This Week in Startups - Wiz passes on Google, Meta pushes open-source AI ahead, Spotify’s earnings, and more! | E1984

Starting point is 00:00:00 I think it's a little bit soul-crushing when you go build features that are what you believe is good for your community. And then you're told that you can't ship them because some company wants to put you in a box so that they can better compete with you. Is he talking about himself or he's talking about Apple? I think Mark Pink has said the same thing about Zuck when they shived Zingha and Farmville and Zingka poker. What we should do is we should take that club again and then scroll down, all the headlines of like Facebook turns off, Facebook kills, Facebook, you know. Yeah, yeah, yeah.

Starting point is 00:00:34 This week in startups is brought to you by Assembly AI. Get maximum value from voice data with AssemblyAI. Build powerful products and features for your end users on the industry's leading speech-to-text models. Get 100 free hours to start building at assemblyaI.com slash twist. Vanta. Compliance and security shouldn't be a deal breaker for startups to

Starting point is 00:01:00 win new business. Vanta makes it easy for companies to get a sock to report fast. Twist listeners can get $1,000 off for a limited time at vanta.com slash twist. And eight sleep, good sleep is the ultimate game changer. The newest generation of the pod, the pod for ultra, has arrived. Head to eightsleep.com slash twist and use code twist to get $350 off the pod four ultra. All right, everybody, welcome back to this week in startups. I'm Jason Kalakanis, an angel investor, former journalist and publisher, now doing random acts of journalism.

Starting point is 00:01:38 With me, my partner in crime, Alex Wilhelm, he writes a substack. It's called cautious optimism. Am I correct? Nailed it. First try. Cautious optimism. And it's awesome.

Starting point is 00:01:51 Every day. You can get a note from him about what's going on in the market, and then three days a week is what we're aspiring to to do news here. I am on vacation for two weeks, but yeah, it's just too much news for me to be totally out of pockets. So I'm working half time as many people who work from home do. What's the style of vacation you like, Alex?

Starting point is 00:02:10 You prefer the like work four hours, take the afternoon off or something like that, and then have two weeks of vacation? Or do you rather have one week of vacation totally unplugged? What's your style of vacation? The audience wants to know. Yeah, no, I used to do very, very serious log off, be entirely disqualified. take the time, breathe. And then once you have children, going on vacation isn't actually going on vacation because you give up childcare.

Starting point is 00:02:39 So then suddenly you're just exhausted all the time, even more so than you are. And then going back to work actually feels like a vacation. So I have never tried the four hours on, four hours off, but I'm about to pitch you later this week, my paternity leave plan. So get ready for very similar ideas because I'm going to balance the two. I am totally fine with it. You have been such a great collaborator here in your first six weeks on the program. And I always say there's like rules for, you know, work and remote work and contracts and agreements.

Starting point is 00:03:11 And then there's rules for high performers. And we're going to talk a little bit about sales and this woman who's trending on TikTok and her whole vibes around sales and taking a lot of time off. But I wanted to start with a little bit of banter. here. You know, we haven't talked since Biden dropped out of the race. So the hot swap has occurred. The hot swap has occurred. I told you last time we were on the show that in the wake of the assassination attempt, people were rallying around Biden. And I said, I'll forgive you the hot swap prognostication because who could see, you know, an assassination attempt coming? And then you ended up being right anyways. Now, I'm not here to fluff the old ego, but I will say,

Starting point is 00:03:52 I expected it to happen. And even though I did, when it did happen, it still felt like a real moment. Yeah, it was a real moment, you know. And I, everybody's so caught up in the outcome of this election. I am a moderate and my vote doesn't count for me living in California earlier this year and now living in Austin, Texas, where I moved to. So there's your announcement, folks. People were wondering where I was going to move.

Starting point is 00:04:21 I'll just tell you right now. Austin, Texas. I moved and I actually have been in Austin for a couple of months. As many of you noticed when on All In or other times, I said I was in Austin breaking my teed eating beef ribs at Salt Lake. But yeah, our family decided to be homesteaders and we bought a ranch. Now I will still be, I will still be doing a bunch of time in the major cities where our startups are, whether that's New York, San Francisco, L.A. or Miami, which a lot of our startups are in. And so It's a very little change in terms of my actual business life, but a big change for our family to live on horse ranch. Yeah. Also, it's a really big change for my life because I now have a free and open house in the Bay Area that I can borrow whenever I want.

Starting point is 00:05:03 Absolutely. Absolutely. Free properties for sale. I know. If anybody wants a modern masterpiece in the Bay Area, a ski house in Tahoe or a loft in the city, all properties are now for sale. Buy two, get 10% off. I'm joking. That would be a lot of cash to care about the 10% discount. But on the show today, everybody, we have the end of the Wiz Google deal. Meta's incredible new AI model. That's open source, Spotify's sparkling earnings. And if we have time, what happens if you just take money and give it to people every single month? But Jason, the biggest story is, WIS and Google are not going to get married. No $23 billion deal.

Starting point is 00:05:42 No happy VCs. Yeah, well, I mean, this could be a case. of going for long-term greed over short-term greed. So the deal for $23 billion to Google does have a risk of interference from different government agencies, potentially, right? All of these deals do, whereas an IPO has zero risk. And after seeing what happened with Figma and Adobe,

Starting point is 00:06:12 if I was a VC in this incredible company that has gotten to this, you know, call it $20 billion market cap, at least the offer was $23 billion, and their last private round, I think was $10 or $12 billion, something in that range. You know, the optimal thing to do

Starting point is 00:06:32 would be to go with what's the safest bet, plus or minus 20% of the valuation, right? Because this is going to be an extraordinary outcome either way. So if I'll kick it back to you, what has a better likelihood of hitting 20% less than $23 billion is minus $4 billion, so let's call it $18 billion. Would you rather go for the $23 billion in a sale and get all of your money at once? Would you rather, say, take $18 billion in a public stock and be able to sell some shares

Starting point is 00:07:08 after the lockup or hold some shares and have that optionality if you were a large share. holder here, Alex. Oh, option B, day, night, weekends, and pick a Sunday, absolutely, for a bunch of reasons. But the first one is, I don't really want to go work for Google, right? I mean, if I'm building a company called WIS, if I'm the CEO, Asoff Rappaport, by the way, do you know what a person who works at WIS is called, they call themselves Wizards? Yeah, I mean, I was going to get there. I was literally put, I was Wizzettes.

Starting point is 00:07:40 Anyways, I thought that was adorable. Okay, congrats to the Wizards. Yeah, but here's the thing. To be an independent company means you could set your own culture, set your own style of business, build products. No one's telling you in some different feet from the company, what you can and can't do,

Starting point is 00:07:56 how to communicate, how to lead, how to hire. And Google is one of those companies that has its own vision in a long time to grow into it, and it has its own way of doing things. And I guarantee you it's not the way Wiz does things because Wiz is small and Google is enormous.

Starting point is 00:08:10 So for me, I think that I would take the best, that you offered. My question is, what's this thing going to be worth? Because the crazy thing is, Google was offering $23 billion for a $350 million ARR company, which is a pretty rich multiple for the current time. Now, it's growing very quickly. Hazaa, fantastic, very happy for it. Let's say that it does do what the CEO wants, which is reach a billion in ARR and go public. First of all, love those goals. That's fantastic. We're going to do a billion and go public. Love it. But today, how hard is it to get a 23x ARR multiple on the public markets?

Starting point is 00:08:47 So the question is, if it's impossible. Yeah, it's, I would, dare I say impossible in the public markets right now. You might have some excitement, weird things can happen. So never say never, but, you know, maybe they'll get 12 to 15 times forward-looking revenue, this billion dollar goal. You know, a hundred times, 350 million is 35 billion. So I guess that offer of $23 billion is pretty rich, right? that's where I ended up getting.

Starting point is 00:09:14 So you said 18 and the future public are 23. Now I'll take the 18. I just hope that it's not worth 12 when it goes public because then the math begins to hurt a little bit more. But let's flip it around. Jason, same question to you. You're the CEO of WIS. Google shows up $23 billion, taking the check, leaving the check. So, I mean, in the current administration, I think I'm going public.

