E125: SpaceX launch, Fox News settlement, "Zombie-corn" exodus to AI, late-stage implosion

1. 0:00 – 28:58

   Antonio Gracias and Gavin Baker join to discuss SpaceX's Starship launch

   1. JCJason Calacanis0:00

      Hey, everybody. Welcome to Episode 125 of the All-In Podcast. On a historic day, we're taping 4/20. Really excited that SpaceX was able to launch Starship, and it made it off the launch pad. An incredibly successful today... day today. We have, of course, with us the Rain Man, David Sachs, the Sultan of Science, David Freiberg, and of course, the Dictator, Chamath Palihapitiya. But two special guests are here.

   2. DFDavid Friedberg0:28

      Special guests.

   3. JCJason Calacanis0:29

      Gavin Baker from Etreides. How do you pronounce it?

   4. GBGavin Baker0:32

      Etreides.

   5. JCJason Calacanis0:33

      Etreides.

   6. DFDavid Friedberg0:34

      House of Etreides.

   7. JCJason Calacanis0:35

      House of Etreides, if you know Dune. And then SpaceX board member, Antonio Gracias, one of the first investors in SpaceX. Antonio, a big day for you. Maybe you could just tell the audience what happened today, why that is so important in the history of this company.

   8. AGAntonio Gracias0:54

      Well, first, I wanna thank you guys for letting me, uh, come on and have a little chat with you about this. Today was extraordinarily important for, for SpaceX, I think for America, and for humanity. And the Starship is the realization of, of the vision had- Elon had 20 years ago, 25 years ago, even as a child, really, to, to go to Mars. And the engineers here at, at SpaceX, and on the entire team working extremely hard, are really just to get this, this, uh, vehicle off the pad, as, as Elon had said, right? The, the... This is a brand new vehicle. Everything about it is new. So the engines, the material science, the structure, the design, all of it new. And the most important thing here was to get off the pad so we can collect data. And this technology platform is the platform that'll allow us to go to Mars. So from a non-engineer's standpoint, why this is important is that what we've proven with this flight, we got past a, um, point called Max-Q, which is the point at which the vehicle takes maximum stress. That's how I think about it. I'm sure engineers would tell you there's a, a lot more, uh, description to it, or David Freiberg give you a better description to it. But it's the most amount of stress in the vehicle, which means that this vehicle will get to orbit. And this is the vehicle that's gonna take us to Mars. So today is the day that all of the hardworking people at SpaceX accomplished a, a goal of making the human race space-faring. When we look back in history, I believe this will be the day when we mark the, the technological development that we broke through and built a vehicle that could actually go to Mars.

   9. JCJason Calacanis2:24

      Now, when we look at it, uh, obviously it didn't make it to orbit. Maybe you can give some context into what is the typical life cycle of a new rocket ship coming out. The Falcon, the original one, has done, I think, 224 missions, 222 of them successful. I think 160 or so actually landed themselves.

   10. DFDavid Friedberg2:45

       Yes, yes.

   11. JCJason Calacanis2:46

       And so you had two or three mulligans, I think, in the development of that. Maybe two, actually. So what can we expect here? When are they gonna stack and rack and launch the next one? Antonio, what's the, what's the timeline here to getting to orbit? What, what, what would we expect versus some of the other projects that we've seen, like the Russian rockets?

   12. AGAntonio Gracias3:04

       So look, this is, um, a brand new vehicle. And whenever we develop a brand new vehicle, it takes a long time in development. You know, my understanding, all this is, uh, as, as, again, a layman and sort of as a, a board member, not an executive here, is that it'll take at least, call it maybe two or three months, two or three months to really get the pad rebuilt and get another vehicle back on for testing. Uh, maybe longer. But it's really important to note here that we've gotten sort of used to the idea that SpaceX launches rockets and all these rockets come back and all those vehicles are stable because the Falcon 9 and Heavy are so stable, and they're so well-engineered, and they're amazing vehicles. They're pr- the most reliable vehicles on Earth in human history. This is a brand new vehicle. This was a huge win. I mean, it was an enormous win for the company.

   13. JCJason Calacanis3:45

       Mm-hmm.

