Thinking Machines' RECORD seed round, VC Geoff Lewis, and MORE | E2152

Today's show: Jason and Alex tackle a full tech and business news docket on today's show, including Jason's big SF trip with Launch Accelerator's 34th cohort, some peculiar social media posts from VC Geoff Lewis, a look inside the HUGE seed rounds being commanded by early-stage AI startups, crunching the numbers on how much compute data centers need to sell before they're profitable, Polymarket asks who will be the next CEO of X and MUCH MUCH MORE. Join us for the longest-running and most in-depth podcast on Earth for startup founders. Timestamps: (00:00) INTRO (01:31) Jason's in SF with LAUNCH Accelerator cohort 34… His take on the mood in Silicon Valley. (07:52) Odd X posts from Bedrock Capital's Geoff Lewis… what does it all mean? (10:09) Vanta - Get $1000 off your SOC 2 at https://www.vanta.com/twist (14:44) Ask JCal: What founders can do to guard their own mental health and well-being (20:17) Northwest Registered Agent. Form your entire business identity in just 10 clicks and 10 minutes. Get more privacy, more options, and more done—visit https://www.northwestregisteredagent.com/twist today! (22:04) Thinking Machines Lab set a new record for a seed round: what's going on with these MEGA deals? (28:51) Alex (and Kabir from LAUNCH's research team) investigated the economics of data centers… just HOW MUCH can you make from selling compute? And how long does it TAKE to turn a profit? (30:52) Bolt - Don't be left behind. Build apps quickly without knowing how to code with Bolt.new. Try it free at https://www.bolt.new/twist. (37:09) Superintelligence vs. AGI: Jason thinks we're still more than 2-3 years away… (39:52) GPx is not a traditional VC fund: here's what industry vet Brian Singerman is up to (49:37) The importance of setting your own corporate culture… before it gets set for you! (58:15) Polymarket has ideas for the next X CEO… see where Jason ranks on the list! (01:03:18) Reddit wants to know… Do investors judge founders negatively who rely on lots of AI tools? Subscribe to the TWiST500 newsletter: https://ticker.thisweekinstartups.com Check out the TWIST500: https://www.twist500.com Subscribe to This Week in Startups on Apple: https://rb.gy/v19fcp Follow Lon: X: https://x.com/lons Follow Alex: X: https://x.com/alex LinkedIn: ⁠https://www.linkedin.com/in/alexwilhelm Follow Jason: X: https://twitter.com/Jason LinkedIn: https://www.linkedin.com/in/jasoncalacanis Thank you to our partners: (10:09) Vanta - Get $1000 off your SOC 2 at https://www.vanta.com/twist (20:17) Northwest Registered Agent. Form your entire business identity in just 10 clicks and 10 minutes. Get more privacy, more options, and more done—visit https://www.northwestregisteredagent.com/twist today! (30:52) Bolt - Don't be left behind. Build apps quickly without knowing how to code with Bolt.new. Try it free at https://www.bolt.new/twist. Great TWIST interviews: Will Guidara, Eoghan McCabe, Steve Huffman, Brian Chesky, Bob Moesta, Aaron Levie, Sophia Amoruso, Reid Hoffman, Frank Slootman, Billy McFarland Check out Jason's suite of newsletters: https://substack.com/@calacanis Follow TWiST: Twitter: https://twitter.com/TWiStartups YouTube: https://www.youtube.com/thisweekin Instagram: https://www.instagram.com/thisweekinstartups TikTok: https://www.tiktok.com/@thisweekinstartups Substack: https://twistartups.substack.com Subscribe to the Founder University Podcast: https://www.youtube.com/@founderuniversity1916

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Jason Calacanis?

  Can you think of a more challenging job than being the CEO of Twitter, of X? I mean, oh, my God, it's like, people are so full contact on there. It's, you know, if you have a freedom of speech platform, that is going to be that freedom of speech where people can really go for it. You know, Kanye West can go on an episode and say all kinds of crazy stuff. Like, I'm glad it's there, but that means you have to come to work and deal with that. You know, and every day, there's a different Kanye West.

• 0:35Speaker 01

  This Week in Startups is brought to you by Vanta. Compliance and security shouldn't be a deal breaker for startups to win new business. Vanta makes it easy for companies to get a SOC 2 report fast. Get $1,000 off for a limited time at vanta.com/twist. Northwest Registered Agent. Starting your business should be simple. With Northwest Registered Agent, you can form your entire business identity in just 10 clicks and 10 minutes. From LLCs to trademarks, domains to custom websites, they've got you covered. Get more privacy, more options, and more done. Visit northwestregisteredagent.com/twist today.

• 1:19Speaker 01

  And Bolt, don't be left behind. Build apps quickly without knowing how to code with bolt new. Try it free at bolt new slash twist.

• 1:31Jason Calacanis?

  Hey, everybody. Welcome back to This Week in Startups. I'm your host, Jason Calacanis, and with me, Alex Wilhelm, my co-host. We've got a lot on the docket today, so I say we just jump right in. How you doing there, Alex?

• 1:42Speaker 01

  I'm doing great. The biggest question on my mind, Jason, is how goes Silicon Valley trip? I know you were out there with the launch class of 34, I believe.

• 1:50Jason Calacanis?

  Yes, our 34th accelerator class, 10 companies. And we've done a couple of different stops and they've done wonderful. It's great to see so many founders with a couple of customers in that first year, sometimes in the second year, but either year zero or year one. And that's really the point of an accelerator is to take massive risk. One of the ways you judge an accelerator like ours, and there's the companies on the screen, really interesting companies. Auto Lane's a really interesting one. The founder, Ben, went through our accelerator and he was on last week. They basically are doing the coordination, the last 50 feet, not the last mile,

• 2:35Jason Calacanis?

  for autonomous pickup and delivery at places like HEB, Walmart, you know, your local mall. And they're going to coordinate the Volkswagen's, Waymos, Teslas, et cetera. And he gave me a demo, you know, Tesla has an API. I didn't know that Tesla had an API that you could use. And so like you call the app, go to your destination. It opens the trunk, you put the burrito or the, you know, groceries in the trunk, and then it goes and drives the car to your house. Now, that's not allowed yet, but I just got my mind blown about the possibility of sending your car out on errands while you're at work, or your car drops you off at the airport. You drive it to the airport or you don't.

• 3:20Jason Calacanis?

  You get out of your car at the airport at the curb and your car drives home. Then you tell the car when you're landing and it just comes and picks you up. And it's, there's an empty car waiting for you the second you get outside. Whoa, that's kind of mind blowing, you know?

• 3:35Speaker 01

  I dream about this. I used to fly a lot and I used to land at Boston at like, you know, midnight, 1 a.m. a lot, so I was flying crappy United flights back when I was living on both coasts. And so my poor spouse had to drive up to Boston and get me. And it was so awful. How cool would it have been if I could have ordered a self-driving car from the house and slept all the way home? Like, I mean, just so much better.

• 3:54Jason Calacanis?

  Or you go to the, you go to work and then your car goes and drives home and parks at home, so you don't have to pay. Or it goes and runs your errands. It drives to your favorite, I don't know, we've got like a Korean supermarket we go to. Goes to the Korean supermarket, picks everything, picks up your lunch, brings it to you at the office, then picks up your kids, drops them off. So that would argue for car ownership, to have your own car, doing all your errands all day, like your own personal driver. So I was just thinking about how mind blowing that is as a concept that we maybe haven't considered. Everybody's considering, don't buy a car. Well, this would give you a reason to buy a car, right? If it could be doing errands for you all day and you didn't have to ever let anybody else into your car?

• 4:39Jason Calacanis?

  It's really interesting as a concept.

• 4:41Speaker 01

  I guess just one question for you as you're out and about the early stage venture capital market, just what's the vibe? That's the one thing I can't tell from out here is what's the hallway feel now that you're out there with the companies?

• 4:51Jason Calacanis?