Starting point is 00:09:38 But I'm telling all my investors that we're doing a secondary at $18 billion. right now and every employee and every founder gets to sell 20% of their shares. So we need a $4 billion, you know, or whatever it is, a $3.5 billion investment to go long and or let's say 10%. So we need $1.8 billion in secondary. Go get me, you know, Josh Kushner at Drive, is that his company where they just spent whatever amount on, you know, buying shares of opening. I find me someone like that who loves to, you know, buy at the top of the market.

Starting point is 00:10:13 it shout out Josh Kushner. Soft bank. You know, somebody who really likes, has the guts to do these kind of deal. So I don't mean that as a dig. You know, if you have the guts to make these kind of a deal, like Josh Fisherner did with Open AI, you know, or Massey Yosh's San did with countless companies, it can pay off in a major way. It just does a little bit of risk. But those people are out there. So find me one of those. And let's just clear 10% of our position. There's four co-founders, I believe, they own roughly 40% of the company. Yep.

Starting point is 00:10:41 You know, so if those four co-founders can take, you know, 100 million, 200 million off the table right now, that gets you a jet, at least a jet card, get you a nice ski house, pay down some debt. I mean, let's go. So easy, brazy, lemon's easy. Jet cards only cost like a quarter million a year. They're not, they're not as bad as people think. Getting an actual jet is expensive. Jet cards is like middle class venture capital. I think, I think this though, this whole saga does kind of show to me how hard it is to do. do 2021 style outcomes through an IPO with 2024 because this is a company that is mind-bendingly good. It's growing so fast. Everyone wants a piece of it. It's got insanely great backers, lots of capital raise, cool culture, moved to New York City, so it's going to be based in the U.S. to have access to the best capital markets in the history of mankind. And we're looking ahead towards a billion dollars in ARR and we're saying maybe the math won't math out when it goes public then.

Starting point is 00:11:38 Like, that just doesn't feel like the place we should be when we're discussing a company that could be, could be, the next decacorn, whatever, and really change the game. I'm confused what's broken. I think what's broken is if you have a lot of cash, you can make five, six, seven percent safe on your money or you can put it into corporate credit like we're just talking about. I know that didn't work for the beyond meat folks, but there's plenty of stronger companies that have corporate debt that pay 10, 12, 15 percent. So really capital is going to find, you know, the returns. And if it right now in a high interest rate environment, you know, doesn't think that these companies are going to be valued really well and that MNA is off the table. You know, there's a major discount.

Starting point is 00:12:22 So, oh my Lord, podcasts, audiobooks, virtual meetings, endless videos. We're producing and consuming more voice data than we ever have. And if your startup is building a product that you, uses voice data, you need to check out Assembly AI. Assembly AI builds speech-to-text AI models that can turn your voice data into new product capabilities. All you need is voice data and a few lines of code. For example, VDIO makes video editing tools that generate the captions you see on our video

Starting point is 00:12:55 clips, right? These are really cool. It saves us a ton of time. Well, that would not be possible without Assembly AI crunching all that audio data for VEDIO. So, Assembly AI, speech-to-text models are simple, accurate, and fast, and they've got the industry's lowest word error rate. 30% of your hallucinations, and it's so easy to use. So if you want to put your audio data to work and you want to build powerful AI experiences for your customers, head to AssemblyaI.com slash Twist and get 100 hours for free.

Starting point is 00:13:28 And join over 200,000 developers who are building amazing apps with voice data. go to assemblyaai.com slash twist a day. I like the idea of being public like Spotify did. If you're public, the amount of discipline you have, and when I watch the video of Daniel Eck today, where he did a video from Sweden on vacation, he just looks in the camera and is like, we're kicking ass.

Starting point is 00:13:52 Let me explain to you how much ass we're kicking. Like, that's what every CEO aspires towards, which is lots of shareholders, lots of expectations and a lot of pressure on the team and then you hit those notes and you get rewarded in the public market for they got, is it over 200 million subscribers and 600 million monthly active users? Something in that range, he said.

Starting point is 00:14:15 We'll get to the numbers in a second, but I want to play the clip. We haven't pulled up for you. Oh, you do. Great. We're going to start at 13 seconds, which I believe is when he starts to really spit facts. Okay. Taking this quarter.

Starting point is 00:14:25 The bottom line is we had a really great quarter. We are now over 246 million paying subscribers, 626 million monthly active users. And financially, we had our biggest free cash flow quarter in the company's history. I really attribute the success to all the innovations that we keep on making to make the product better and better and stickier and sticker. First of all, he's stealing my style. Like, big glasses, bald head, too close to the camera.

Starting point is 00:14:58 That's me. Daniel, come on. calm down. But it is a flex to just stand there and be like, hello, do you see the Swedish landscape behind me? Great quarter everybody. Like,

Starting point is 00:15:09 uh, okay. I mean, this is where we're at is just CEOs popping out their phone. I don't know where the comms department is. He's clearly not reading from a teleprompter. I mean, he may have rehearsed it once or twice,

Starting point is 00:15:18 but he's just like, hey, let me tell you how we're doing here. Now, they went out with a direct listing. Their stock is, I think, at an all time high.

Starting point is 00:15:26 They've got free cash flow. And so, you know, Bill Gurley, my, my pal Bill Gurley, always talks about how when you go public, you know, this forces a level of focus. And that's what they wanted for the ubers and door dashes of the world. And it just, you know, and for Stripe,

Starting point is 00:15:43 etc. And it's so easy to stay private for so long that you have to admire the people who do go public, who take the hits sometimes on valuation. They, they don't take the easy sale. And then they get to this point where now Spotify has a lot of influence. And they have, you have, you have, I think, reduce their dependency on the music industry a bit by having this unlimited audio book product, which I haven't tried yet. I now subscribe to three musics as I find myself accidentally. So I'm going to have to consolidate and have Apple music for free as part of my Apple family plan, because I went on that Apple family plan. That gives you everything. Then I got these high fidelity headphones as I talked about a bit. Shout out to my friends at

Starting point is 00:16:30 headphones.com, not a sponsor yet. And I bought Kubuz, QU-B-Z, which is this incredible high-fidelity service where you get like the flacts, like the original audio files from the studio. Real difference when you're using proper equipment. And then, of course, I have Spotify because I've loved that in my whole life. And now I find myself like, I use Spotify for playlists. I don't use the Apple music. And then I use Kobaz for when I'm listening to Miles Davis or Dire Straits or

Starting point is 00:17:00 some really high fidelity, amazing stuff where I want to hear the nuance and I want to actually hear the music. So I am enamored by his business. I think the podcast stuff he did was bold and audacious. Now maybe they had some missteps, but that's part of learning.

Starting point is 00:17:16 And now that they have that robust advertising business there and they let Bill Simmons's podcast and Joe Rogan's podcast in the 2.0 version of this, now they're available on YouTube and other platforms. And I guess they just take the advertising revenue wherever they can get it in super distributed, not exclusive. That's a brilliant idea, too. So, you know, I think they could run the table on podcasting if they wanted to.

Starting point is 00:17:37 I mean, if they came to me and said, hey, this week in Starbucks makes this amount of money, would you like us to take it over and we'll pay you, I don't know, they paid me a 30% premium or something? I might take it. We've been sold out. I might take it and just be like, okay, yeah, you take over the answer. Sorry to my head sales team. We're probably listening to this. Yeah, I was going to say, Jamie's going to have a viewpoint on that one.

Starting point is 00:17:58 Anyway, I'm just saying, like, if you were a podcaster and they come to you and they look at your revenue and they say, hey, plus, you know, we'll give you one point X and X is a three handle or more. Yes. You have no choice but to consider it because they would be guaranteeing and taking out risk. And so it's kind of brilliant what they're doing. And I think the sky's the limit for them now. They own audio. And I don't know why they're not doing video production more. They should be trying to compete with YouTube.