   14. AGAntonio Gracias3:45

       An enormous win for the country, just getting it off the pad and collecting the data. And now we know it works. We just have to get it stable now and get up to orbit. So it's, it's really... It's, it's a hard problem, but it's a solvable problem from here. And the, the... we learned here is that this vehicle does work.

   15. JCJason Calacanis4:03

       Amazing.

   16. DFDavid Friedberg4:03

       Can you guys talk about, like, the impact of this vehicle, cost to launch, payload, like the big metrics that are... that kind of help realize that outcome? Gavin, I think I saw you did a bunch of really good tweets on this. You shared some of the metrics that, that I, that I, that I thought were, were really succinct and really helpful.

   17. GBGavin Baker4:20

       Yeah, sure. So when this is... I think it's a long road to full reusability. Um, you know, the first step will be Mechazilla catching the booster. Doesn't, it doesn't have, uh, legs like the Falcon 9. And then the second step will be landing the Starship, which is really hard. But once you do that, they should be able to send over 100 metric tons to orbit at a variable cost of under two million dollars per public data. This, these are public statements.

   18. AGAntonio Gracias4:53

       I would never confirm or deny those statements because again-

   19. DFDavid Friedberg4:55

       (laughs)

   20. GBGavin Baker4:55

       ... this is Gavin's, this is Gavin's math off, off the back of the envelope.

   21. DFDavid Friedberg4:57

       Two million dollars is the cost to get 100 tons into orbit. That's, that's the, the metric?

   22. AGAntonio Gracias5:02

       Variable cost.

   23. DFDavid Friedberg5:03

       Variable cost, right.

   24. AGAntonio Gracias5:04

       So I, I think the point Gavin is making, if, if I might just-

   25. GBGavin Baker5:07

       Please.

   26. AGAntonio Gracias5:07

       ... add is that it is a step function change. It's not like a small change. It's an enormous change.

   27. DFDavid Friedberg5:12

       Right. Can you compare that to the numbers before for folks to understand?

   28. GBGavin Baker5:16

       The Falcon 9 mass to useful orbit is 17 metric tons. And the variable cost that, you know, I, I have seen Elon tweet about is somewhere around 15 million dollars. So you are lifting more than five times the mass to orbit. And based on other statements, that 100 metric tons is a very conservative estimate. And you're doing it at, call it, 10% to 15% of the cost. So this is a, you know... We can all do the math, but, look, we can envelope it. Um, you know, roughly a 50X-

   29. AGAntonio Gracias5:58

       Huge.

   30. GBGavin Baker5:59

       ... change.
2. 28:58 – 40:38

   Fox News settles with Dominion Voting Systems, will pay $787M

   1. JCJason Calacanis28:58

      We can talk about the Fox settlement with Dominion. That was something you and I talked about last week, Sax. I'd love to get your opinion on that. They paid $787 million in this settlement, uh, for defamation. It didn't go to trial. Uh, you were hoping it would go to trial. How do you feel a- I mean, and- and this is an extraordinarily large settlement, so... Yeah. But what are your, what are your thoughts on this? You wanted to see this go to the man, go to the Supreme Court and maybe see the laws in the United States change eventually. Um, catch us up on your reaction to this ginormous fine. It's a big speeding ticket, yeah? Yeah. Well, it's funny you- you call it a speeding ticket, 'cause when I saw this, it reminded me of a scene at the beginning of Apocalypse Now, where Martin Sheen says that, "Charging a man with murder in this place is like handing out speeding tickets at the Indianapolis 500." I mean- (laughs) ... the analogy here is that the media is so dishonest- (laughs) ... whether it's CNN or MSNBC or The New York Times, I mean-

   2. CPChamath Palihapitiya29:55

      Or Fox.

   3. JCJason Calacanis29:55

      ... they're constantly inaccurate or- or whatever, and you know, so... (laughs) Still defending Fox.

   4. CPChamath Palihapitiya30:00

      (laughs)

   5. JCJason Calacanis30:01

      So for this one network to get a fine of like 800 million, it's like pretty- pretty incredible. Um, I mean, they should be handing out a lot more of these, in my- in my view, not just to Fox. But yeah, look, I would like to see... To that end, I would like to see the standard in New York Times versus Sullivan revised by the Supreme Court. The standard is actual malice, so you have to prove not just that the press told a lie, but that there was malice behind it. And that's what this trial would have been about and, um, Fox settled. Well, it should have changed too.

