And they bet yes that it would come out on November 18th, netting them$91 ,000. But that's just the start. They also made a series of yes and no bets on will certain people be in the Google top five most searched people. They bet yes for Pope Leo,$174K profit. They bet no on Zoran Mandami, the mayor-elect of New York City, plus$45K. They bet no, Donald Trump, plus$171K.

27:38And they bet yes on Jimmy Kimmel, plus$15K. And this was all in a pretty short time window. And when you're 22 out of 23, as Forbes reported, it looks pretty gosh darn suspicious. I think it's the polite way to say that. Let's be clear here. Prediction markets are distinctly different than trading public stocks or betting on sports. There are going to be different sets of rules here. The reason prediction markets are unique is that there is no concept of insider trading in prediction markets. There is insider trading. If you worked for Google and you made bets on the stock price of Google based on the release of Gemini 3.0, that would get you in big trouble.

28:23And by big trouble, I mean gel. And insider trading, Martha Stewart, plenty of examples. And that one was probably a little ticky tacky. And they went after her, obviously, as a celebrity. which is a bit unfair, but everybody has to face the music, I guess. I don't know the details of that one, but it did seem a little bit harsh for a tiny gain. I think they probably shouldn't need to go to jail. They could have just given her a big fine and made her do a commercial. But anyway, putting it aside, these prediction markets are different. And if you're on the Warriors or the Knicks, as we just saw with the Justice Department going after people who might have been placing what are called prop bets over unders on the number of rebounds, that also is illegal.

29:03So I think this is where the market is going to learn what's allowed and what's not. So where does this one sit and how are people framing it in the press? Are the prediction markets giving any guidance here or, you know, Polymarket, Calci, other prediction markets, obviously, is going to be a whole bunch of these with the success of Polymarket, which I have a partnership with and I'm a shareholder in, full disclosure. You can't have an effective sales team without an effective and above all organized sales process. Trust me on this one. I've seen thousands of startups basically fail because they had messy, chaotic systems in the sales department.

29:41It's one of the biggest red flags there is. This Week in Startups, it's about helping founders. So I want to tell you about the tool we use to stay organized here on This Week in Startups. It's called Pipe Drive. They bring your entire sales process together in one clear, easy to understand and centralized space, giving you just one place to check for all the customers and deal information you need. I'm a big fan of data, right? Especially here at This Week in Startups. So teams that use Pipedrive close an average of three times more deals every month. It could be you. Join over 100 ,000 companies using Pipedrive.

30:13And when you use our link, you're going to get a 30-day free trial. No credit card or payment is needed. Just go to pipedrive.com slash twist to get started. That's pipedrive.com slash twist.

30:48prediction market, does that action change the market enough to constitute manipulation? The consensus appears to be no, which means that enforcement here is going to come down to individual companies trying to protect their own information and not letting their employees trade on it. Because once someone finds out, Jason, that an insider from, say, Google is making a bet about a product release, then they could take that information as a third party and trade on it themselves or change their own release schedule. I'm sure OpenAI is going to pay a lot lot of attention to this person when they bet on future Gemini release schedules because it's going to tell them exactly what's going on.

31:23So just as I think sports teams and sports leagues want to ensure that their players are not corrupting the league's performance, I think that companies are not going to have to be super clear about what their employees are allowed to do because they don't want to accidentally leak secrets through this loophole. And it may be that this particular user is not a Google employee. People seem very convinced that they are. But when they do get found out by Google, if they are an employee, they're going to get canned because one, they're making Google look really gross. And two, they're effectively telling the market ahead of time when they're going to drop market defining products that Google wanted to keep under.

31:59Yeah, there are rules about this inside of companies already. Apple is very famous, as well as Tesla and anybody who releases products on some cadence of explaining this explicitly to employees. And also they all have security departments. So the security at Apple, security of executives, I'm talking about security of corporate security of secrets. This is like one of the most locked down organizations. And when the person lost the iPhone 4 and Gawker bought, not Gawker, Gizmodo bought it and committed essentially what I believe is a felony to buy stolen stuff. I don't know where that ever hashed out, but I remember the Engadget team coming to me and we're like, hey, we got offered this too.