Starting point is 00:18:24 They should be trying to get Mr. Beast next. They should be trying to get these YouTube channels. and say, stop publishing on YouTube, publish to us. Oh, okay. I see what you're saying. You're not saying they should get into video production. You're saying they should get into video hosting. Yes.

Starting point is 00:18:37 Okay. They do host video versions of this podcast and all in. And I think they opened it up to everybody now. We've been in that beta for a couple years. But why not go to a YouTuber who you think is awesome for your existing ad base? Let's say they got an advertising deal with Samsung, you know, and they already sponsor Call Her Daddy, and they sponsor Bill Simmons, and they sponsor, Joe Rogan, like they're,

Starting point is 00:18:59 they're, I think those are their top three. Yeah. Okay, so they got the ring on all those dialed in, but Samsung wants more. So they say, okay, great, we're going to take Marquez. It's Marquez, right? Yes. Who does, like, the YouTube channel, just go to Marquez and be like, hey, what are you making on YouTube?

Starting point is 00:19:14 We'll, uh, give you a guarantee. You can still post to YouTube. We just want it here first. So give us, you know, 24 hours on Spotify and then everything goes there. How does Marquez not take it? I mean, if they offered them 1.5. Well, Jason, right now, Spotify, is going through a period of financial strictness.

Starting point is 00:19:32 So while I like the theory crafty, and because I live in Spotify from the moment I wake up to the moment I go to sleep, totally on board, but they've ended up working on cost control and margin expansion. So, Mr. Beast, I can't imagine being cheap. Didn't he also do a deal with Amazon Prime or something like that?

Starting point is 00:19:51 Yeah, that was for a different show, though. So, you know, anyway, I think these opportunities are out there. They would obviously have to be disciplined in doing them. Probably the first round of, deals they did, maybe they were less disciplined and learning. But now that they understand the model, you know, if they lose, you know, 10s, low tens of millions of dollars across a couple of these deals, but they lock in, you know, advertising and it grows audience and they get more members, they've got two ways

Starting point is 00:20:19 to win, advertising, and then subscribers. So now let's say this week in startups and all in became part of Spotify. I'm not having any discussions about that. So I'm just speaking completely hypothetically here. But if they did make an offer for both of these, one of which I own 100%, one of which I own 25%, have one of four votes on,

Starting point is 00:20:37 here I'd be like, hmm, at one point, again, three, four, five X are existing revenue. I would have to take a look at it. And, um, they would then have to make point three X, extra on the subscribers. So if you get some number of net new subscribers,

Starting point is 00:20:57 okay. So, you know, they could make the same amount in advertising and then actually convert some number of subscribers. And there's historical precedent for that working. So did you ever watch Top Gear, the old BBC show? Yeah, of course. Yeah, yeah. That was like the number one show the BBC ever produced, right? I believe, Dr. Hill and that.

Starting point is 00:21:14 Yeah. And so the three guys ended up leaving. It went over to Amazon Prime and you had to sign up for Prime to watch their new show Grand Tour. And you know what people did? They signed up because that show was, that show had just a rabid audience. So there is precedent for this. It'll be curious to see where Daniel Eak takes the company now because he's kind of done what he set out to do, which was cut costs in the last quarter. Operating expenses were off 16% year over year.

Starting point is 00:21:38 That's one at every $6. That's a lot. And their operating income went from negative $247 million to positive $266 million. Their free cash flow went from $9 million to $490 million. And the real thing is their paid subs rose 12%. Their revenue grew 20%. because they raised prices at the same time they were cutting staff. So their profitability has just gone bonkers because they managed to grow the top line and

Starting point is 00:22:06 shrink the cost basis. I mean, it's kind of a beautiful moment for them. So I wonder when they're going to have the straight up, like belief in themselves to take a risk again. Because I feel like they spent a lot of money on podcasts, maybe a little bit too much. And then they had to talk to Wall Street and say, hey, we're going to get cost of control. We're going to bring this back. Don't worry. They did that.

Starting point is 00:22:25 But now do they stay where they're. are with the happy Wall Street and stockback in record highs, or do they take another risk and say, we're going to double down on video, audio books, whatever it is to break this relationship. Yeah, we had this as one of our themes that we've been working on, you and I, for 2025, which is the static team size. And that is something I brought up with the CEO of Etsy. We just had the CEO of Etsy on the pod. It hasn't come out yet. Great guests, by the way, and great notes. But the static team size is, I think, one of these themes that I've been talking about now for two years, and it is clearly infected every company. What is static team size?

Starting point is 00:23:07 It means you keep the same number of headcount. You empower those individuals to be 20, 30 percent more efficient each year with tools like AI, and then you fire the weakest performers and replace those with higher performers. and instead of trying to grow headcount because you can and because everybody has done that since Google started this trend and Facebook matched them in this war for talent, you don't compete in the war for talent. You take your existing static team size,

Starting point is 00:23:37 the same thousand people, or 2,000 people in the case of Etsy, or how many people it is here, and you just say, we're demanding excellence, we're giving you tools, we're going to give you more stock options, we're going to keep the team size the same. If we're more, if we're 30% more productive each,

Starting point is 00:23:51 which is, I think you and I become 30% more productive every year, right? I feel that way. If you're 20, 30% more productive each year, I would rather get raises than increase the number of people of the company that I have to manage or that need to be managed and that we need to hire. Hiring is brutal and time-consuming, managing people, brutal and time-consuming, just find great people you trust,

Starting point is 00:24:16 empower them and be done with it. So you raised, so one, generally speaking, I agree. I do think our productivity is rising, especially if you live at the forefront of new tooling. Like, I don't know. Since I've joined the Twist family, I've learned how to use both Notion and Coda. You know, they've been fantastic. I now can do a lot with them that I didn't do before.

Starting point is 00:24:36 I used to live in Google Docs. I still kind of do. Yeah, I do think we improve over time. The reason why I think it doesn't quite work out the way people wanted to is that when a company can grow revenue and keep headcount flat, they don't tend to change their salary increase chart. And so people tend to get a bit more work without the kind of more than a two or three percent raise.

Starting point is 00:24:58 And this is why I think we see so much job hopping and so forth. So if a company said, we're going to keep headcount very low and we're going to grow and we're going to take a big chunk of that new profit and we're going to pay ourselves with it versus we're going to sit on it. That is actually, that would be actually a really good, you know, for the for the employment side to be saying, hey, we kept, we did our part, we did more with less, and then we achieved 30%, hey, let's, you know, we have a 2% standard raise, 3% standard cost of living raise.

Starting point is 00:25:29 hey, maybe we should talk about 4%. Maybe we should talk about 5%. Maybe we should talk a little bonus here, maybe some equity. And I think, actually, that is a really healthy dynamic. And so, yeah, you do hear about the complaining, like, you know, oh, you know,

Starting point is 00:25:43 we laid everybody off and now I have to do more work. You know, the people who are valuable today are the ones who learn the tools. And just, you know, when we do show notes and research, I'm taking podcasts or cutting and pacing the transcript. I do this myself. I take the transcript. I put it into Claude and I say, give me the 10 most important bullet points here with citations, yada, yada. And so I take the transcript of somebody who's going to be on the show, like here's the Etsy founder. You know, here's their

Starting point is 00:26:13 video from four years ago. And I'll watch that video. And sometimes you guys give me the summaries. And I just say, hey, what are the major bullet points and give me the timestamps? Boom. That's, you know, a half day for a producer. It's five hours of producer time to listen to that bond and right many of those notes. Yeah. So, you know, I do think it's a, there's a great dynamic here for both sides of the table to, you know, get a little extra comp and a little extra efficiency in these companies.

Starting point is 00:26:39 And then you know what can happen with all the other challenge people? They can get UBI. How's that for the transition? I used, you stole it. I stole your transition. I'm sorry. To the Facebook area. We're going to talk about some AI models.

Starting point is 00:26:50 But I was like, you know what? I think this isn't trending in a, the worker money direction. Let's just take the UBI. Let's go. Listen, a strong sales team can make all the difference for a B2B startup. But if you're going to hire sharks, you need to let them hunt. And you can't slow them down with compliance hurdles like SOC2. What is SOC2?