   6. DSDavid Sacks30:31

      Yeah. I think that-

   7. JCJason Calacanis30:33

      What would you, what would you change to? What would be a better standard in your mind?

   8. DSDavid Sacks30:37

      I think that if the media makes a mistake, they should have to correct it. And I would say the correction needs to be at the same level that they publicize the original story. So, if they make a mistake, if it's untruthful and it damages someone's reputation, and they refuse to post a correction, then I think they should be liable. That seems fair to me.

   9. JCJason Calacanis30:59

      If you were to put something on page A1, if you were to give it five minutes at the start of a show on whether it's Rachel Maddow or Tucker, whoever makes the mistake, could be anybody, if they made the mistake in the first five minutes top of the show, they don't get to bury it on the website. They don't get to bury it on page 7. They gotta say it upfront, "Hey, listen. We made a mistake."

   10. DSDavid Sacks31:18

       Yeah.

   11. JCJason Calacanis31:18

       "Here's the mistake. We're correcting it."

   12. DSDavid Sacks31:19

       "Here's the correction." Yeah.

   13. JCJason Calacanis31:20

       I think that's reasonable.

   14. DSDavid Sacks31:20

       Correction should get same level of publicity as the original story. And by the way, if they correct it, that would be like a safe harbor. But if they refuse, and they publish a lie, and it damages somebody, then they should be liable for that.

   15. JCJason Calacanis31:33

       I think there should also be some room around-

   16. DSDavid Sacks31:34

       That's what I think, but...

   17. JCJason Calacanis31:35

       Yeah, I'm in agreement with that because they, there is the trick in journalism to bury the correction. And, and there's another trick that I think needs to be looked at, and it's a little more nuanced which is, uh, somebody makes a mistake or an accusation in a, in a, in a publication, and then the next publication says, "Oh, The New York Times, FOX News, CNN said this." But they don't check it themselves. So they're using another publication as like a proxy to kinda give them some level of protection, and they should have to on a firsthand basis.

   18. DSDavid Sacks32:07

       Right. No, that's the game they play. You're right. So they start with a super s- shady source.

   19. JCJason Calacanis32:13

       Yeah.

   20. DSDavid Sacks32:13

       That, you know, it just attributes it to some sort of anonymous source. Then the second, uh, most shady publication quotes that one, and then the third most shady quotes that, and then it kinda goes-

   21. JCJason Calacanis32:24

       They just let-

   22. DSDavid Sacks32:24

       ... through the whole food chain.

   23. JCJason Calacanis32:26

       Yeah.

   24. DSDavid Sacks32:26

       So you're right. Reposting something in the echo chamber because other publications are doing it, you're right, that should not be protected. They should have to do their own sourcing, basically.

   25. JCJason Calacanis32:35

       Yeah, or just some base level of fact checking if you... And now if, if you called it FOX Opinion, that's slightly different than calling it FOX News, so maybe a part of this is branding up. Chamath or Freeburg, do you have any thoughts on this? Just in general?

   26. CPChamath Palihapitiya32:48

       FOX I think has only four billion in cash, so they just spent 20 some odd percent of it paying this off. And if I were a shareholder, the question I would ask is, "Did we actually make enough money to justify having to pay almost $800 million of our cash balance?" The answer is probably no. And this is the first of a bunch of lawsuits that they have. The next one, which is I think is like Smartomatic, which is another voting machine company. That's an even bigger lawsuit actually. That's a $2.5 billion lawsuit. And so, uh, it's... Could it be the same outcome in another settlement? And so now, all of a sudden, you would deplete half their cash, all for a lie? To do what? At some point, some smart business governance needs to kick in over at FOX and they need to realize that this stuff just doesn't make economic sense. Maybe they thought it made political and ratings sense, but you can't justify that when it costs $2 billion.