32:37And I was like, we're not going to commit felonies. Thanks, guys. But we'll take a pass. I mean, I'm being serious here. I remember them saying, like, can we buy stolen secrets? And I was like, absolutely not. That's a felony. Like, can't commit corporate espionage. Now, Nick Denton had a different position on it, which was, let's see what happens, I believe, was his interpretation. How did Gawker fare long term there, Jason? Yeah, I think, yeah, you got to be thoughtful about these things because they can be what we call in the gambling business, the risk of ruin. You don't want to ever get to that point.

33:07But employees need to know if you putting aside the prediction markets, putting aside the CFTC, your employer, you've already signed documents that you're not going to share information. In a way, this is sharing information and you're not allowed to use this information. So this is just going to be in everybody's employee contract. They're going to explicitly put in corporate documents. You are not allowed to participate in any prediction markets related to our work or the information you're exposed to. This will be in documents very soon. And you're not allowed to share it either, because this is going to be a bummer in some ways, because what makes the prediction market so powerful is that people participate in them.

33:41This is the thing, right? This is the idea that the sharps, the people that have more information than the average person can essentially tell the world what's going to happen. But if you take that out, Jason, and I don't mean to talk ill since artists are doing very well, but I wonder what the value of the predictions will be if you take out insider information, because then it's just reaction to market news. You know what people have? Friends. So what's going to happen? Friends, cousins. friends of friends, here's what typically happens is a person on a sports team is like talking to their friends in group chat and they're like, oh yeah, don't bother coming to the game tonight.

34:14LeBron's injured and the injury report hasn't come out yet. And they're like, yeah, no, don't come to another game. Don't buy tickets to this one. I mean, LeBron's not even playing. Then they're talking about like, oh yeah, LeBron's not playing. And then a friend of a friend places the bet that LeBron is not going to hit their 32 points and they win or like the Lakers aren't going to win. And all of a sudden, you're in the money. So that's kind of the danger. Those kind of things will happen. People talk, people chit chat, et cetera. If you're a journalist, you have great inside information. And like you're one of those like old school journalists as a network, I suggest you quit your job and you just go full time Polymarket.

34:46And there was a guy that was trending. I don't know if you saw this, but there's somebody who's just doing very small bets and making a fortune on Polymarket. Did you see this? I did not see the small bet guy. Yeah, there's a story that's been trending or buzzing around about somebody who's pretty sharp. I'm surprised we didn't pick him up as our interesting person of the day. But there are a number of people who are full-time on these prediction markets. And they were. When there was in-trade, there was a company called in-trade in the 2000s, early 2000s. People became addicted to doing that.

35:14And I think it was in London, but it got shut down or it closed. But there are people who are going to do this full-time. I think there's also a startup opportunity here. All jokes aside, if you can figure out who the traders are on Polymarket that do events and information, you can aggregate that and sell it to hedge funds. You can make a lot of money. I mean, and all this stuff I believe is on the blockchain for these companies. So that's why when this trader in question, Alpha Raccoon or whatever changed their name, we knew exactly who they were. And then I, because it's all in the blockchain, you can't scrub that out, which is the power of, you know, decentralizing and so forth.

35:43But it's interesting. All right, let's move on. Let's talk about a couple of companies from the recent Y Combinator, fall 2025, class. A couple of stats for folks out there who are curious. Where is YC getting companies from these days? Well, out of 156 launched companies, 130 of them are from the US and Canada. 22 are remote, Jason. Four from Europe and one from Southeast Asia. And that's it. So essentially, it's San Francisco in the cloud or nowhere. I think you're not shocked by that. Yeah, exactly. Okay. Yeah, I mean, it makes sense. What I'll say is, you know, these are finishing schools.