Starting point is 00:27:09 Well, any company that stores customer data in the cloud needs to be SOC2 compliant. If you don't have your sock too tight, your sales team can't close major deals. It's that simple. But thankfully, Vanta makes it really easy to get and renew your SOC2 compliance. On average, Vanta customers are compliant in just two to four weeks. weeks. Without Vanta, it takes three to five months. Vanta can save you hundreds of hours of work and up to 85% on compliance costs. And Vanta does more than just SOC2. They also automate up to 90% compliance for GDPR, HIPAA, and more. So here's your call to action. Stop slowing your sales

Starting point is 00:27:41 team down and use Vanta. Get $1,000 off at vanta.com slash twist. That's Vanta.com slash twist for $1,000 off your SOC 2. I want to stay on that last point though. So if you had a company that was growing, let's say 30% per year, growing headcount, 5% per year. Because you're going to add another person in the mail room, whatever. And then the company keeps doing better and better and better. Everyone's happy.

Starting point is 00:28:03 If you offered people something like a pension or something, like an old school style of pension, I wonder if you would have like akin to 100% employee retention. Because that would be something that no one else offers. Nobody has pensions anymore because we all, the biggest problem with pensions right now, why it would be insane to offer a pension? is because with the gains in AI,

Starting point is 00:28:23 we actually don't know how long people are going to live. Oh, you're talking about the long tail of life extension. Life expectancy could be 110. You are for a pension now, like the New York Times. I mean, this is an incredibly weird topic to bring up, but, you know, we know the tail end of the pension people at the New York Times, like probably John Markoff or, you know, I'm trying to think of the old school tech journalists, Saul Ansel, who came before us.

Starting point is 00:28:49 You know, the people, Walt Mossberg, people who were, 40 now to my 53 and you're 35. Congratulations, happy birthday. You are very old. As I told everybody, man, to get those 18 years back, I'd give you every penny. They got pensions. They got pensions at the New York Times. Like actual pensions.

Starting point is 00:29:08 And it was like, you're going to get 50% of your salary indefinitely. And it's probably your last three years. So, like, you hit key comp of 150 a year at the New York Times or 125 a year. they took those last three years, they averaged them out or something probably, and then you're going to make $60,000 a year, which is what the Cub reporter is getting on the way in. So it's almost like John Markoff or Saul Hansel, if they did hit those pensions, I'm assuming they did, Steve Lorr, some of those folks. Yeah. They're probably getting $60,000 pensions right now. And they count as headcount. I mean, it could bury a company if they live to 110. So morally, though, do you think that pensions are a good, or do you think that they are something that is morally neutral and therefore the fact that we've moved away from them isn't a loss? They're not aligned properly and they will bankrupt any company or system that employs them or, you know, or that strategy because people used to live to 65 and the average lifespan was 71 for a man.

Starting point is 00:30:08 So you'd work to 65, you get your gold thing and you had six years of pension. if your wife happened to live, you know, you know, an extra eight years, I think probably life expectancy in these cases were like 72, 78, something in that range. So they looked at it like, hey, you put in your 30 years and we paid another six, and we paid a quarter to your spouse,

Starting point is 00:30:28 because there were some sort of triggers where like your spouse could get a portion of it, you know, what really has to happen is we need to do what they do in Australia, which is the super annotation funds, where you're forced to put 12% of your salary every year, 9% of your salary into essentially what's a 401k. Because of freedom in the United States and, you know, a lot of weird things like democracy

Starting point is 00:30:50 and it being bought and sold by either side of the aisle every four to eight years. We don't have a thoughtful approach to retirement. The really thoughtful approach, you need only look at Australia. Everybody in Australia is happy. They have no stress because they had a gun put to their head and they said, you have to put 9% into this super annotation fund, you can't touch it. And everybody complained.

Starting point is 00:31:14 And then they watched the stock market rip and they're like, hey, you know what? I'm going to retire. It's 60. And I'm going to give my job to a younger person. So they actually have a high functioning society there where people, in order for people, you know, in a static team size, how do you get the slot?

Starting point is 00:31:30 Somebody has to retire. Somebody has to retire. Right. You have to give up the seat. So somebody gives up the seat. You and I retire. Two new hosts of this week in startups. Boom.

Starting point is 00:31:38 And the cycle goes anew, and that's what superannotation would do. No, I'm here for it. But we do have some data out on a new UBI study. Not Ubi. It's universal basic income, giving people money for free. I know if you aren't American, that's blowing your mind, but they did it. And the experiment was for three years involving 1,000 lower income people selected from Illinois and Texas. They got $1,000 a month.

Starting point is 00:32:05 There was also a control group that got $50 a month, just to show kind of like two different data. sets. And this is all done through Elton's, Sam Altman's, nonprofit, open research. And unlike with Open AI, they actually did drop a lot of data and information about this. We have parsed through it.

Starting point is 00:32:20 They broke it down into individual buckets of findings, including entrepreneurship, health, home, and a couple of other things. I have a bunch of information from the study itself here, Jason, but I don't know exactly where you would like me to start. We can talk about results. We can talk about

Starting point is 00:32:37 expectations. Where do you want to dig on? Well, I'll first want to start with, you know, Sam Altman does interesting things with his money. So he put $14 million. I understand of his own money into this. And I think it was while he was a Y Combinator, uh, was president of Y Combinator and some other group of people put in the rest of the 60 million. And I think this is a worthy test because I have deep concerns about free money. Um, because, you know, I worry about motivation. However, I'm smart enough to know that all experiments are worth trying to get more data.

Starting point is 00:33:14 Yeah. And so I am thrilled that this occurred, and I'm very interested. I understand some of the points here, but I'm just interested to get into it because what people seem to think about UBI. And talk about Preciant, you know, we're now really talking about UBI at this moment because of what we expect from AI. So UBI, three little letters, AI, in the words of Trump, two little letters, okay? Very important. A.I. Two letters, okay?

Starting point is 00:33:43 You can't just drop a random Trump, but I'm trying to say no. Sorry, sorry. Nasty, nasty lady running secret service, okay? I mean, slope, rupe, okay? I don't know if she's qualified or not. Speaking about hire and fire fast, she didn't make it through that congressional hearing. I think that was the most panned performance by any elected official I've ever seen in front of Congress. no partisanship, just pure anger and just...

Starting point is 00:34:09 I mean, it was infuriating to me when I saw the clips I saw of her, because, you know, I like to think that as Americans, there's like a civic duty pride, especially in the pinnacles of leadership. You know, when you're the CEO, when you're the director, when you're the president, whatever it is on the board.

Starting point is 00:34:31 And all I saw there with her obfuscation? Obfuscation, thank you, observation. I was, she's got a pension. Back to pensions. She's got a retirement. And it was so obvious to me, because any time you want to understand the world, just understand the incentives. If she resigns, she probably gets less than if she's relieved of duty. Like any other person, there are triggers.

Starting point is 00:34:58 So if they said to her, we want you to resign, and she says, okay, that's interesting. They're like, so you resign? Nope. Because if I resign, my pension is this. If you relieve me of duty, I get this acceleration and I get my pension. I guarantee when we double click on this, her whole farce of not answering questions is going to be based on her personal economics in filibustering. So she went on to that, you know, that whole rigamarole was done just.

Starting point is 00:35:35 so she could get some amount of cards. Somebody should have just come in and just said, listen, we'll give you everything. You're relieved of duty. You can call a resignation, go all whatever you want. And I learned this when I was young in entrepreneurship. I got mentored. And a human resources person said,

Starting point is 00:35:55 here's what I want you to do. Because we were firing somebody. And we were firing somebody, paradoxically, for harassment, a woman harassing another woman in a company. And it was like, I'm 27 years old running my first magazine. And one 23 or 24 year old woman told another woman about her anatomy and her body. Like, I was just like, what was going on?