   27. JCJason Calacanis33:43

       There is a concept here, I think, Sax, that, uh, is particularly interesting. In Finland, when you get a speeding ticket, speaking of speeding tickets, your speeding ticket is proportional to your net worth. And so, (laughs) these NHL players who like to speed and a Nokia executive was given the equivalent, and this is a bit excessive, of a $103,000 fine for going 45 in a 30 zone on his motorcycle, and an NHL player got a $39,000 fine two years earlier. I think that this is part of the problem is sometimes fines don't match the crime and/or the, they're not proportional enough for the person to feel them. And so then we kinda can joke that they're speeding tickets. One very rich person said to me, you know, when they would, they, I'd watch them park, you know, incredibly illegally in a small town and I said, uh, "You're gonna get a pretty serious ticket. That's a, it's a pretty gnarly spot to be parking." And they said, "Oh, you mean the VIP parking charge?" And I was like, "Yeah, that kinda sucks, but okay." So there, there might be some concept here of, um, a proportional fine, and I think the EU is starting to work on that as well, because, you know, Big Tech was ignoring...

   28. CPChamath Palihapitiya34:54

       This is gonna embolden a lot of people to take matters into their own hands and sue some of these media companies if the lie is egregious enough and it negatively impacts them enough. I think it would embolden a lot of folks. So, Sax is right that up until now most folks haven't done anything about this, but if you feel like the media has really wronged you in a meaningful way, there's probably also now a lot of private equity organizations that would do the lawsuit finance or hedge funds that would do the lawsuit finance-

   29. JCJason Calacanis35:26

       Yeah. Like Peter Thiel did.

   30. CPChamath Palihapitiya35:28

       And so now you're... Yeah, it's like you're free rolling it, right? So you, you get one of these folks to pay it. They get 50% of the gains, you get 25%, the lawyers get 25%, and you go and you litigate. I'm sure that you'll see more, not less, because this fine is, it's really big. And again, we don't know the end of all of these 2020 election hoax shenanigans because this is the first, not the last of these.
3. 40:38 – 53:49

   AI update: Reddit to start charging for training data, how the next startup bubble will materialize, navigating the AI "dust storm"

   1. JCJason Calacanis40:38

      I think, some of the stuff we're seeing with AI. We had talked on this show before many times about the corpus of data under which these models are being built. Well, Reddit announced plans to start charging companies that use its data to train AI models. The co-founder, Steve Huffman, who came back after Condé Nast, Nast had bought Reddit and then, um, sold it back to the founders, and paradoxically, I believe Sam Altman has a major investment in Reddit. He said, "More than any other place on the internet, Reddit is a home for authentic conversations. There's a lot of stuff on the site that you'd only ever say in therapy, AA, or never at all." A lot of people use pseudonyms, obviously. And the Reddit corpus of data is incredibly valuable, but we don't need to give all that value to some of the largest companies in the world for free. Crawling Reddit, generating revenue, and not returning any of the value to our users is something we have a problem with. It's a good time for us to tighten things up. Your thoughts, Chamath, on what we talked about, two different episodes. Maybe we'll play, drop a clip in here, Nick, if you want to, in post. It's gonna be the large datasets. Quora, Yelp, the App Store reviews, Amazon's reviews. So there are large corpuses of data that you would need. Like, Craigslist has famously never allowed anybody to scrape Craigslist. The amount of data inside Craigslist as but one example of a dataset would be extraordinary to build ChatGPT on. ChatGPT is not allowed to, because you, as you brought up robots.txt last week, there's gonna need to be an ai.txt. Are you allowed to use my dataset in AI and under... And how will I be compensated for it? Will the rights to the data, will s- will Google just say to Quora, "Hey, we'll give you a billion dollars a year for this dataset if you don't-"

   2. DFDavid Friedberg42:19

      They should.

   3. JCJason Calacanis42:19

      "... give it to anybody else."

   4. DFDavid Friedberg42:20

      They should.

   5. JCJason Calacanis42:21

      Chamath, those were our previous discussions that we just played. What do you think?

   6. DFDavid Friedberg42:24

      It's so incredible. We are witnessing such an important moment for Silicon Valley, but frankly, how the world works, and it's just, everything is changing. That's what I'm just in awe about. That, you know, we talked about this, and we were basically spitballing something two or three months ago.