36:14YC is the most notable of them. I'm not necessarily saying I say the best because I think there's five of them now that are competing to be the best. We're one of them, Watch Accelerator. A16Z has a speed run. PerVC, Marr, and Pejmon. Arc from Sequoia. Antler has a great reputation. Techstars. Dave Cohen's back running it and rebuilding it. And I saw him when I was in Tokyo. They're doing a great job. You should consider applying there. So there's five or six that I would say are the best. And I think they're probably all equally good if you are a founder, not in Silicon Valley, first-time founder, or a second time founder and your first project, you know, never became like venture backed or had, you know, serious backing.

36:55And what these programs do at their core is their sorting function. For every hundred people that apply, one get in, where in our case, I think it's like one in 150 or one in 200. I think YC is pretty upfront that it's one in a hundred. So they don't do a perfect job picking companies, obviously, like most unicorn companies don't go to Y Combinator, but they do hit unicorns on a pretty regular basis, probably a couple of years. Historically, VCs love the sorting function they provide and they love to draft off of that. And so that's what we're going to do right here. We're going to talk about the companies I see here, 100 B2B startups, 17 consumer, 14 in industrials.

37:34That makes sense. Business to business software, people like to accept those to programs. Why? They very quickly get to revenue as opposed to consumers, which take years. Consumers are lightning in a bottle, one in a thousand work out. B2B startups, one in 50 work out. So you have a higher hit rate with B2B, but you have more outliers with consumer. But tell me about the top couple of companies and what they do. Yeah. So one that Lon and I both really liked is called Lightberry, called the social brain for robots. And the idea here is if you have a humanoid robot, if you have a robot in your house, you want to program it.

38:05What if you could just talk to it and then teach it just from voice? So think of it as the no code for voice AI for at-home robotics. It's a very nascent idea. Clearly the market for this isn't there yet because we don't have enough robots in our homes, but I love where they're looking. They're looking a couple of years ahead and that's probably where you need to start in robotics. A company called Dome, Jason, is super cool. It's a unified API for prediction markets because there's Polymarket, there's Kalshi, there's Opinion. There's a lot of these now. And so what if you want to have all that information?

38:32What if you want to use it? What if you want to learn from it? They're building one single web hook for that. So Dome, a unified API for prediction markets, great idea. They're going to need to have, if the information is public, they're going to be able to not scrape it, but they'll be able to collect it. But they're going to need probably permission from these markets if they want to have API access. But a lot of times, small ideas, things start as small ideas and it's a wedge and it opens up into something bigger. So I kind of like the idea, actually. They're skating to where the puck is going.

39:05That's what I was thinking, kind of looking ahead a couple of years. Speaking of looking ahead a bit into the distance, Zephyr Fusion wants to take fusion power, Jason, and put it in space. Now, you're saying we don't have functional net positive fusion power on the planet. Correct. But we are going to need it eventually up in space if we're going to have the in-space economy that we all expect because solar power will only get us so far. Love that. again, sometimes with a startup, you want to skate to where the puck is going and you have five years of build time because this opportunity seems super crazy.

39:36But if you can raise the money to do something super crazy like this, as long as you don't run out of money. So a startup like this needs to keep their burn low and have the right investors. So you need to have like a Steve Jurvetson. And I think his fund is future fund now. And I think he told LPs, be prepared for a 20-year hole period, as opposed to the classic 10 to 15 years. So you just need to match the investor to the window. And that would be what Zephyr used to do here. Also reminds me of Star Cloud. Previously, they had a different name, but they're building the data centers up in space, Jason.

40:05Sure. Also going to be a very patient process. You're going to want long-term capital for that. A couple more of my absolute favorites, Icarus. This is high-flying solar planes for defense context. So not satellites, not normal flight heights, much higher solar power, stay up in the air for weeks, months on end. Really cool idea. People have done this before. This was done 20 years ago. There was a plane that used solar that went around the world. So this is a startup doing that idea. I think that the gist here is, even though we have SpaceX bringing down launch cost dramatically, even though Blue Origin is getting the new Glenn to launch and then land its first stage, it's still really expensive to get space.