Starting point is 00:36:19 Like, why am I in the middle of this? And that person complained and wanted to sue. It was nuts. So the person says, here's how you handle it. You take a check for one month of salary. And then you have a letter here of the person being terminated. And then in this one, you have the resignation letter for them to sign the separation agreement,

Starting point is 00:36:40 walk into the room, say, sorry, it didn't work out. I know it's a complicated situation. There is one month severance if you sign this piece of paper. Here is your termination if you don't. And no month severance. And your desk is being cleaned out right now. You're in the conference room. Your desk is being cleaned out right now.

Starting point is 00:36:57 All your accounts are turned off. So you're no longer working here. And I just, do you hold the two envelopes? And the person goes, thank you for my severance. and they signed the piece of paper and we're done. Is that how it went? Of course. Of course. Yeah. Nobody. I mean, and I said, we will never speak of this again because you resigned. It was your choice.

Starting point is 00:37:17 Right. You made some decision to leave as opposed to we fired you for cause. Here's the letter. Yes. But there's one last clause you didn't add, which is, and don't call us for a reference, because it's not going to, it's not going to go in your favor after we've gone through this and we've let you go out. I wonder, I think if you make a threat like that about references, maybe that could be, double actionable. I think what you do is just we will not, I think that, yeah, the last time I had to deal with that humorous thing is we just don't give references because it's too much liability. So a lot of companies today have a rule. We do not give references good, bad, or

Starting point is 00:37:53 otherwise because of liability reason. We will tell you when the person started and when they left. Those are the two pieces of data you can have. So anyway, let's get into this UBI because I do think this is super interesting. 8 Sleep has released the Pod 4 Ultra. Besides heating and cooling, you know all about that. Now it elevates automatically, which is so nice for reading, or if you want to do a little Netflix and chill, it is the best way to prevent snoring,

Starting point is 00:38:19 which your spouse is going to be thrilled with. So Pod 4 Ultra can cool you down on each side of the bed to 20 degrees Fahrenheit below room temperature, so you and your partner stay cool, even in a heat wave. When you have the Pod 4 Ultra, man, it is like, being just absolutely cool, calm and collective. You're going to get a great night's sleep. Also, the pod four has an adjustable base that fits between your mattress and your bed frame. So, you've got reading and sleeping positions to help you unwind after a hectic day. Not only that,

Starting point is 00:38:49 eight sleep has amazing sensors. And those sensors are going to help you track your sleep time, the sleep phases, HRV, and your heart rate. So it's basically a wearable. And it's going to give you that great sleep score so you understand how you're sleeping. You know, for me, it's all about getting that extra hour of sleep every night because if you got to make decisions the next day, you want to be sharp. If you're going to be on camera, you're going to be a leader of a company or a podcaster. You need great sleep. So here is an awesome offer for our audience. 8Sleep.com slash twist. Use the code twist to get 350 bucks off the pod for ultra. What a generous discount there at 8Sleep.com slash twist. Currently they're shipping. United States, Canada, UK,

Starting point is 00:39:29 Europe, and Australia. So you can get it down under now. Okay. Well done to my friends at 8Sleep. Tell me what you found interesting here. NPR did a little bit of digging. I'm going to quote them because they have good verbiage. So overall, people who got the cash payments worked a bit less, 1.3 hours less per week on average. But that does include some people who are logging 50 to 60 hours a week. So some people were working marginally less,

Starting point is 00:39:53 but they were already working time and a half. And keep in mind that these residents had an income of below $28,000. So $12,000 a year is like, you know, like $50. It's huge. There's also some data... And by the way, this was tax-free because it was a gift in some way, or it fell under the gift clause, I believe. So they were not paying tax on this $12,000. Okay.

Starting point is 00:40:18 And then also, the study found that it did drive it more interest in entrepreneurship, but it wasn't until the third year of payments that some folks really kind of got that bug. And I was trying to figure out why there was a lag between cash showing up and more. entrepreneurial spirit, and I think I figured it out. Okay. If you're just getting by, barely, and you get some more money, you're going to fix the big things first. Your kid's shoes are falling apart. Your car is broken.

Starting point is 00:40:46 You're going to really get everything patched up. Broken dishwasher, whatever it is. Yeah. Maybe your AC doesn't work. Maybe you're behind on the light bill, whatever it is. You're going to work through all of that, build a little base, save a little money, and then you're going to start thinking, I should do something with that money. Then you're going to work on your company, your idea.

Starting point is 00:41:04 They did notice that there was a greater impact on entrepreneurship and kind of like the desire to start a business and ideas for a new business amongst women and minorities. And so that was kind of cool to see, bigger impact there in terms of driving entrepreneurship interest. And then something that I really, really found fascinating was, and this is a quote from the study itself, recipients reported decreased problematic alcohol use and some types of illicit drug abuse. We also find notable improvements in stress, mental distress, and food security during the first year of study, comma, but the effects fade by the second year. So people get used to free money is what you take from that. And so this is a perfect example of a mixed bag, right? And so if you were to tell me what I expected, it would be a mixed bag because not everybody is the same.

Starting point is 00:41:54 I'm sure some people get the money and they take it for granted and they spend it on booze. And I'm sure some people get the money and they say, you know what, I've always wanted to be an entrepreneur. And once I get, you know, my credit cards paid off, I'm going to think about starting a business. So it's a mixed bag and probably what needs to happen now is you do a second study where you build on this one and try four different modalities. And I think, you know, if I were going to build another study on top of this, instead of offering $1,000 in free money, a month and doing just strictly UBI, I would tie it to, hey, you can get these UBI payments if you take an entrepreneur course, right?

Starting point is 00:42:39 So something's tied to it, right? And that could be, I think, you know, the combination of UBI plus something else, you know, paired with it, I think could be interesting. And the reason I think that's interesting is because having watched what I saw happen in the Middle East where I met a lot of young people, you know, your age or younger, who had the King Abdullah scholarship or this scholarship,

Starting point is 00:43:07 they had told their citizens, go anywhere in the world, not only will we pay your tuition. You don't have to apply for a scholarship, but just pay your tuition. We'll pay for your apartment. Not only will pay for your apartment and your living expenses, we're going to give you a stipend.

Starting point is 00:43:22 In other words, UBI. What happened? A whole generation of people in Saudi, Abu Dhabi, Doa, you know, the entire region blocked to the west,

Starting point is 00:43:34 got degrees and graduate degrees, and then came home with massive knowledge to build their economies. And so I love the idea of tying it to education and then seeing what the outcome there would be,

Starting point is 00:43:45 right? I mean, what do you think? What would be, you have an idea of how to level up the study and what you would do next? Well,

Starting point is 00:43:51 I would change the geographics. I would change the dollar amounts. I would change the frequency. I would also set up one program with triggers. So, for example, you know, we're going to give you a thousand bucks a month, but if you get a job that pays 20% more, we're going to jack it up to $1,500 a month. Just tag on some incentives and just see how fast people react to them because we talk

Starting point is 00:44:11 a lot in economics about price elasticity and you kind of demand elasticity. I don't think we have enough data points here to actually see how people's behaviors are willing to change or able to change under these circumstances. So I want to go back to what you said. I love an experiment. I love seeing this. I love the fact that it also had some really cool impact on people's lives. So going from the macro to the micro, one of the study participants said that when he learned about the cash transfers,

Starting point is 00:44:40 he went to his boss and said that he wanted to cut his hours so that he could spend as much time as possible with his four-year-old son. And that's the one thing that I think that these conversations often miss is that it's impossible to measure the GDP of joy. and if people get a little bit more time to be alive and themselves and just off the grind for a little bit, just a little of reduction in the noise and pressure of life, I mean,

Starting point is 00:45:10 I struggled to not view that as incredibly valuable to my fellow humans. I don't know how to measure it, but I would like to focus on both helping people work, found companies, and thrive in the economy, and boost joy if we can. I mean, you have to ask yourself, how big are the entitlement programs

Starting point is 00:45:29 we already have in the U.S. and, you know, we have welfare, we have food stamps, we have unemployment. There's a long list of things that are done already. So people get a little triggered by UBI, like, oh my God, we're giving free money,

Starting point is 00:45:44 it's a wealth transfer, all the stuff. It's like, it's already happening, folks. It's just got a lot of different names and programs on it that are incredibly expensive to run and are incredibly easy to game. So one test I would love to see a state run is to say, what are the collective value?