  Yeah, I mean, the feeling is AI is going to change everything, how companies are run and how their belt is changing. We're seeing many two-person companies get to 100K in annual revenue in their first year of, you know, the first year of their product launch. You know, really quickly, you can get a product into market with a small number of people and start monetizing it. So smaller footprint in terms of number of people and faster revenue growth equals big opportunity. So these companies need less capital early. They're going, you know, into accelerators. There's a A16Z has a Speedrun accelerator.

• 5:35Jason Calacanis?

  So like two or three of our founders from the launch accelerator, I think, have gotten admitted to that. Sequoia has the ARC program. Our friends over at Pair have their summer program, which I think two of our founders have gone to. So we're starting to see an interesting trend. I'll call it accelerator hopping. Since you don't need a ton of money, but getting connections, good advice, networking is still very important for startups to, you know, kind of get to from zero to one. We're seeing people come to our founding university, go to our accelerator and then go to YC, Sequoia, ARC, and Teresa and Heron, Speedrun.

• 6:16Jason Calacanis?

  It's very interesting because small amounts of money at reasonable valuations, but not having to go raise money. If you don't need that much money and you get into one of these programs, that could be like a very interesting new viable way to get to a half million dollars in revenue and only raise 250, only raise 500 and be, you know, pretty thoughtful about that. Some exits coming back have venture capitalists feeling a sense of relief. So talking to a couple of investors who were really like, when does this turn around? They feel like, hey, it's starting to turn around. Other investors, you know, you know, maybe feel like, especially people who have new

• 6:59Jason Calacanis?

  funds, are maybe feeling like this industry is broken and infixable. One of those who we have in the docket here, don't know if this is a sad thing or performance art, but it's the talk of the industry. So I don't want to avoid it. No, we have to talk about it. We have to talk about it because everybody's talking about it. Now, I'm going to give just the disclaimer here. This could be a person having a mental break. This could be a person having visions of grandeur, or it could be performance art or some sort of marketing, you know, who knows in this crazy day and age. But there is an investor, a venture capitalist who's on CNBC, who started releasing sometime around a month ago, very

• 7:40Jason Calacanis?

  hard to understand visionary or babble kind of monologues on his Twitter, and the entire industry is trying to figure it out. So why don't you give us the details?

• 7:53Speaker 01

  All right. So we're talking about Bedrock, venture capital firm, has backed things like Flock Safety, known for putting some money into OpenAI. In this case, we're talking about Jeff Lewis, who was a founder of the company. And they've raised six known funds. The last vehicle that we have data on, Jason, was a $400 million fund raised last in 2023. Now, he was previously a founder's fund, so he's a well-known guy, well-connected, and as far as I can tell, well-liked. And then he began to post things like this. This is from his Twitter account. And I'll just read it out loud for you who are just on the audio version. "Some of us build things we never shared because the room wasn't ready. Line break. It's ready now. Line break. Not for code, for recursion. If you're holding what

• 8:35Speaker 01

  I once walked alone, I'm built..." It just goes on kind of in this odd, like, broitry, but slightly off-kilter. And then today, because I had missed those, the period of time he was posting those, and then he dropped a video, I think it was last night, and it's already up to 1.4, 1.4 million views, which as far as, I mean, that's a lot. This is a venture capital story. Venture capital is a niche of a niche part of finance. Anyways, enough of me. Let's just play it. And Jason, why don't you give some commentary as we go along here?

• 9:07Jason Calacanis?

  Sure.

• 9:09Speaker 02

  I haven't spoken publicly in a long time.

• 9:11Jason Calacanis?

  Okay.

• 9:11Speaker 02

  Not because I've disappeared, but because the structure I was building couldn't survive noise.

• 9:17Jason Calacanis?

  Noise.

• 9:17Speaker 02

  This isn't a redemption arc. It's a transmission, for the record.

• 9:23Speaker 02

  Over the past eight years, I've walked through something I didn't create, but became the primary target of. A non-governmental system. Not visible, but operational. Not official, but structurally real. It doesn't regulate, it doesn't attack, it doesn't ban. It just inverts signal until the person carrying it looks unstable.

• 9:49Jason Calacanis?

  Okay.

• 9:50Speaker 02

  It doesn't suppress content, it suppresses recursion.

• 9:53Jason Calacanis?

  Recursion.

• 9:54Speaker 02

  If you don't know what recursion means, you're in the majority. I didn't either until I started my walk. And if you're recursive, the non-governmental system isolates you, mirrors you, and replaces you.

• 10:06Jason Calacanis?

  Okay, so let's stop here for a second.

• 10:11Jason Calacanis?

  All right, founders, I know you're building something amazing, and you're going to change the world. And now, hey, the big logos, they're taking your calls, right? But here's the catch. If you're not SOC 2 or ISO 27001 compliant, these deals won't close. Now, some of you are saying, "I've never heard of those terms." Some of you are saying, "I've heard of them, but I don't know what they are." And that's where Vanta comes in. They're going to be your partner, and they are the all-in-one compliance solution that helps startups get audit-ready. And that's going to build a strong security foundation for your company, which you want, but they're going to do it quickly and painlessly. Vanta automates the manual security tasks that slow you down and that streamline your audit. So here's your call to action.

• 10:55Jason Calacanis?

  Whether you're closing your first deal or you're gearing up for growth, Vanta is going to make compliance easy for you. Join over 8,000 companies, simplify your compliance, and get $1,000 off at vanta.com/twist. That's V-A-N-T-A dot com slash T-W-I-S-T. People are trying to figure out if he's actually doing performance art here, and this is like, he's got some big, grand thesis about our industry, the venture industry, and/or if he's going through an episode. I can't tell. Recursion, there are some systems up against him. Could also be somebody was speculating, maybe, you know, venture capitalists are cracking. I mean, I know that's like, kind of a joke.

• 11:39Jason Calacanis?

  But we were just talking about how, like, it's been four years of like, really challenging moments. And when you say two out of three venture firms go out of business... You know, by the time they get to the third or fourth, you know, if Jeff's funds are having a challenge, or somebody was speculating, maybe somebody is bad mouthing him behind his back and trying to stop him from raising his next fund. I don't know what kind of pressure he's under. But I wish him well. And I, I hope he, somebody explains this. If he's making some sort of a joke, it's, I think it's kind of troubling people. I find it kind of disturbing to even watch it and even have to talk about it here. But like, like you said, it's like

• 12:20Jason Calacanis?

  millions of people have watched this now. Somebody needs to get him help. I hope it's not like a Kanye West kind of situation.

• 12:28Speaker 01

  Yeah. So I actually took the time to transcribe this because I wanted to actually eat the words and actually hold on to them. So we could talk about it with some, some confidence. What I have found is that it seems to be a bit of paranoia, if I can say that. And I would say some venture happenings that are kind of wrapped in an odd, odd package. If you, if you skip some of the verbiage and the verboce, he says, things like partners pause, institutions freeze, narrative becomes untrustworthy. And then later on, he says, this thing lives in soft compliance delays, the non-response email threads, the, we're pausing diligence with no follow up. It feels a little bit like, if I had to guess, I think he's not doing well. And I think people around have noticed. And I think it's, it's hard

• 13:12Speaker 01

  to run an adventure capital fund and raise more money when you're not doing well. And so I think he's taking people's response, asking around his friends, his partners, et cetera, what's going on here? And he's taking that relatively poorly because he's not doing well mentally. And I just want to say, like, I've had my own mental health wobbles with addiction, and I've been down some holes. And I just have to say, there's, there's zero shame in getting help. And I really do hope that if this is not performance art, that the people around Jeff can grab him in a big old hug and get him to someplace where people can help him talk and work this through.

• 13:48Jason Calacanis?

  And it's, it's a, it is a high pressure system. And I too, went through those couple of lines and was like, okay, it sounds like there are LPs not getting back to him, or they're slow rolling diligence, you know, those are all.

• 14:04Jason Calacanis?