40:41So what can we do high stratosphere that is similar to low-Earth orbit satellites and what's the value trade-off there? I think it's a fun place to be. And people are betting, I think, a lot more on launch costs coming down, Jason, than on the ability to use planes up at the stratosphere. What's really interesting about this lesson is when people do experimental things in the world decades ago. So there was a project called Solar Impulse. This happened like in 2011. It was a Swiss long range experimental solar powered aircraft project. The goals were to be the first to circumnavigate the Earth by a piloted fixed wing aircraft using only solar power and to bring attention to clean technology.

41:22Here's one of the images of the first iteration, Jason, of this, clearly a stripped down, lightweight, high efficiency looking design. What's really interesting about this, if you think, and by the way, it was Bertrand Picard, he created that project and he was very famous, like who was like an environmentalist, et cetera. Anyway, this was all done under the auspices of like clean energy, like as a proof of concept. But as time goes on and you think about The efficiency of solar panels, how much they've changed in 15 years, efficiency of batteries, how much they've changed, and then autonomy and chip technology and AI.

41:57These things are going to be able to fly extremely high without pilots, without oxygen, without a bathroom, without safety concerns. You can just fly these things 50 ,000, 60 ,000 feet where there's less atmosphere, and they could be real game changers, and they could be flying all the time. They could be flying in perpetuity. So I think the interesting thing is, how, not can it circumvent the globe, how long can one of these stay up there? And what could they do up there while they're up there? Really interesting technology. I like that one a lot. Two more from us, and then I'm gonna bring up one from producer Claude.

42:31Play Health. This is all about perimenopause care for women. There's been a lot of talk, I think, in the, I don't know, the US community about perimenopause as a thing to pay more attention to, to provide more accurate care, or just better care for women. So I love that there's a startup working on that. It seems like a pretty good market that's not oversaturated. No, there's been a lot of this data. A lot of these apps, actually, women's health apps. The challenge with these is people aren't willing to pay for them. So, you know, with all of these, you're going to really need to make sure that insurance companies are willing to pay for this kind of stuff.

43:04And they tend to be a bit niche. So while it does impact everybody, the key to the success of this company is will insurance pay for it? What will they pay for it? And is it defensible? All that great stuff. So it does seem, looking at their website, they do have a page about providers and they understand this. So that's good. Somebody who advised the company or the founders are very smart and they understand like, hey, to make this work, it's not just about building the app. Because a lot of people build a meditation app. A lot of people build a fitness app, period tracker, pre-menu-pause tracker, all this stuff.

43:37You have to be able to figure out a sustainable, profitable business model. So it has to provide enough value that an insurance company will pay for it or enough value that consumers will pay for it. And then you have to ask yourself, both of those things are hard, right? So if you're going to do these kind of health care things, again, find an investor who understands working with, and this would be my counsel to them, and YC's fell with smart people. So I'm sure Gary Tan and his team said to them, like, the first question you're going to get with investors, how does this scale and who's paying for it and how profitable is?

44:11What are the gross margins? And so it's going to be challenging. But if you figure it out, what if it works? So I like it. Another one that might have a hard time monetizing, Jason, is Sunflower. But I like this idea. It's using an app to help people form new habits to break away from addiction. Have seen this many times, but you went through this kind of thing and did you use one or how appealing would this be to you as somebody who quit alcohol? So what people don't know is that rehab is mostly just AA with a wrapper around it. And so I think that what we don't have is a variety of options for folks out there for whom the main most popular method doesn't work.