Starting point is 00:46:03 Okay, we got 10 million people live in this state. We are spending this amount, a billion dollars a year on these entitlement programs that one million people take advantage of. So we got a million people and they're collectively getting $10,000 a year. Let's take, you know, 100,000 of them and just give them the money on $1,000 a month. on $1,000 a month, and they can either opt to get the money or be in the entitlement program.

Starting point is 00:46:30 Oh. See what I'm saying? Because I do feel like welfare, food stamps, you know, all of these things create a game with a lot of overhead with a lot of constituents

Starting point is 00:46:42 who are now, you know, um, have incentives to grow those programs. So you got somebody who's working at, like we have in San Francisco, the homeless industry, complex, well, they're going to ask for more money each year.

Starting point is 00:46:56 They're going to spend the money they have each year. And it wants to grow. As opposed to, if you said to a group of homeless people, uh, or people without shelter, whatever, whatever the, for the correct way to say homeless, unsheltered people. Thank you. Or as we called people in the 70s junkies, uh, in this case, you know, here's the pile of money that you were going to get anyway.

Starting point is 00:47:17 Jason. I think, you're like, here's the thing. I didn't get it quite right. Is it supposed to be this? Never mind. We're going back to the 70s. We're going back to the 70s. He's just straight up.

Starting point is 00:47:25 I mean, if you shoot heroin, you're in junkie. I know, that's super valid. I'm not, I just, the wording. They called it junk, and if you enjoy junk, you were junkies. Anyway, the point is, why don't we try something like that? And, you know, every time you take a scientific, thoughtful approach to this, everybody gets triggered. The right, the left, the people who are benefit from it. And I think, of all the money we spend, I wonder how much actually gets to the person in need.

Starting point is 00:47:53 And it's probably less than 50 cents on the dollar actually ever makes it to the person. There's an amazing comedian who I'm not able to find the clip for this segment because I didn't know where I was going to get here. But he says that he was giving money to a homeless person. And someone said, oh, you know, don't give him the money. You know, he's just going to use it on drinking drugs. And the comedian said, well, what did you think I was going to spend it on? And then the person said, no, no, give us the money. And then we'll make sure they'll get it.

Starting point is 00:48:21 and we have to have like a gorgeous, you know, $90 square foot office space in Soma. And we have 17 executives here, you know, who are wearing somehow Chanel and Prada and they get $150,000 a year to work at a homeless organization. I'm like, oh. Mistake error. But what you're describing is block grants, but not in the federal money to state block grant, but in state to individual block grants. It's essentially an option to grab, not to grab, that sounds slightly rude, but just to accept the same level of care, if you will, but in a cash form versus a service form. Just like school vouchers, right? And so what these kind of concepts do is allow the individual

Starting point is 00:49:04 to have sovereignty as opposed to be part of the nanny state. So do you want to have to get in line, to get your welfare check, to get your unemployment, to get your food stamps, to get whatever it is? And I'm not saying, I don't think those. program should exist. I think it's great that we have a society with the safety net. What I'm saying is collectively, is that money efficiently deployed to the person who needs it? And so I think your idea of tying it to outcomes around your employment or your savings, etc. Like, why not give the $1,000 and say, we're putting 200 of it into your retirement account, 200 of it into, you know, the QQQ or whatever index fund,

Starting point is 00:49:47 and then you're getting 600. Now, you can't take that 200 out for your retirement until you're 55, and then you can only take out X percent per year, 5% a year for whatever. And then the other 200, you know, you can take it out 10% at a time, but you can't go below 5,000 in your account, right?

Starting point is 00:50:07 Now, that could get really interesting. So they're forced to watch $500,000, $500,000. $5,000 in their, you know, wealth-front account, e-trade account, wherever it's custodian. And now you have to watch that grow every year. And now you're part of equity. So that can just be so much more powerful in terms of educating people. And that's really what, you know, I think is the missing component here is people who are at the bottom don't understand economics. And J.D. Vance talked about this in Hillbilly, uh, elegy a whole bunch, which,

Starting point is 00:50:43 was social capital. What's social capital? It means the people around you and the circumstance you're in, that socialization educates you and provides opportunities. If you are growing up and your parents are attorneys, like the chances of you knowing how to become a lawyer are 100%. If you grew up in Appalachia or in deep in Brooklyn and you never met a lawyer in your life and your parents, you've got a single parent and they're, you know, on welfare or whatever, you don't even know how an attorney gets a degree. And J.D. Van said that in his book, which is he didn't actually know that lawyers needed to go to law school. He didn't understand what law school was before he, you know, got to Yale in other places and, you know, whatever. You can think

Starting point is 00:51:29 what you want about the person. It's a good book to read to kind of get some insight into, you know, some of the challenges poor people have. Have you read it, by the way? Oh, I own a copy. I have not read. Reading. You know, I read so much nonfiction during the day that at night I am a 100% fiction guy, and I really need to get off of that. But there's so many good books to read. It's a quick read.

Starting point is 00:51:52 It's a quick read. It's easy, breezy. It's like, and it's a great listen, by the way. It's got a great narrator on Audible. Shout out to my friends in Audible. Dude. So anyway, great job on this. All my money.

Starting point is 00:52:03 I love Audible. You know what I do is I get the platinum every year, $200. I've had it for over 10 years, so I've spent probably $3,000 on audiobook. You know what? Wait, wait, wait, one good idea in your life. One change in your life pays for the whole thing. What is,

Starting point is 00:52:15 what is platinum? There's a platinum program. I don't know if they promote it anymore, but they give you 20 credits for $200 or 25 credits for $200 and it winds up being $8 a book. But, you know, sometimes the books on Audible are 35 or 25 or 15. It's like, it's variable.

Starting point is 00:52:30 They just abstract that away. You have 20 credits. It nets out to 10 bucks a piece. And I've had years where I get through the 20 audiobooks by, you know, the summer. and I just re-up it earlier. And they actually give you the option, buy five more credits, 10 more credits.

Starting point is 00:52:45 So I just have always 20 credits sitting in there. I don't even think for a second, if I'm going to read the book. I just think, this book I might read at some point. I'm going to buy it, put it in my library, download it and have it there

Starting point is 00:52:57 for my next flight. Boom. Yep. That's my strategy. I cannot agree with that anymore. I do a slightly differently, but the same result. But you have a board meeting,

Starting point is 00:53:05 so we're going to move on to Meta's new AI model because this is, probably apart from the Google Whiz thing, this is the biggest news of the day. So everybody,

Starting point is 00:53:14 Meta has a new AI model. It is called Lama 3.1, 405B for 405 billion parameters. My brain,

Starting point is 00:53:25 it just forgot that word, mid-sentence. It was trained on 15 trillion tokens and 16,000 Nvidia H-100

Starting point is 00:53:32 GPUs, so an enormous cluster, an enormous data set. And meta says that this is the first openly available model that rivals the top AI models when it comes to

Starting point is 00:53:43 state of the art capabilities in general knowledge, steerability, math, tool use, and multilingual translation. The thing that I think matters the most here, Jason, is meta is challenging the top dogs using an open source approach. So before open source was lagging open AIs and anthropics, here we have meta going right to the top of the charts with an open source model, which I think actually very much changes the game and makes some of the closed source AI companies look a little bit expensive. Yeah, and if you are working in a large corporation in America

Starting point is 00:54:19 and we were sitting here 30 years ago, open source was like this weird thing that you didn't trust. You bought IBM, you bought Microsoft, you took the safe route. Then open source became the safe route because you weren't tied to a vendor. You weren't tied to their price gout. or their release schedule,

Starting point is 00:54:38 you could actually get in the code if you needed to and alter it. So once it became stable to use open source, and people felt comfortable doing it for their web servers, for, you know, their databases, it was like one step after the next

Starting point is 00:54:52 that corporate America, and that's what this is really about. If you are a corporation, whether it's a startup or it's, you know, G.E. You trust open source today, more, then you trust investing in a big company's solution. Now, that doesn't mean people don't still use Oracle databases.