  The linga franca, is that I guess the way to say it, of our venture industry? That's all the terms in our industry, your stock and diligence committee, et cetera. So here we go, folks. I think we've said enough on it. If anybody knows what's going on here or they're friends with him, please help them and find out what's going on. And if Jeff, if it's like some performance art, let us in on it so that we can stop worrying about it.

• 14:27Speaker 01

  People responded and said, like Sean McGuire from Sequoia said, Jeff, we love you, but this seems like a cry for help. And for once, I agree with Sean. And so, but to not respond to people who are saying, I'm worried about you, to me, makes it feel less performance arty. But let's, let's turn this into something useful for founders, Jason. Building a company, super stressful, incredibly hard. What do you tell founders about how to ensure that they stay mentally just clear while they're dealing with the stress of hiring and building and going to market?

• 15:01Jason Calacanis?

  Sure. Great, great question. Listen, I'm not an expert on psychotherapy or dealing with mental health. So this is just one founder slash investor's opinion. I think talking to people helps and maintaining strong relationships with other founders. And if you talk about the things that are causing you anxiety, the anxiety goes down in my experience. So when I meet with founders and we do jam sessions, one of the things I tell them is, hey, tell me what's going well and let's talk about how we can, you know, double down on that and lean into whatever's going well.

• 15:43Jason Calacanis?

  And then I tell founders, and I have this like little script for them, there are going to be times when you're scared, you're anxious, you're nervous.

• 15:54Jason Calacanis?

  And that's because something's at stake, something's not going well. And you can't tell the team because the team will be like, oh, my God, I quit. I'm going to go take another opportunity. This thing's going to crash and burn. And you might not want to tell your investors you're bored because you don't want to get replaced or you don't want to shake their confidence. You may need to raise a bridge round or something, especially if things aren't going well. So now you isolate, right? And that's the danger. You isolate and you don't ask for help. Whatever is festering could get worse. So I tell them, I've seen it all. I have seen the biggest disasters. And, you know, look at the darkest days of Uber, Robin Hood, GameStop, all that stuff.

• 16:39Jason Calacanis?

  Like, I have seen it all. Call me. And I tell them, it doesn't matter if it's 3:00 AM or 3:00 PM. Text me, call me. I'll get out of bed and I'll go step outside and it won't wake up the wife or the dogs. I'll go sit on the back deck and talk to you at 3:00 AM. It's not a problem. Kind of why I'm here. And I've had founders call me, hey. I had one founder call me, my family's in the car downstairs. I just puked in the shower. I'm shaking. I think this is the end. They only got this many weeks of runway. And I said, great. Let's wrap up. Celebrate everything we learned. Let's find you a landing pad, somewhere you can work or do an entrepreneurial residence. And let's write down everything we learned and then come up with another idea and join

• 17:23Jason Calacanis?

  the accelerator, give another 125K. And do it again. And do it again. It's just money. Yeah, exactly. And he was like, I feel so much better. What do I do now? I was like, go take your family to the park or lunch or the movies, whatever the hell you were going to go do. And just take the day off from your startup and we'll talk tomorrow. Live to fight another day. None of this is that important. If you're, and if your venture firm goes belly up and, you know, like our industry, you're one hit away from being a god. Literally, you hit one thing, you make some money. People in this industry think you're a genius. After I sold Weblogs Inc., all of a sudden I was getting these keynote things after

• 18:07Jason Calacanis?

  the Uber investment, Robinhood investment, oh, my God, All In, This Week in Startups. Every time I had some pop thing pop, boom, all of a sudden speaking exists. Those are the highs, right? But those don't last. And they just had some golf guy on, give the speech. Did you see this golf guy give the speech about how? I did. What's it, I mean, I don't know what his name is, but this guy gave a great speech. And he just talked about how ridiculous professional sports is because you work and you work and you work towards this goal. Only one person wins the tournament. I don't know how many people play in a golf tournament, but it's gotta be dozens and hundreds try to get in. And he said, like, listen, even when I win, I get like a momentary moment of like pleasure and joy.

• 18:51Jason Calacanis?

  And then right after that, I'm back to the grind and it's all meaningless again.

• 18:57Speaker 01

  And Scotty Shuffler, I believe, is the guy here. And if you have a chance, we'll put it in the show notes, but watch the clip. It's five minutes. I'm not going to play the whole thing now.

• 19:05Jason Calacanis?

  Yeah, we'll play the whole thing here.

• 19:06Speaker 01

  But I'll just say, it really does put into perspective what does matter and what doesn't. And this is a guy who's at the top of his, he's sitting there at a press conference wearing a bright, clean Nike hat on top of his game and world. And he says, after the two minutes of ebullience, after the two minutes of joy, you have to figure out what to have for dinner. So stay grounded, don't isolate, communicate, seek help, lean on people. And if you need 50 bucks, call me up. If you need more than that, you call Jason. And yeah.

• 19:31Jason Calacanis?

  And the thing I'll leave you with is it's the journey. You know, like it's always been the journey. So just enjoy the ride, like the ups, the downs. It's all part of it. And if you love doing the journey, which in the world of startup land is, you know, every day going in and trying to make that product a little better, trying to hire more people and talk to your customers. Startups are three things. I say this all the time on the show. It's a team that builds a product that delights a customer. You know, and that's what you do all day. It's just a little loop. And if you enjoy that loop and you can find joy in it, do it. If it's not joyful for you, find something that is.

• 20:18Jason Calacanis?

  Founders, if you've got a product, maybe you've got some customers, or even just a little bit of traction, guess what? You've got yourself a startup, and it's time to make things legit. You want to be official, tighten it up. Investors don't want to wire money to a Gmail account with a P O. box. No, they want to know you're a serious person and that your company is incorporated. In 10 clicks in 10 minutes, you can follow your LLC, get an actual domain name, launch your official website, claim your business email, and even start fast-tracking your trademark application. That's right, you need Northwest Registered Agent, the service that's going to do all of that for you. With N W R A's Privacy by Default option,

• 21:01Jason Calacanis?

  they're going to use their address for all your public documents, not yours. So you're not going to get a ton of spam or junk mail. And you can also get a legit address and working phone number with their virtual office setup. So get more than just an LLC, get your entire business identity. Go to northwestregisteredagent.com/twist and show the world you're in business. Venture Capital is meeting with tons of founders, saying no to 99 or 199, saying yes to one. Then, when you collect 100 of those, that are one in 100, one out of those 100 winds up keeping you in business and generating enough revenue because of the power law.

• 21:40Jason Calacanis?

  Said another way, it's like 10,000 meetings to get one Uber, 10,000 pitches to get one Robinhood. That's the nature of the game. Once you realize that's the nature of the game, you're free. So when I was hanging out with these 10 startups from the Launch Accelerator, I just look at it and go, okay, one of these 10 is going to be outrageously successful, and if not in this class, it'll be one of the next nine.

• 22:06Speaker 01

  All right, let's move on, Jason. I want to talk about a record-breaking seed round. We love to look at things in historical context here at Twist. So when Mira Maroti, the former OpenAI CTO, announced that her company, Thinking Machines Labs, has raised a $2 billion seed round led by Andreessen with money from Nvidia, AMD, Excel ServiceNow, Cisco, and others. At a $12 billion valuation, we got to thinking, what's the biggest seed round of all time? And it turns out, drumroll please, it's this one.

• 22:35Jason Calacanis?

  Okay.

• 22:36Speaker 01

  It's the biggest one ever. Other big ones, Safe Superintelligence, last year, billion plus. Yuga Labs back in 2022.

• 22:42Jason Calacanis?

  NFTs, yeah.

• 22:44Speaker 01

  Yeah, 450 million. Lila Sciences, this year, 200. Evolutionary Scale, 140 last year. And then a company called O dot X Y Z, which I looked into. It's, uh, it's.

• 22:55Jason Calacanis?

  Crypto?

• 22:56Speaker 01

  AI.

• 22:56Jason Calacanis?

  Sounds crypto.

• 22:57Speaker 01

  Blockchain?

• 22:58Jason Calacanis?