44:47And so to me, as a person who also believes in harm reduction and helping people just live better lives, if Sunflower is effective and it claims that it is, and if it can get into the hands of more people, then the amount of human misery that it could potentially take out of society, it is huge. So I'm just very optimistic about them taking this approach. I don't know if this is going to be the company that cracks it or if it's going to monetize super well, but - You're exactly correct. And what I'm going to point out here is with what happens with venture capitalists, a very important lesson for founders.

45:17If you're doing something that is feel good, and on all these lists, like when I asked my team to put a list together, the press puts a list together, they will pick a couple of the feel good ones. And it was climate for a while. It could be DEI. It could be women's health. It could be sobriety. All these things make you feel good, right? Oh, great. Helping society. You just said it. If there's multiple options, reduce suffering. Please understand, you're going to get 100 % of your VC meetings, and they're going to lie to you. Venture capitalists are going to lie to you. And they're going to say, I couldn't get my partners around the investment.

45:48However, I'm going to be as supportive as I can. Please let me know how I can help. And it's like, well, the way you can help is by making an investment. You're a venture capitalist. Like, we don't, I don't need a high five from you. You know, like I need money. So be prepared. And I told my team internally, because what was happening is we were accepting a lot of feel good startups because the team wants to make the world better. I said, I don't, we're not here to make the world feel better. We're here to get returns. And it's the venture capital is really hard right now. And that's what we're seeing.

46:14A lot of venture firms are going out of business. They're in like a stasis mode. They haven't been able to raise a new fund. So they're just living off of management fees and managing the existing portfolio while slowly getting rid of partners. That's like these, you know, you talk about zombie companies. There are zombie VC firms right now. So as the management fees taper off, you go from four partners to three to two to one. And there's just somebody like, you know, wrapping up the fund halftime. That's what happens if you make too many of these feel good bets. That's why it's really hard right now.

46:41So how do you counter that? You have to be a cutthroat capitalist. And the people who are attracted to these opportunities tend to be the feel-good people who want a virtue signal, or historically, and who want the social cred for it. So when Molly was here and we were working on climate, the big problem I had was the valuation expectation was really high. The majority of the entrepreneurs, I would say, were not cutthroat killers who were business builders who wanted to build profitable businesses. they kind of were kind of activist-y and they wanted to save the world. They wanted to save the whales.

47:15I would put it in the save the whales category. So you had a high price and the wrong type of founders. And that's why climate became such a toxic space for returns for LPs. LPs know that. They quietly have moved away from DEI, quietly moved away from that. And these will have that same thing. So I'm not saying this because it's how I feel about it. I'm trying to explain the game on the field to those founders. The game on the field is, Don't give yourself credit for any kind of high fives or getting included in the list here because we want to have like, oh, you know, when lists are built like this, oh, let's have some of the future tech.

47:49Let's have some of the feel good. Let's have some of the like obvious business things. People put lists together. You put lists together at TechCrunch. We were there. I put it together when I was at Engadget, Silicon High Reporter. You have like a little diversity list. Don't give yourself credit for that. Give yourself credit for a customer being addicted to your product and being willing to pay for it and having a high gross margin. Because these founders get led astray. The VCs are not helping them by saying, oh, great idea. Oh, great idea. So when I meet with them, I say, respectfully, love the mission, love that this exists in the world.

48:21What's the way you're going to print money and have a defensible monopoly here and be cutthroat and kill your competitors and, and, and. And a lot of times the founders, the virtual signaling founders, their status seeking founders who want to be the CEO of the company that saves the wells. They want the personal glory for themselves. I see. Correct. And so VCs, this is the inside discussion that happens at the poker table, at the, you know, post YC drinks, post tech stars drinks, you know, in the decision making room at YC or A16Z speedrun. Oh, it's a feel-good company. We're not here to feel good.

49:01How do we let them down easy? Take two meetings, give them some advice, introduce them to a friend, some university professor, whatever. Tell them your spouse is going to use the product, whatever. Tweet for them, retweet them. But don't put our money in it. So just understand, you have a 10 times higher bar. That's reality. If you're in a feel-good space, you're going to win all the startup pitch competitions. And man, I've seen it over and over again. These founders, Alex, they start to believe that they're great, but not because they have a great business with great margins, with great virality, with great go-to market strategies, the great team with great technology.