Starting point is 00:55:14 Of course, they do at scale. The SLA is incredible. Doesn't mean people stop using Office for an open source version of Microsoft Office. That's a completely fine product, majority of corporations use it. But it is an alternative that puts a little pressure on the proprietary solutions. And that's what you're seeing in the market today. You know, inside.com, I'm working with Z.

Starting point is 00:55:37 who is our new CEO over there, and we're building a new app, and the app is related to news and community, and we're using a little AI in it, and we're constantly having this discussion. Do we want to use, you know, the API from closed AI, I'm sorry, open AI.

Starting point is 00:55:52 Do we want to use perplexities API? You know, you test everything, but then you're like, at the end of the day, you want to have one of those companies rug pull us. Now, Facebook's been the ultimate rug puller. Remember Zingda and Sings? some of these other companies, they rub pulled everybody.

Starting point is 00:56:09 Yep. So, you know, they do not allow access to their API anymore. They don't allow you to have access to the graph anymore, but they're behind. So when you're behind, you open source. When you're ahead,

Starting point is 00:56:22 you go proprietary. None of these people are honest about that. It is an absolute lie that they are pro open source or they're pro closed source. They are pro, whatever, wherever they have an individual. advantage. And the, you know, if you look at Google, Google's like Android, open source, everything's

Starting point is 00:56:42 incredibly. Like, like, can you open source your API for the algorithm and search? They're like, oh, no, no, that's a black box. We can't do that. We couldn't even if we wanted to. It's impossible. It's impossible. It's impossible for us to do that. And it's like, okay, but duck, dot go and Brave has this new awesome search engine that has, by the way, shout out to Brave. I love that browser. It's my default browser. Highly recommend Brave. Their VPN. And then, you know, know, saving yourself from all advertising and tracking. Unbelievable product. They're not currently a sponsor.

Starting point is 00:57:13 They were on the past. I've used Brave on all my devices since day one. I am absolutely militant about protecting my privacy in that regard. Maybe two, but they have a new search engine with an API. So it is possible to have a search API, by the way. And so, yeah, I think this is super significant. And what's most significant about it is that Zuck is engaged. Very engaged.

Starting point is 00:57:36 You know, he's on vacation. He just got off the beach. You can tell his hair he's got, either he's using the sea spray, salt spray for his hair. It's a new innovation since you lost your hair. And I'm barely got whatever I got left here. But I've used the salt spray. It gives you like a right off the beach look. But I have the feeling that's his beach house.

Starting point is 00:57:54 I was looking at the wood paneling. I think that's a, I think he's in Kauai right now. I don't think you have to use salt spray. If you own half the world, I'm pretty sure you just go to your private. If you own half a Kauai and you built a fence up and built a, bunker, 100 feet below ground. Yeah. I'm guessing he got off the beach and surfing,

Starting point is 00:58:12 wave surfing, whatever he's doing, to put on his gold chain and to do this podcast. That tells you everything you need to know. The dude realizes he blew 40, 50 billion on VR headsets and AR that nobody wants.

Starting point is 00:58:30 I mean, all that people want are the Raybans to take pictures of their kids. That's the beginning and end of it. So that product's DOA. Maybe in a decade, it won't be. I think in a decade if the Raybans actually work, sure, why not? Sure.

Starting point is 00:58:43 But they don't work as currently thing. And now he's just so engaged. You see it in his eyes, he's looking at this going, oh, I'm actually, I'm doing something that could grow my company and that is worth the investment now. 16,000,

Starting point is 00:58:58 I think you said it was 16,000 H-100s or something. Yes, those go 20, 40,000 a pop. Maybe put it at 30,000. It's a half billion dollars in hardware. And he's not done buying that hardware. So, you know. Compared to VR, though, AI is cheap. You know, it's a, it's a bargain if you're a matter.

Starting point is 00:59:17 A bargain. And he still owns the H-100s. By the way, it's not like you use them once and throw them away. No. He still got them. So, you know, that's 16,000 cluster will be 160,000 will be 1.6 million and continue on from there. So absolutely fantastic for everybody that this is occurring. most of all starts.

Starting point is 00:59:35 I'll bring your point and my point together. So your point about Zuck being locked in, engaged, ready to rock, super dialed in. And my point about the importance of this being open, we have a clip from Zuck talking about his, was from the podcast you're referring to. And he's talking about his response to Apple and how that led to him thinking about open source and building in a more free setting. So John, can we run that clip? I was a little shocked by how directly you called out Apple and their closed approach.

Starting point is 01:00:02 Can you kind of expand on that and where Apple has been a blocker for meta? Yeah, I think it's a little bit soul crushing when you go build features that are what you believe is good for your community. And then you're told that you can't ship them because some company wants to put you in a box so that they can better compete with you. Is he talking about himself or he's talking about Apple? I think Mark Pink has said the same thing about Zuck when they shipped Zinga and Farmville and Zingka poker. What we should do is we should take that clip again and then scroll down all the headlines of like Facebook turns off, Facebook kills, Facebook, you know. Yeah, yeah, yeah. But, I mean, there we have, Mark Zuckerberg, 30 or 40 minutes of one of the five or six tech CEOs who matters time is a non-trivial thing to get your hands on.

Starting point is 01:00:51 Yeah. And that is an impressive amount of communication and candor. So I think that Zuck, to your point about investing money in something that is working for his business, I bet you right now he's walking a couple inches taller because. Everyone's been raving about this before 3.1, 4 or 5B came out. People were talking about it. It had amazing stats. People wanted to see it in practice. And we have a chart here from meta, just showing the overall scores so people can see why this one matters.

Starting point is 01:01:20 Basically, and I'm going to adjust this down for everybody, each column here is a different model, and each row is a different test. You want to have higher scores. And as you can see, Lama 3.1, 405B competes well with what we're saying. seen from Open AI and from Anpromic, etc. So I think he's, I think after getting beat up for years over his ARVR bets, no one really wants the Metaverse, I bet you Zuck feels reinvigorated by this. And that's why he has that broccoli hairdo.

Starting point is 01:01:50 He thinks he's a zoomer. Well, and, you know, losing sucks and everybody mocking you for changing the name of your company and blowing 40 billion. Like, you know, he's a human. you want to be loved, you want to be respected. And, you know, he's never been.

Starting point is 01:02:09 He was always kind of like, no matter how much money he made, no matter how big Facebook got, everybody always thought, he's just a memmick machine. Is that the right word, mimic theory? Yeah, I know you're referred to.

Starting point is 01:02:22 Like, memmic theory, you know, the thing that T. T.L. was always in of, like, copying, and then caring about everybody around you thought, he just copied MySpace, friendster, it better, never had an original idea, bought Oculus, bought Instagram, bought WhatsApp. You know,

Starting point is 01:02:37 there's very few things that actually Zuckerberg didn't steal in the history of Facebook. It's all either acquired or stolen. There was no original idea there, right? Like, Pope? Like, there was no original idea. Even the news feed, like, he didn't pioneer it. Friend Feed did. They bought it. So people don't remember the history of this, but I think, you know, he told his team, I don't want your ideas, I want you to just copy it. And that's why the Instagram founders left, because they were like, we have our own ideas. We're pre-win.

Starting point is 01:03:06 They were just like, just do stories, kill Snapchat, right? And there was this joke in the valley that Evan Spiegel was the hardest working person at Facebook. Because he was their pre-product manager. Exactly. He was like the product manager. He was working two jobs because all they did was they stole ephemeral messaging. They sold stories.

Starting point is 01:03:28 They just kept stealing everything that he did. streaks, etc. You know, I think Zuck feels like he's not an innovative person. He's just an executor. And I think he now wants to actually put a stamp on something. And, you know, maybe he will. So good for him. You know, I hope it works out.