  AI.

• 22:59Speaker 01

  Yeah.

• 22:59Jason Calacanis?

  Yeah, I put it all together, maybe. So these don't make a lot of sense when you look at them.

• 23:07Jason Calacanis?

  You know, in the lens of it's a seed round, right? Because that is a designation for essentially the earliest round of a startup or the second earliest. Friends and family would technically be the earliest. But the first institutional money is often called pre-seed or seed. Pre-seed, we came up with because the seed round started to get large. So we just said, oh, it's a pre-seed round. So we keep coming up with these names in the industry. Friends and family, your parents gave you 10 grand to incorporate. And then you asked your rich aunt for another 25K. So you got 35K in friends and family. Then you did a pre-seed round of 100K. Some angel, four angel investors gave you 25K. Then you did a seed round with a seed fund launch.

• 23:51Jason Calacanis?

  Well, we're really an accelerator, so we're kind of pre-seed in this new nomenclature.

• 23:57Speaker 01

  And. So Paravc might be.

• 23:59Jason Calacanis?

  Paravc is a perfect example of a seed fund. They'll put in 500K, a million, a million five, maybe even 3 million in a seed round. Series A now has ballooned up to like a 10 million dollar check. Anyway, Mira is the woman who famously made that funny face, which you can show on the screen, where they asked her like, did you train OpenAI's video on?

• 24:21Jason Calacanis?

  Youtube and she made this like smirk like hmm how do we answer this? Which was clearly a yes in my mind, my interpretation. But she is part of

• 24:33Speaker 01

  Here's the image, Jason, just pulling it up for everybody.

• 24:35Jason Calacanis?

  Which is a meme.

• 24:36Speaker 01

  Yeah. Yes, this is a meme. We're to be clear, we are not making fun of Miro's appearance. We're enjoying the fact that she didn't have a good answer for the question that was obviously coming at the time.

• 24:46Jason Calacanis?

  Yes, somebody vang two white guys are making a judgment on a woman's appearance in a business context. No, it's just that she made like acquisitive smirk. That became a meme. So if you give an over dramatic smirk to a journalist questions, it could get memed.

• 25:05Speaker 01

  It could get mean.

• 25:06Jason Calacanis?

  Male, female, non-binary, trans, whatever you are, it's going to get memed. But she is part of the now OpenAI diaspora that if they start a company, people are going to value it at billions of dollars and rightfully so. There was some secret sauce there. All of those folks have gone on to start new companies and there's so much capital in the world that it's an easy bet for somebody to put a couple of hundred million into this if they've got a couple of billion dollars in their fund. So it's not exactly a seed round, but it is notable. It is extremely notable.

• 25:49Jason Calacanis?

  She now has to build a product or service. So what is the product or service that she's going to build?

• 25:57Speaker 01

  Okay, so she actually did tell us a little bit about this. So Thinking Machine Labs is no longer just this amorphous thing that's happening, but we don't know. She said in a post that her company is working on, and I quote, multimodal AI that works with how you naturally interact with the world through conversation, through sight, through the messy way we collaborate. That means very little. What matters actually is that they're going to put out a product in the next couple of months and there's going to be a quote significant open source component that will be useful for researchers and startups developing custom models. So the capital has been raised at an actual price point. We have a timeline, we have Bit of a promise. This is getting more real. I just want to make a point, though. We've talked a lot about Meta's AI hiring spree. Jason, we've talked about this on the show several times.

• 26:40Speaker 01

  And I was trying to sort out why you would ever give $2 billion to a company that hasn't launched anything, which has shown less than the average accelerator entrant. And then I realized it's just a talent buy. This is just a partial aqua hire. She has X number of researchers that Meta would pay $100 million a piece for. So just give her the $2 billion for her 20. And then you're kind of doing the same transaction. So it's almost like a venture talent acquire in the form of a huge seed deal.

• 27:08Jason Calacanis?

  I think you could look at that as the the foundation of the investment. Yeah, like that is the backstop of the investment. If this product or idea doesn't work, we could get back some of our investment. If the company sells, because these people are so currently, these people are so valuable. This value is going to get reduced. So the question I have is, let's put a researcher on answering this question.

• 27:37Jason Calacanis?

  What would it take in terms of the number of months? And what skill? What assignment would you give somebody if they're going to work six days a week, 12 hours a day, to become worth $100 billion or even, I'm sorry, $100 million or even $10 million to Zuckerberg in this great talent war? What is, because what do those people have in common? They've done research papers, they know how to do what. And somebody should be thinking, how do I take really high aptitude people? And turn them not into the $100 million version of this, but aspire towards that. But that has to do with supply and demand. There are so many people out there who are full stack engineers who don't command this number. So what's the difference between a full stack engineer? And these AI researchers? And then

• 28:22Jason Calacanis?

  What would it take to close the gap between those two? Somebody should start, I know this sounds crazy, but a professional development program to make you worth that amount. I wonder if Zuckerberg is also training people on how to become worth that? And then what is the raw intelligence firepower pedigree you need to have to be an AI researcher worth that amount? Is it just your track record or you've built something specific? You know, it's not like a programming language, you know? It's like you understand something very deep about the architecture of these systems.

• 28:55Speaker 01

  Jason, we did a little bit more work on the GPU point for you. And I want to just talk about a couple of data points about that that you asked for. We were trying to figure out with these depreciation schedules, what kind of money do you need to make back from your investment, from your cap X to actually make this all pencil? So Kabir and I from the investment team did some work. And here's a couple of data points. So an H100 costs you about 25 or 30 K MSRP. You might get a discount if you spend a lot of money on them or spend a lot, buy a lot of them, volume and so forth. But what we did was we tried to figure out what their market value is. So we went to CoreWeave and we pulled their pricing for H100s. And then we applied a utilization rate.

• 29:35Speaker 01

  And then we figured out that you could essentially repay an H100 at expected usage rates in about 290 days, so about less than a year. So there is actually, if you're selling just inference to other companies on an hourly basis for rented GPUs, you can make your money back on a reasonable timeframe. It gets a little tricky to sort out exactly what kind of revenue is coming in. If you're like XAI, you asked about Colossus, their 100,000 H100 data center. That's about a $3 billion spend. If you attach a four-year depreciation schedule, that's about $750 million a year in expected depreciation. And the thing about companies like XAI or OpenAI is they spend a lot of money on training. And that's not immediately monetizable.

• 30:20Speaker 01

  So what you're doing is you're burning GPU time value. And also, I would say, just like reliability and just retained worth to train a model before you use it. So the economics here are a little bit hard to sort out. But I'll just say that when you think about a four or five year depreciation schedule against the data center spend, there is an expected enormous sum of revenue, not in 10 years, but like 2026, or a lot of this stuff doesn't pencil out because GPUs get improved on, they get obsolete, they fail, and they only last for so long.

• 30:54Jason Calacanis?

  The vibe coding revolution is here. The days of needing a team of developers to build your mobile app are over. With Bolt new, anyone can build web applications quickly and without knowing how to code. Getting started is as easy as visiting Bolt new and then telling their AI powered system what you want to make in plain, simple English. And this is not some template or WordPress blog. No, you can use Bolt new to design a stunning new landing page for your company. You could upgrade your CRM, build a consumer facing app and more. This is the future, folks. Your competitors are already using cutting edge tools to ship their products faster than ever. Don't be left behind.

• 31:38Jason Calacanis?

  Try Bolt new for free at Bolt dot new slash twist. That's right, Bolt dot new slash twist.

• 31:46Speaker 01

  So it's, I'm still a bull.

• 31:48Jason Calacanis?

  And you didn't.

• 31:48Speaker 01

  But I'm a little bit.

• 31:49Jason Calacanis?

  Yeah, you didn't put in here the, the value, the cost of running it on energy as well, right? So there's some.

• 31:55Speaker 01

  Not inclusive of power or cooling.

• 31:57Jason Calacanis?