49:41No, because they want to start up competition. And they won because the person on stage is like, oh my God, oh, it's so precious. You're saving the fucking whales so this is to save the whale trap don't get caught in the saving the whales trap man the whales are doomed i think the whales do pretty good actually i think since we started stop polluting the oceans and people stopped hunting them as much well as well as doing better we're no longer using uh whale oil which is i think a really great thing for the whales isn't that just as a total aside you guys two guys from the northeast isn't it incredible that the northeast East and Nantucket and like that whole corridor was the Saudi Arabia, the Qatar of the whatever 17th, 18th century for blubber converted into oil to run lamps so people could read books at night.

50:31They literally caught whales, cut them up into big pieces and then rendered the fat down on a boat, brought the oil back and then you put it in your lamps. Think about like, think about the air quality in your house, Jason, if you're burning whale fat. Now I will say my house in Providence, It's a historical old town. It was initially owned by a sailing captain. So a little fun bit of history there. Now, before we move on, Jason, we asked producer Claude to also take a look at the YC Fall 2025 list. And I want to highlight a company that both Lon, myself, and producer Claude all liked at the same time.

51:04So right here, we have the top list from Claude. Claude.ai is exceptional. So we said here, rank the most promising companies from the fall 2025 Y Combinator cohort. Shout out to my guy, Gary Tan. Come back on the program anytime. You haven't been on the program in 12 years. He's on which you think are most likely to pull through and get additional funding and growth in the future. Oh, I love the prompt. And he says, I'll search for the information about it. Gets all the results. Let me search for more specific information about the class. Great. Let me fetch more specific details about F25 companies from the Y.

51:37Oh, from the directory. And here it is. Top tier. Most likely to break out. Multi-factor security infrastructure. So the one I wanted to highlight from this list that I just had on the screen is Hyperspell, because both Lon and myself also have this company. It's very interesting. So I thought the three of us all together worth highlighting. Hyperspell creates essentially a separate memory layer that it caches that sits between the user and AI. So if you think about AI just as an engine or a brain and you as the person using it, you want to make sure that it has access to your stuff. But what you don't want to do is sit there and wait all the time for it to go read your email again.

52:09So Hyperspell pulls in your email and your calendar and your context, all of that and provides a contextual blob that then you can use as your own kind of like quick twitch information source for AI usage, which I think makes AI both more personal and faster and therefore better twice over. Other people are working on memory, Jason, in LLMs, but I think this is a particularly cool idea and a good place for a startup to build. Great idea. AI infrastructure, always a big winner. And what I see here is it didn't say how it's making its decision here, But it has founded by, you know, for the multi-factor company, ex-CIA officer and NASA scientist with PhDs, solving critical AI agent security with post-quantum cryptography, already has a strong investor backing.

52:52So it knows some of the signals that we use here in Silicon Valley. So great job to our friends at Anthropic. And producer Claude is amazing. All right, Jason. The New York Times is suing perplexity because they're mad about copyright. Now the people mad at Perplexity include New York Times, Chicago Tribune, Amazon, the cease and desist, Reddit, saying that they're scraping all their content, Cloudflare, alleging that they're going around all the robots.txt norms, Forbes got mad at them, Dow Jones sued them. Question for you is really simple. Is Perplexity doing the smart thing here by trying to just go forward and ignore the legal threats?

53:24Or would you tell them that it's time to slow down and maybe make some friends? There is a legal framework that is going to emerge. and I love the Comet browser and I use it and I'm trying to use some other ones. Nobody's got as good as one as Comet right now, but I love these agentic ones. Now, if I'm looking at a webpage on my computer and I cut and paste it, I edit it, I can do whatever I want on my computer and perplexity as a tool can do whatever it wants. Now, there is a gray area here, which is if a user looks at 100 LinkedIn pages, 100 pages of the Wall Street Journal, New York Times that are behind a paywall, Does a large language model, et cetera, have the ability to put that into their training or to use that or to scrape it in some way?