Starting point is 01:03:46 There is one last element of this that I think is important to bring up. We have talked about AI safety, regulatory capture, AI regulation, and then the closed versus open source AI debate. Zuck makes a very interesting point. And I read his lengthy essay announcing these models in his own. philosophy. And he says this. I'm going to just quote this. There is an ongoing debate

Starting point is 01:04:06 about the safety of open source AI models. And my view is that open source AI will be safer than the alternatives. I think governments will conclude it's in their best interest to support open source because it will make the world more prosperous and safer. So, doesn't like Gopold's close model, thinks

Starting point is 01:04:22 open AI, if you will, is the way to move forward with development. I got to say, I wonder what this does, the value of your anthropics and your open AIs, because they're not better and they're more expensive. I mean, until they launch a hosted API version of, you know, there are models and a consumer $20 version and an enterprise $20 version, that's really what Zuck needs to do or somebody needs to take that model and do it. There was a note that you couldn't use his stuff if you had

Starting point is 01:04:56 over $100 million or maybe it was $250 million users. So this open source model, you're not allowed to just take it and put it in an at-scale company. But really what has to happen is if they're going to defeat OpenAI, they need to have an app that does, it matches the OpenAI app, which, you know, I think tens of millions of people are paying for Open AI enterprise like we do at launch. And, you know, some number of consumers pay for it as well. So I don't know why that doesn't exist yet, but that would be the next question. card, I think probably by the first quarter of next year, there'll be a meta-a-I app, a standalone app, and you'll be able to access inside the apps, right, which you already has that little search bar, but look for a, you know, a chat GPT app competitor. And that's really where

Starting point is 01:05:45 I don't know why these other models don't have the focus on the fit and finish of an app, because I have been using that every day, multiple times a day. I will take out my phone and ask questions using the voice like the her mode uh scarlet johansen mode uh i'll call it from now on so i'll use scarlet jansom more trouble just to get more trouble just a bigger settlement i'm trying to get scarjo a bigger settlement shout out to my car joe um but you know the scarjo mode and um also just research mode you know you and i do the show i have chat chpt open and i was when we were doing our discussion of ubi i just said entitlement programs tell me what the big to refresh my memory, and I always just say, do citations. It gives me the citations. I feel

Starting point is 01:06:33 relatively good now reading those statistics in 4-0. But you know, you can't do that with this model. You don't have that. You know, Grock doesn't have that fit and finish yet. So, Claude, I don't even know if Claude has an app yet, but it needs to be in the app format. That is the interface. So I think look for, that's my no strachanus prediction is we'll have a competitive meta app and it will be 10 bucks a month, 99 bucks a year or ad supported in some way and that's going to be

Starting point is 01:07:03 paid a complete. And then we'll see who's, we'll know, just going to see who's better at apps. Sam Altman and OpenAI or Zuckerberg and Meta. And I think we already know the answer to that question of who can get more

Starting point is 01:07:20 users for their app. Zuckerberg releases a competitor to OpenAI. This app, I think Open AI loses 50% of their value as a company. I think they literally get cut in half. So that's my other prediction is Open AI's valuation gets cut in half. By a competitive app, Zuckerberg will copy that app pixel by pixel and then add a couple of features and he'll make it free. Did you say, there's your Nostracanus.

Starting point is 01:07:44 Nostracanus? Because I thought you mispronounced it at first. I'm like, oh, you got that one wrong. No, no, it's chasing. After I got the hot swap correct, now my friends are calling me Nostrcanus. Oh, yeah. But in fairness, I told them they had to call me no shrikanis. I was not, you know, I was not going to bring up the question of astor-turfing in regards to that comment.

Starting point is 01:08:04 I can't improve on that. So we have to go. A couple of matters for our friends. Episode 2000 is coming up. If you are subscribed to the Twist Newsletters, don't forget, we are collecting everyone's favorite episodes, moments and themes from episode 1,000 to 2000. We are also, Jason, you told me to ask you, we have a Twist 500 party coming up. Oh, yes. You know, you guys are doing such a great job on the Twist 500. Go to Twist500.com. A shout out to our friends at Coda who make a great product. They're not the official sponsors of this. We just love with our product. They have sponsored the show before, and I just think the world of their product. But if you go there, you'll start to see, you know, it's getting a little more advanced. You'll see that we now have it by stage. We have it by topic. And, you know, we're going to just keep growing this and figuring out different views. And I love to throw events. And so, you know, the show, the advertising on this week in startups once again sold out.

Starting point is 01:09:00 So we sold out the year, Alex, once again. And, you know, it's only July or whatever. So this happens to us. We run out of inventory. So my poor sales team is like, can you come up with a new product? And so I was thinking, and I'm just brainstorming here, but I'm thinking of doing a Twist 500 party, maybe doing in three cities. You know, we do like a New York, a Bay area, and in Austin, Texas. and we tell, and we do it three times a year,

Starting point is 01:09:27 anybody who works at this at a Chris 500 company will get two tickets to the event. So, you know, for the management team, let's say, top 10 employees, founders, whatever. So then I was thinking, well, what isch could you do there? And I was thinking about making it a one day just party, then I was thinking, you know,

Starting point is 01:09:46 in the early days when I did TechCrunch 50, which was the idea I came up with Mike Arrington, your old boss, or I don't know if you ever worked for him. I've actually spoken to my glarington for a total of, I think, 30 seconds my entire life. Congratulations. That's more than enough. He was nice to me in those 30 seconds. If you gave it another 90 seconds, you might have had a different outcome.

Starting point is 01:10:08 It's a joke, folks. It's a joke, folks. He's a pretty promulgional guy. Anyway, putting it aside, what I did in that event was it was a gift back to the community and I said, let's give everybody a free table. in a trade show area, and, you know, we'll just have 100 tables and we'll have 50 people on stage and everybody launch a new product. We wound up renting the tables, I think, for $1,000 for two days just to cover the cost, which was more than that. But, you know, there's a little cost to the trade show area.

Starting point is 01:10:37 So I'm thinking of two ideas. Some sort of a party, like a lawn party, food trucks, no agenda, just for the Twist 500, which is the top private technology companies in the world, defined as, because people have been able to. asking me, our definition, we're defining these as the companies we think will have the biggest return on investment. So just so we're clear, we believe these will have the biggest outcomes. In other words, if you own shares in them as an employee, as a founder, as an investor, consultant, or if you were to buy them in the secondary market, we're going to say, we believe, Alex and I that these are the ones that will get the biggest return. Now, we're not going to be right on that. You have to make your own financial decisions, but we're just going to take a stab at that,

Starting point is 01:11:19 right, of the 50,000 companies, these are the top 1%, therefore, with our editorial judgment, we think it, we have the biggest outcome. So that's what I'm thinking about. I'm open to people's ideas. And I was also thinking about maybe

Starting point is 01:11:31 renting something like the San Mateo Fairgrounds and just having three or four stages. And then all 500 companies can get a booth or like a tabletop where they show what they're working on. Now, you know, Andrel or Stripe may not care to have a little booth there. but the other 200, you know, who are up in commerce series A and below may really appreciate that.

Starting point is 01:11:52 Yeah. And then we invite, um, you know, the top 100 companies or 100 companies we think are interesting to give a 20 minute talk with a 10 minute Q&A. And so you got three stages, two days and make it just like a big festival, right? So that's my thinking right now is some sort of event to celebrate these companies and let them come together. And, um, yeah. As long as there's a big focus on letting the people just mingle and,

Starting point is 01:12:18 and connect and cross-pollinate. I'm so in. Also, I love the food in all those places, so I'm in. I will be the first person there. I'm totally into it, but we have to go, everybody. Don't forget, like and subscribe on the YouTube's. We did not get to audience cues.

Starting point is 01:12:32 That's because we ran out of time. I have a list from the team. I'm going to try to respond after the show. We appreciate you all. Goodbye.

This Week in Startups - Wiz passes on Google, Meta pushes open-source AI ahead, Spotify’s earnings, and more! | E1984

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.