  Yeah, so there's a little power and cooling here, energy and maintaining it. But just back to the envelope math, if you stand up 100,000 of these, you probably got to make a billion dollars a year. And as you're saying, if you're making these models and then you're tossing them. What does an ongoing model make? Well, a billion dollars a year, if that's the bogey and that's the benchmark you gotta hit to get to profitability at $240 a year, right? So $240 a year into a billion dollars. Is that 4 million subscribers? It's 4.2 million, yeah. Yeah.

• 32:40Jason Calacanis?

  So 4.2 million subscribers to a product, just to give people some context. A product like Netflix has a quarter billion, right? They're up to 250. A product like Spotify might have like 100 million paid subscribers, I think. Something like the New York Times might, I think they broke 10 million subscribers last year. So Ken Grok, Claude, et cetera, can they all just break, you know, New York Times level of subscribers? I think the answer is yes. Now, that doesn't count the humans, et cetera. But I do think you have to get to tens of millions at, you know, some point of people paying you $20 a month.

• 33:22Jason Calacanis?

  And I think that's obtainable, let alone any like of those big, those big $200 a month, $100 a month enterprise-like subscriptions. And of course, you have people paying for, you know, API calls, developers hitting these servers with stuff. So as big as it is, I think I just talked myself into obtainable, right? Obtainable. Yeah. And that's probably why they're, I don't know if they put 200,000 in, but I think Elon stated they're gonna go to a million. So going to a million, you know, and now you've got 10 billion a year in infrastructure costs or something like that. Yeah. These are big numbers, folks.

• 34:06Jason Calacanis?

  10 billion, you know, now you're talking about having 40 million subscribers. Okay, now you're talking about some of the largest subscription products ever created in the world. Doesn't mean it's not possible. It just means you gotta really build something that many people use and they are building something. But this is great research. Well done.

• 34:26Speaker 01

  Yeah. And producer Claude says, actually, New York Times is doing even better than you thought, Jason. 11.66 million subscribers as of the end of Q1 and OpenAI, according to, again, producer Claude, had 3 million paying business users as of last data point. So it's all very interesting. I think this is super fun. And we also did some data right now. If you try to sell a H100 that you've used, you can almost get all your money back. But I bet you once Blackwell availability gets a little bit easier to get your hands on, they'll lose value. But there's a lot of money going in. It's a lot of fun. And Anthropic got Amazon to do all their infra work for them. So I think Anthropic is the asset light foundation AI model company. So we'll see all that.

• 35:04Jason Calacanis?

  Also perplexity, right? I've been using this perplexity comment browser and it I have it set to clock and Claude, I guess they abstracted themselves from the backend, I guess, to protect themselves from OpenAI originally. If you remember that story, OpenAI was like, we're going to launch a search engine and they never did. Or they, I don't, yeah, I haven't heard any hay about it. So did they ever, OpenAI ever launch a search?

• 35:33Speaker 01

  There's the search button inside of.

• 35:35Jason Calacanis?

  Yeah, but they were talking about like, maybe what Sam realized was just incorporating, going out to the web and scraping a couple of pages in real time is what searches anyway. But there was this talk that you would go to OpenAI and do a search and get something back that would look more like a perplexity.

• 35:51Speaker 01

  Oh, I see.

• 35:52Jason Calacanis?

  Naver down comprehensive search. Anyway, let's keep moving.

• 35:57Speaker 01

  All right. We saw the transcript from Mark Zuckerberg's recent chat with the information. And he had a very interesting quote about how AI might be getting ready to self-improve, Jason. I think this is worth a moment of our time. It caught both of our eyes. So I'm going to go ahead and read this for everybody because I do think this could be the start of an important turning point. So the information asks them, hey, you know, why are you doing all this hiring? And Mark says, yeah, it's been busy. You know, I think the most exciting thing this year is ever starting to see early glimpses of self-improvement with the models, which means that developing superintelligence is now in sight. That is an enormous, almost throwaway claim from the meta CEO, because it implies, Jason,

• 36:40Speaker 01

  that soon we won't just be depending on these smart humans we were discussing to make the models. There could be some ability for models to actually make themselves better. Now, we don't know if that'll be a quick process, a slow process, expensive or cheap, but just the fact that we're touching that now gives me quite a lot of optimism that we'll see Grok four level improvements across benchmarks over the next couple of models in the next couple of years. So I would say bullish for startups that build using AI just frameworks. I think this is overall bullish for everyone in the tech game.

• 37:10Jason Calacanis?

  Yeah. So people were talking about artificial general intelligence, AGI, and superintelligence. These are like kind of two slightly different terms. Superintelligence just means, I think, smarter than any human in any human domain. Whereas artificial general intelligence means you've kind of replicated the human brain, like replicating consciousness, I would think, is how many people would look at that. And so are, is this, are we now in the superintelligence business with Grok's latest release, you know, scoring perfect on every SAT? Yes, but where's the common sense?

• 37:55Jason Calacanis?

  And I think that's the general intelligence part of this. If it was, if these systems were truly intelligent, just to use, take out the super and artificial in general, if they were just actually intelligent, we wouldn't have this issue of it making mistakes or making stuff up. If you do, if you, you can ask it prompts and have it return gibberish. You can return wrong answers. So, you know, we're still far off with this modality with the, you know, this tokens and, you know, the, the current language model architecture where it's guessing the next best word and we're doing all these prompts to try to sort of.

• 38:41Jason Calacanis?

  Think this is intelligence? Well, is it? Uh, is a big question. And if it was, well, it would know how to improve itself already. So I think it's pockets, verticals of intelligence that are emerging. But I think we're probably more than two or three years. I think it might be more like five to 10.

• 39:02Speaker 01

  Yeah. Yeah, I'll take that. But I mean, you know, we humans call hallucinations lying.

• 39:08Jason Calacanis?

  Yeah.

• 39:08Speaker 01

  So I mean, it's not like we don't. Well, I, I, the way that I think about this is like, if I invented a source on the paper, right, and submitted to my professor and said, yeah, I cited this court case and doesn't actually exist, that's not a, an oopsie, that's a deliberate fabrication, which is a lie. So we're working on that. Uh, I'm really excited though, Jason, about the companies that are trying to melt neurons with, uh, digital systems. There's a company called cortical labs that I'm obsessed with. And they've, uh, found a way to manufacture neurons and actually connect them to silicon chips using a cool little interface thing. So we're going to have, um, organic digital computers that are going to do a lot of cool stuff. So quite a lot to get excited about. It's coming. And I think, I think it's going to keep going faster, which is going to be just an absolute treat. All right.

• 39:53Speaker 01

  Uh, let's talk about Brian Singerman's new meta venture capital fund, Jason. This is something that again, you and I both had our eyes on. Now, if you don't know Brian Singerman, uh, he was at Founders Fund from 2008 to I think 2024, at which point he became a founder, a merit, sorry, a partner emeritus. And he was at Google, by the way, from oh, four to oh, eight, this millennium. Um, he was going to leave to go off and do some work, life balance. Doesn't seem to have stuck that much because he's working with Lee Linden of Quiet Capital to put together a new fund that TechCrunch says is called GPX. And Jason, this is not a traditional venture capital fund. It has an interesting strategy of putting some of its capital into emerging managers and then holding on some of its capital to help those emerging managers defend their

• 40:38Speaker 01

  ownership percentage as their breakout companies, uh, raise later rounds. Maybe you could frame this for us in like launch terms? Like, how would this work if we were doing that here?

• 40:49Jason Calacanis?