54:08The answer is obviously no, because that would break the terms of service. But if a user uses it, yes. So we're in the middle of this dance, specifically, I believe, around this, because it does feel fair to users that they should be able to, if I paid for the New York Times, summarize the New York Times with the assistant that I could have from ChatGPT, Grok, or anything. And I can certainly cut and paste the New York Times story and post it and tell you this and say, summarize this. Now, am I breaking the law by doing that? No. When you're a user, you have the ability to remix and use content that you buy on your desktop, generally speaking.

54:44As long as it's not for commercial output and using. But are these things actually benefiting from it? And that's where there's going to need to be some rules on the road. And so perplexity is the first out there. And I think that they're probably on the margins being very aggressive about it. New York Times, as I said, is not going to be fooled this time around. When they had Google scraping them and search engine scraping them, they made a mistake. Rupert Murdoch said, somebody can go find the quote, that Google was a parasite or something to that effect. And, you know, if we all just blocked Google and made them pay, you know, the industry would do much better.

55:20Because Rupert Murdoch was right. Google would be nothing if they couldn't index all this stuff. But Google was fair for some period of time in the minds of publishers because they sent traffic and they didn't take too much content. Now that's all done. People are summarizing these stories. There needs to be a licensing deal for these. Back in 2009, Murdoch said he might block Google searches entirely from his properties. And he said that they have kleptomania and is acting as a, quote, parasite for including news content inside of Google News pages. Right. And he had said stuff like this earlier.

55:50So they're not going to wait because, remember, France, Canada, and Australia all had rules around Google News and having to pay a license fee. And then Google said, well, we just won't have Google News in that country. It's no sweat off our back. We don't make any money off Google News. Long story short here, people need to sit down around a table and come up with a reasonable licensing fee. And there needs to be a tag on The New York Times that, you know, perplexity proposes. and they should just say, hey, we will give you$1 ,000 per story. You produce 100 stories a day. So we're gonna give you 100 ,000 a day.

56:18We're gonna give you$3 million a month, $36 million a year to have the New York Times in perplexity. And the New York Times would say, $36 million a year, we'll take it. Yep. Now, if they said, we wanna give you$100 per story and we're gonna give you$3.6 million, the New York Times is gonna say F off, like Logan Roy said to me, you know, my birthday a couple years ago. Shout out to Brash. Did you see that, Logan Roy? I did not. We'll play it at the end of the show. F off is the right answer and the right thing for New York Times to say to open AI and everybody else perplexity. And that's what's happening here.

56:48And I advise them on this very show. Don't back down this time. Hold all the LLMs accountable equally. And just for the LLMs, you can't have a$500 billion company like ChatGPT or a$300 billion company like Claude, which we just mentioned, partner of the program. You can't have - A$20 billion company like Perplexity. without paying for the content that's going to make it great. So just make that decision to say 10%. I think minimum 10 % of the revenue of these should go to content providers. A minimum. I might even say 50%, like 55 is what YouTube gives providers. So they should just give a percentage of it.

57:24It's a benefit to those programs to have those partnerships. And I showed on this with my Nostra Chicanis ideas. Log in with your New York Times account to perplexity. Authenticate. authenticate with ChatGPT to your Disney Plus account. Just have an authentication process. And if people have a subscription, you can then access New York Times data and have it represented there in a very proper way. That would be so amazing. You would have a competitive advantage. People are paying Reddit now, right? They pay Reddit. On that point, Meta just announced a number of deals with media properties for its own AI product.

58:00This includes CNN, Fox News, Le Mans, People, Daily Caller, Washington Examiner, USA Today. more coming, and they're going to pay when they cite those articles in their AI products. This is a great model. Microsoft also signed a deal with Business Insider recently. We're starting to see this kind of shake out. The worry, though, is if I want to put on the other hat, if you will, steel man this a little bit, it seems to be that the companies who are making these deals are the ones most able to afford them. Meta, Google with Reddit, et cetera. Perplexity has less profitability to play with, but at the same time, that doesn't mean theft is okay.

58:32Now, Jason, I do have your Logan Roy clip. Do you want to see that? Sure. Play it for fans. Here it is. If you're a Succession fan, you'll know who this is.

58:43Hey, everybody. It's your boy, Logan Roy, today on the pod. Yes, I'm telling you all to fuck off. I'm not doing this. Jason, you wish you could be as rich as me, but you won't be because you don't fly in helicopters. but your team at launch wants us to wish you a very very happy birthday and a big fuck off to you i love it that was a great birthday gift from pres shout out to my guy pres who works for me and now has a great company micro one reached a 100 million dollar arr threshold i think you're an investor in this company jason we did the seed round we were the seed investors in this company i wrote this up and then remembered that so it does pass editorial guidelines the reason why this stood out to me, Jason, is that it's another company in the world of connecting experts to AI companies who want to have their knowledge to help them train and tune their models, doing incredibly well.

59:35Micro One started this year at a$7 million run rate. Now it's over 100. It was at $50 million when it raised a Series A in September. And it fits into a broader category. Mercore, which we've had on the show, Episode 2159, is reported to be about a$450,$500 million run rate, same idea. We've also seen scale AI before Meta half bought it, half decapitated it, grow incredibly quickly. They were on the show episode 1005 and search AI also in the space, profitable, growing very quickly, over a billion and trillion revenue. My question here is not to be a Debbie downer because this is insanely impressive, but is this a long-term business?

1:00:12Are they going to need these experts forever? Yeah, because the world changes over time and we're out of data to put into these LLMs. Everything that could have been scraped is scraped. We're now fighting over the scraping of it per the last story. So what's left? Get some experts in here and say, hey, experts, there's a deal with Warner Brothers, our first deal. You're a media expert, Lon. Walk us through the deal. Oh, there's a 10Q. Interpret it for us, Alex. Oh, there's a startup here. Interpret it. And for experts, there is going to be a new job category. So when we talk about job displacement, which I've talked about a bunch, there'll be new jobs created, The new job that's going to be created is experts training LLMs.

1:00:52And then there are these intermediary companies like Micro One that do a great job of managing that process because you need to have these independent contractors accumulated, tested, paid, all this stuff. And it's going to, I would say it's the second inning of these businesses. This is the future of AI. The future of AI is not scraping data anymore. All been done. It's not synthetic data because synthetic data only works in certain areas, like making a simulation of roads. You can't make a simulation of like people analyzing M &A, like fake M &A deals. That doesn't make any sense. So what you need is to have experts who were maybe journalists, teachers, PhDs, developers, whatever it is, come in, give their expertise, and then look at the searches that are occurring.

1:01:39Look at where things are breaking down, where people give the thumbs down to a physics question. Oh, that was a bad response. hey, now train the LLM to fix it. There's a war going on for who can have the best LLM. When we talk about the ranking switching on all these leaderboards, how do you think you move up the leaderboards? Expertise, Mark Rewind, expert at this. Very proud to have seeded this company. They're doing great. I think they have a bright future. I think it's probably gonna go 20X from here. And I think it's a great category. And there's gonna be a half dozen important players in it.

1:02:08And we got lucky enough to bet on one of the top three. So here we go. Very impressive. I'm curious to see where the revenue goes next year. If they keep growing at this pace, they're going to be doing about a trillion dollars in revenue in about, I don't know, six quarters. Okay. All right. Well, from your words to God's ears. All right, everybody. Another amazing episode of This Week in Startups. We'll see you next time. Bye-bye. Bye, everybody.