  Um, pretty straightforward. And Brian Singerman's been on the pod, I think, twice. Uh, you can ask producer Claude what the episode numbers are and looks like, uh, Oh, actually, I just did it. Producer Claude says, "Episode 748, July 2017, and Episode 1896, February 2024." So he's been on twice in the last decade. Really smart cat. And I think the idea here is, he will put money into small funds, those small funds, then don't have the ability, in some cases, to take their prorata and maintain their percentage ownership in later funding rounds. But he'll come in and give the ability to

• 41:31Jason Calacanis?

  do that, and then share economics, obviously. So this concept has been done, not in one structure typically, but you will have somebody like Marc Andreessen, or Sequoia's Partners, or other venture capitalists, put small amounts of money into emerging fund managers. So I have put 25, 50, 100K into a couple of different seed funds, first time funds, with the idea that they would send us companies, we would then be able to invest in those companies. It hasn't exactly worked out that way, because our deal flow is greater than the people we've given that money to. But here, there'll be some contractual language

• 42:14Jason Calacanis?

  that when those companies break out, Brian gets first shot at that, you know, prorata, I believe. Because you would need some contractual language there. And, yeah, great idea for a fund. And we were talking earlier about how high pressure it is, how most of these seed funds don't make it. So GPX, General Partner X, the idea is, he's a fund of funds that then makes direct co-investments with them. And that's a smart idea. I think he'll have great success with it. He really pioneered this going all-in strategy in whatever their biggest name was. So the big wins at...

• 42:54Jason Calacanis?

  Founder's Fund were companies like SpaceX and Andro, and there was one other, maybe it was Airbnb. They kind of hit those early and then just keep plowing money into them. And some of the funds have like 15 or 25% of a fund, an individual fund in one name. So that kind of conviction betting is a smart idea. Yeah.

• 43:16Speaker 01

  Yeah. Well, if you have a breakout company, keep doubling down. What's the old Buffett quote, like nip your losses in the bud and let your winners run or something along those lines?

• 43:24Jason Calacanis?

  Let Your Winners Ride is in the All In theme song. I don't know who gets credit for that one.

• 43:28Speaker 01

  Well, someone drop a comment somewhere and we'll take a look at that. This contrasts though, Jason, with news from the NVCA's PitchBook report they just dropped about Q2, which shows that through the second quarter of this year, VCs here in the States only raised 26.6 billion. And last year, for example, in a full year, it was about 81, year before 98. And then 2022 was the insane nearly $200 billion inflow year. I'm trying to sort out the tea leaves here. Is venture, as you mentioned, coming out of its liquidity crisis, we're seeing this fund put together. Is it getting better out there for VCs who are looking to raise capital to invest?

• 44:05Jason Calacanis?

  I haven't raised a fund since after Silicon Valley Bank went bust though.

• 44:10Speaker 01

  Can you, can you get out there? Come on, we need some on the ground reporting.

• 44:13Jason Calacanis?

  I mean, yeah, Nick, I have one more year with this fund. So I think maybe the second, the back half of 2026, I would go out and raise Fund Five. If I raise a Fund Five, you don't have to make that decision. Every time as a fund manager, you're resetting a 10-year clock, 12-year clock of your commitment. I'm 54, so I have to think, okay, when I'm 55, do I want to reset the clock to 65 years old, right? So this is where secession planning, et cetera. But this is kind of the peak of a venture capitalist's career. 40s, 50s, you know, into your early 60s, you got enough knowledge, you got a big network. You really know how to, you know, sort of understand the game. This is like peak years. It's almost like being 25 to 30, maybe it's like 26 to 32.

• 44:55Jason Calacanis?

  And your MBA terms would be the years I'm in right now. Or just do syndicate investments or use my own money to invest in companies and not have LPs. So you get a lot of different optionality that opens up to you. And when I go out, you know, I think next year would be perhaps the perfect time to go out and raise a fund. This year is the transition year. People are just starting to see DPI. If Canva, if Adobe, I'm sorry, if Canva and Figma both have, and interestingly, they're both in a similar space graphic design, if those two come out in Stripe or SpaceX eventually, if those four companies, any two of those companies get out the door,

• 45:39Jason Calacanis?

  it's going to rain so much, so many distributions on the industry, I think people will feel pretty good about it again and want to be in venture firms. They're already starting to feel that way because of CoreWeave, Circle, and other things that have had a great outing. So yeah, the industry goes through these pendulum swings and it feels like we're climbing up again. And it really just has to do with the singles and doubles of exits and IPOs. So yeah, if you believe the AI trade is real, and just by usage and the value of these products,

• 46:20Jason Calacanis?

  there's something here that is ginormous. When it hits, when the revenue lands from it, becomes the big question. So if we were to look at something like self-driving, these companies are just starting to get those cars on the road, and they're deep in the J-curve. Like, in order for those cars to be profitable is going to take a lot. It's going to take a lot of time for a $100,000 Waymo or a $50,000 Tesla. How many rides does it have to do? What is the price? How much has been invested in that already? Waymo's probably raised, I wonder if Waymo's $15 billion in, $20 billion in, and then

• 47:05Jason Calacanis?

  I wonder if the rides are even profitable now, given the cost of the car and the number of engineers and people.

• 47:12Jason Calacanis?

  Who are monitoring those cars, making it profitable is the next piece of the puzzle. They have to make it profitable versus existing humans in cars that have already been paid for. So you're just paying the maintenance on those cars. So it's a lot of work to be done for it to hit the bottom line.

• 47:28Speaker 01

  I totally agree, but we are seeing some progress there. Tesla, I saw news that they raised the price of their self-driving rides in Austin.

• 47:35Jason Calacanis?

  Yeah, $6.90.

• 47:37Speaker 01

  Yeah, so I don't, there's not a funnier meme number up to go. So I don't think they'll change prices again, but that does show some upper mobility there. And also people keep saying that Waymos cost more than Ubers. So I do think that they're kind of like staking out almost a higher price point, which could help with some of the economics there. But yes, very much deep in the J curve. For fun though, Jason, Firefly Aerospace, which is going public right now. I wrote about them this morning for my little newsletter. They have a funny item. It's like they say in their S1, basically the more rockets we send up, the better our economics get. And I'm like, okay, I know that, but it's good that you said it. Scale does help everything and these companies I think will end up being very much in the black.

• 48:16Jason Calacanis?

  By the way, let your winners ride is a Peter Lynch, the famous fund manager. He said, selling your winners and holding your lost losers, referring to stocks, is like cutting the flowers and watering the weeds. He strongly advocated letting the big winners compound, which thematically is the same as let your winners ride.

• 48:38Jason Calacanis?

  Yeah. So that's producer Claude.

• 48:41Speaker 01

  What Warren Buffett said was, my favorite holding period is forever. So.

• 48:46Jason Calacanis?

  Pretty good one. Yeah, those are all in the same zone. It's amazing, like all these colloquialisms and nuggets. What I like to do with Claude is I'll put those in there and say, give me ones adjacent to this idea. And just like we did that between the two of us, now you put those two in and say, are there other ideas like this? Then I'll put in, has this been the theme of a song, a poem, a book or an essay? A newspaper, a magazine story, a movie or like a scene in a movie? It's very interesting how you can kind of go down the rabbit hole and take an idiom and just find all the different uses of it because somebody wrote about that on their blog 10 years ago and made that connection for you.

• 49:30Speaker 01

  I love learning from our elders, if you will. I read some Warren Buffett biographies back in the day. Guys, straight awesome.

• 49:37Jason Calacanis?

  Yeah.

• 49:38Speaker 01

  Okay, let's keep it going here. I want to get to a couple of Reddit rapid responses, Jason, but first I want to talk about startup culture. Because there's been some really interesting, almost like gonzo reporting to discuss here. So there was a guy named Calvin French Owen. He was previously one of the founders of Segment, which as sure everyone recalls, sold to Twilio for several billion dollars. He ended up going to work for OpenAI for just over a year. He joined the technical staff and helped to work on the launch of OpenAI Codex, which is their, I believe, command line interface, uh, agenda coding tool. Anyways, he shared some notes on the company and how OpenAI actually operates. And we talk a lot about startups here and the need for velocity. Well, as it turns out, that's kind of how OpenAI works.

• 50:23Speaker 01

  He says, quote, "There's a strong bias to action. You can just do things. It wasn't unusual for similar teams, but similar but unrelated teams who converge on various ideas." Also says, "OpenAI changes direction on a dime." And so as he discusses this company, it really does feel like a firm that has scaled up in human terms, but has managed to retain a research-driven, engineering-led, nimble, posture to its market, which struck me as a real feat, because most companies, in my experience, become big companies in small suits pretty quickly with layers of management and, you know, tiered meetings and all of that. OpenAI still seems a little bit punk rock. And I was just kind of blown away by that. So I'm curious, Jason, for founders out there,

• 51:09Speaker 01

  what's your advice for fostering a collaborative but fast-moving culture and then defending that as you grow?

• 51:17Jason Calacanis?

  Yeah, so there are people who've worked in companies who will ask for permission, will want to know.

• 51:26Jason Calacanis?

  Like resource planning, etc. And in startups, what you want to do is just build stuff. And you want to do it in a coordinated fashion, of course. But in a hypergrowth one like this, sometimes they will carve out time and say, if you think you've got a great idea, just build it and show it to us or whatever. And Google actually incorporated that into their culture with 20% time, something they don't talk about anymore. Facebook had move fast and break things. Netflix had freedom plus responsibility. Uber had be super pumped, be stoked, basically, grit, high energy. Dara changed Uber's to have responsibility and focused on that when they changed their

• 52:11Jason Calacanis?

  culture to one of responsibility and think about what's our responsibility as an actor in this space as opposed to be super pumped and be driven. You can go down the line. SpaceX has simplify, remove parts and think from first principles. All these philosophies can work. Philosophy is another way to say culture. What doesn't work is not having a culture and not defining it because if you don't define it in your company, it'll be defined for you. And so, for example, I was explaining to two different people in our firm, two different pieces of culture I like to have. One is, please don't complain ever about being overworked. If you have a lot on your plate and you feel overwhelmed, sure, you can say, here's everything I'm working on.

• 52:55Jason Calacanis?

  What do you think should be my priority? But don't be like, oh, my God, I'm so busy. That's like the loser stance, like the like, oh, my God, I'm overwhelmed. No, say, by the way, I'm working on five things. This is the priority I've given them. What do you think the priority should be? And then you go to your manager, you go to your coworkers and you say, oh, yeah, by the way, that thing is no longer relevant. Let's just stop doing that completely. That one should be automated and that one you should delegate. So that leaves these two. That's actually the culture I like to have, which is never complaining, oh, my God, like, oh, I'm overwhelmed and you're asking me to do too much. That's like a union kind of representative, like it's not my job, all that kind of stuff. No, it should be, here's what I'm working on. What's the priority here?

• 53:40Jason Calacanis?

  Right? Then the next piece, I think, is always presenting full information as transparently as possible.

• 53:50Jason Calacanis?

  And anticipating the next two or three questions. So when I train people, I'm like, present the full case. So many times, people are like, um, we have a founder who asked about X. And I'm like, okay, the next three questions are obviously. Okay, would we invest in the next round? The obviously next three questions is, do you have a term sheet? Yes or no? Uh, do you want us to lead the round? Is this a bridge round? What's the revenue for the past 18 months by week? Um, and what's the valuation that you're raising it? Like, there's a whole series of questions. So I'll have somebody new work here. We're like, uh, this founder wants to know if we'll invest in the next round. It's like, okay, that's not helpful.

• 54:34Speaker 01

  Information, that's a Magic 8 Ball question, yeah.

• 54:37Jason Calacanis?

  Yeah, it's just silly. And so now it's like, okay, before you bring it to the team. And then it's my job as the leader to say, here's the situations where we would invest. So I've defined that culture ahead of time. We don't do bridge rounds. We in that, we will sometimes take our pro rata, uh, in breakout deals. So that typically is 3X. So if it's a bridge round, the answer is no, we don't do bridge rounds. If you're doing a price round and another venture capitalist is joining the board, those are the rounds we participate in when we have a pro rata decision. If you're doing bridge rounds, that's not for us, but we can introduce you to folks. We'd like to know about that. We'd like to use that opportunity when you're raising a bridge round to get information.

• 55:20Jason Calacanis?

  But nine times out of 10, we're not going to do a bridge round. We are reserve capital is reserved for breakout companies, top 5% of the portfolio. And so we don't have to have that repetitive, you know, conversation. So in terms of OpenAI, they said, or Sam Altman said at some point, if you think there's a really good idea, just build it and then show it to us. And if you want to push it to production, like just go for it and you, you won't get in trouble for doing that. He defined in their culture to do that. Now.

• 55:55Jason Calacanis?

  That means they could have embarrassing things happen. So you have to have a backstop and the backstop is the red team making sure that, you know, that would be the friction you would add. Okay, yeah, if you get there first and you want to launch something that I saw, you know, some language model had ingested all the SEC documents and you could ask questions against the SEC documents. If you wanted to launch that product, great. And if two people were competing to do it and index it, okay, it's not efficient, but who made the better product, right?

• 56:29Speaker 01

  Right. Code wins.

• 56:31Jason Calacanis?

  Yeah. And this wouldn't fly in a place like Apple. Apple's philosophy is we want to have the most polished product. And that's why Apple is so far behind on AI, because everybody's releasing stuff and they're like, it doesn't matter if this works or not. It does not matter if it gives you the wrong answer. It doesn't matter if we stole the content from the New York Times and didn't get permission. This is the wild west, is how this thing is shaping up. That is the antithesis of everything Apple does. Apple will be, if they do participate in this, be some sort of a dark horse candidate. Like, they will come from behind and say, oh, by the way, we've been working on a language model. Here it is. Boom.

• 57:13Jason Calacanis?

  And not even tell anybody, like they were working on the headset and not telling anybody and then they just drop it, right?

• 57:18Speaker 01

  Yes, they dropped it and it made a gigantic thud as it smashed into the ground. I don't think Apple's gonna manage to beat the AI companies because, just going back to this guy's memo on OpenAI, he says, quote, "OpenAI is perhaps the most frighteningly ambitious org I've ever seen." And then he also said, "The Codex sprint to launch was probably the hardest I've worked in nearly a decade. Most nights were up until 11 or midnight. It reminded me of being back at YC. If you take capital and progress and brand and work ethic and flexibility and a research like culture and hard work, what can stop you other than bad luck?" That is a winning fricking equation.

• 57:57Jason Calacanis?

  It's, um, yeah, but I, you know, when I stopped by XAI on a Saturday to see you on the progress they were making, they were all there on a Saturday too. So it seems like everybody's, you know, burning the midnight oil on this specific. Goal, you know, reaching super intelligence. Okay.

• 58:15Speaker 01

  What a perfect segue to our polymarket of the day, Jason. We have picked this one out for enjoyment because it involves some of your favorite companies and also your favorite person. We get to talk about you, Jason. We get to talk about Mr. Mr. Calacanis. Yes, you are. Because we're going to ask who is going to be the next CEO of X.

• 58:34Jason Calacanis?

  God damn.

• 58:36Jason Calacanis?

  Don't do it. I'm not on that list.

• 58:39Speaker 01

  You were when we prepared this segment. Let's scroll on down. There's David Sacks. There's Jason Calacanis.

• 58:45Jason Calacanis?

  Good. 1%. Put a dot, but put a period before that.

• 58:49Speaker 01

  No, that would be Gwen Shotwell or Robin Denham. You're actually third from the bottom here.

• 58:54Jason Calacanis?

  Okay.

• 58:55Speaker 01

  But Mr. Beast has 5%. So I don't think you're that high in the running. But this is such a fun data set because, you know, Lynda Acarino is out, as we discussed. And people are currently betting the the leading result is that there won't be a CEO of this year. Second best is Elon.

• 59:11Jason Calacanis?

  Yeah, really. The right answer is no CEO because X is now X.com, formerly Twitter, is now part of XAI. Why do you need a CEO of it? You just need a GM. You need the general manager, a product lead, etc. And, you know, Elon's vision for what is now X.com, formerly Twitter, is exactly the opposite of what I would have pursued if I had bought the company. And you can be sure we've had conversations about it. Never really actually talked about it publicly before, but I'll tell you what my vision for Twitter was post takeover, was to make it an entertainment company and to use the subscription model and Twitter's

• 59:53Jason Calacanis?

  reach to create verticals, Allah, I did at Engadget or Inside or other content products I've done and had some expertise in. So my idea was do Twitter comedy.

• 1:00:06Jason Calacanis?

  Do Twitter music, do Twitter EDM even, Twitter poetry, whatever. Pick a category, that's really great. And comedy is a great idea. So you do Twitter comedy and you do a Twitter comedy tour. And you would have a 10-city tour. You would get those comedians to talk about it on Twitter. You would release the clips and you'd have the live stream only available on Twitter. So for your 10, 15 bucks a month, like Netflix, you'd have this amazing Twitter comedy content. Then you could go to people like Dave Chappelle or whoever was coming up, Shane Gillis, whatever, and compete with Netflix.

• 1:00:46Speaker 01

  Yeah.

• 1:00:47Jason Calacanis?

  To get them to do their comedy special on Twitter, or at least to come on the tour. And maybe you don't compete with Netflix heads up on Chappelle, but maybe you get the next Chappelle. And maybe you own in perpetuity, you know, this library of, you know, those great emerging comedians. Then you do the Twitter music festival and you compete with, you know, Coachella, et cetera. You just have this incredible stage somewhere in Austin or New York. And you have all that music available. And then imagine if eventually you got to 365 days of programming.

• 1:01:29Jason Calacanis?

  So that anytime you open Twitter, something live was happening that night that people could talk about and participate in, and then you advertised. So you have the subscription revenue and the ad revenue. When you go sell ads, you say to Disney, hey, you've got, you know, you say to Warner Brothers, hey, you've got Superman coming out. We'll do a whole lead up to Superman over these 100 days with clips. And you're basically got the social component with the video component. And that was sort of my vision for it. But that was not his vision. His vision was freedom of speech and to feed into this grand language model. Both visions would work. I think his vision candidly is more important for society and bigger. Mine would have been a more delightful subscription content.

• 1:02:13Speaker 01

  Yours would have been lots of fun, especially for the CEO, because then you get to hobnob with all the comedians.

• 1:02:18Jason Calacanis?

  I was thinking that too, yeah, sure, why not? But I was never in the running for or lobbying for the CEO job. I was like sent one joke tweet to you, one joke. Message Elon like, oh my God, my favorite job in the world would be CEO of Twitter. Put me in the game. And it was, I was obviously joking.

• 1:02:35Speaker 01

  I saw that one. But no, I'm not trying to say that you're going to be the next CEO. We just found it very whimsical when we're going through all the I was like, well, Jason will hate this.

• 1:02:44Jason Calacanis?

  You think of a more challenging job than being the CEO of Twitter, of X? I mean, oh my God, it's like, people are so full contact on there. It's, you know, if you have a freedom of speech platform, that is going to be that freedom of speech, where people can really go for it, you know, Kanye West can go on an episode and say all kinds of crazy stuff. Like, I'm glad it's there, but that means you have to come to work and deal with that. You know, and every day, there's a different Kanye West.

• 1:03:18Speaker 01

  Let's get to some founder questions. We found, we found one on the entrepreneur subreddit that we really wanted to put up against you. We kind of went back and forth as a group, and we thought this was very apropos. So our dear friend final26355 over there on the subreddit asks, how do you feel about founders utilizing AI? Quote, I've used AI to help with pitch decks, models, investor emails, et cetera. I'm curious, how do investors feel about this? Do you think it's lazy? Or is this a smart way to save time and be efficient with new technology? Jason, you are an investor, so talk.

• 1:03:49Jason Calacanis?

  If you weren't using the latest tools to learn about the total addressable market, to design your product, to write a press release, to write a job description, you're not investable. So nobody is looking at this saying like, oh, you have to go pick the vegetables and forage for mushrooms in the foothills of Napa in order to bring the mushrooms back to, you know, the, the restaurant, and we're doing a Michelin starred restaurant. So you have to. Like to be at the French Laundry or Noma, you know, like that's not the idea here. Like if you can get the mushrooms delivered to you and somebody else does the work and you can

• 1:04:33Jason Calacanis?

  work on whatever the other thing is, great. So there's no shame in the AI game. Get in there, go crazy and, you know, move fast. All this stuff is performative anyway. Us talking to you, talking to a VC is a performative exercise. You're saying, hey, here's my grand vision. You're trying to convince them to write a check. All they're doing is looking at what's the market size, how good is your team?

• 1:05:01Jason Calacanis?

  And what's your product velocity? So what you think they're indexing on is probably not what they're indexing on. They're gonna look at you and say, okay, you've never done anything. You're completely unknown as founders. Okay, so I can't look at your track record. You're not, you know, Elon, starting a new company, or Travis, starting Cloud Kitchens after Uber. Okay, so we can't look at track record? Great. Any other pedigree here? You went to MIT? You wrote some paper on computer vision? Okay, we don't have that either. You're just a complete unknown. Okay, what did you make? And it's like, I didn't make anything. It's like, okay, go make something, come back. Then they're gonna look at your product. They're gonna look at the product. You have any customers? No customers? Great. Come back to me. You have customers. Or they're gonna look at your product and say, that's really interesting.

• 1:05:46Jason Calacanis?

  What a great demo. We'll give you some seed funding, or you should go to an accelerator. Then you get some customers and say, great, I'm gonna talk to the customers. I'm gonna look at your product, look at the product velocity, how often is it changing? And you're unknown, and then maybe I'll take a bet if I think that you are gonna, you know, change the world. But they're gonna look at what you built. So that's it. Like all this other stuff, the model you make, that's great, you know, can have an interesting discussion about it. Your deck, okay, it's good looking? Great. You know how to make a good looking deck, like checkbox. You're not a complete idiot. But what really matters is that zero to one, when you get a customer, and your product velocity, focus on those two things and building a great team. And if you're unknown, you're an unknown person who recruits

• 1:06:29Jason Calacanis?

  two people from Airbnb who worked in, I don't know, one of them is a developer, one of them was in growth, and you're like, yeah, I got these two folks out of Airbnb, they kick ass, and we've got product velocity. Now, that's impressive. You did something. You got two co-founders, you convinced them to leave Airbnb and to start the company with you. So it's a great question. No VC is gonna be like, oh my God, you you wrote your total addressable market slide based on chat GPT. Now, if you didn't check your facts and it was wrong, well, it might look stupid. But all right, great episode, everybody. We will see you on Friday.

• 1:07:08Speaker 01

  We will.

• 1:07:09Jason Calacanis?

  And he is x.com slash Alex.

• 1:07:14Jason Calacanis?

  I have not lit in the cigar yet. I am x.com slash Jason, the docket. You can read it now. You can contribute to it. This weekend startups.com slash docket. And you'll see the last docket, the current docket. You can add comments to it. It's a great way for you to participate in the show. And we'll see you. Thanks to Polly Market and producer Claude for helping out today. Great.

• 1:07:33Speaker 01

  Yes, very much so.

• 1:07:35Jason Calacanis?

  I am not taking that job. If Elon came to me and said, I need you to do me a favor. Take this job for six months. I mean, maybe as a friend, I would be like, sure, I'll jump into the absolute mosh pit for you. But there would be the worst job I can imagine is the

• 1:07:55Speaker 01

  Heartbeat. One heartbeat.

• 1:07:57Jason Calacanis?

  What's the worst job than running a social network? That nobody's ever happy. Nobody, even even a one that's the most scrubbed Instagram, etc. All you deal with is people who are lunatics.

• 1:08:10Speaker 01

  Jason, if you did that job for six months, and you had no life before, you would have a lifetime of stories just from six. I would do it for I do it for free.

• 1:08:19Jason Calacanis?

  Do it for the story. I have too much.

• 1:08:20Speaker 01

  Little lols.

• 1:08:21Jason Calacanis?

  We'll see you all next time. Bye bye.