   7. CPChamath Palihapitiya42:45

      ... and not but 60 or 90 days later these things come to pass. Right? We talk about something in one week and then, you know, 14 days later, it's completely upended, like how impressed Saxe was about plug-ins, and then plug-ins were rendered useless and somewhat impotent two weeks later by AutoGPTs. It's just so profound, I think, what's going on. So Google to- uh, today announced that they're going to merge two organizations that I thought were so orthogonal to each other.

   8. JCJason Calacanis43:18

      Disparate.

   9. CPChamath Palihapitiya43:19

      Brain and DeepMind. The cultures just seemed so totally different. But now they're merging those two things together, something that I thought would never happen. They did it. So all these competitive pressures are so real. I was in LA for the Breakthrough Prize, I was flying home with somebody on Saturday. I won't say who it is, but they are right at the bleeding edge of a lot of this AI stuff, and they let go a third of their company, replaced it with an agent, within six weeks of training it. So how is this not gonna affect everybody else, I guess is maybe the bigger question. And then I go back to what I said last week, which is that we've talked about a lot of the positive things and I think it's important to make sure that people understand that there are a bunch of non-trivial negative things. And I think I shared one on Twitter, which was around this company that used an AI model to build a library of 40,000 toxic compounds that could kill all kinds of numbers of humans. So there's all kinds of really, really tough things going on right now that I think, to me, means it's the moment where I have the least sense of how to do my job. And so I've tried to kind of, like, put a pin in everything and just go back to learning mode.

   10. JCJason Calacanis44:34

       It's a bit humbling is what you're saying. Like-

   11. CPChamath Palihapitiya44:36

       It's incredibly humbling.

   12. JCJason Calacanis44:37

       ... the entire rule book, even for us as capital allocators, company formation, i- i- it has you, it has you, Chamath, wondering-

   13. CPChamath Palihapitiya44:44

       But even, even as a, as a-

   14. JCJason Calacanis44:45

       ... how to do this.

   15. CPChamath Palihapitiya44:46

       As a CEO, I'm like, should I be using models to do parts of the workflow inside of my business, inside the portfolio of companies that we're invested in? Am I supposed to go into the board meeting now on Monday and say, "Hey, XYZ person just did one, two, and three, and cut OpEx by a third"? Do I demand the CEO do that? Do I force change if they don't do it? Do I say... So I don't, I don't, I don't exactly know what to do.

   16. JCJason Calacanis45:17

       The carousel is spinning increasingly faster and we're all on it.

   17. DFDavid Friedberg45:20

       Yeah. I'll just say a general point, uh, which I kind of made at this event today. It feels like the pace of change is so high that you're kind of in a dust storm. You don't really know where you're gonna end up. So it's very hard to sit as an investor right now and say, "I'm gonna pick these things," 'cause, you know, two weeks later, you just don't know whether that path even exists anymore, 'cause the dust storm washes it away or, you know, blows it away. And I think that there will, as a result, there will be a lot of money lost by investors, by companies building in this space. Net net, the index for investing in, like, .com companies, the majority or large amount of money was lost during that era. But from that era also emerged a handful of winners, and those winners ended up creating extraordinary value. I think we're at a point in time right now where we could see 10 times the value generated in this phase of technology advancement than we saw during the internet and the advancement of the internet. And if, if that is true, I think you'll end up seeing certainly the same thing happen, which is the index will lose money, but the few winners will accrue such extraordinary gains. The problem is you can't deterministically pick those winners today-

   18. CPChamath Palihapitiya46:41

       Totally.

   19. DFDavid Friedberg46:41

       ... because of the dust storm problem. You just don't know the path.

   20. CPChamath Palihapitiya46:44

       Totally.

   21. DFDavid Friedberg46:44

       So if you were to meet Jeff Bezos versus some CEO of some .com selling pet stuff back in 1995 to '97, '94 to '97, would you have recognized Jeff Bezos was gonna stand out? Would you have recognized Larry and Sergey were gonna stand out? Would you have recognized Bill Gates was gonna stand out, or Zuck? At the end of the day, this is harder than it has ever been in terms of predicting a technology cycle. But what we still know to be true is that the capital will be allocated i- within a company. The operations will be run, managed, and driven, and led by an individual or set of individuals. And that's effectively what I think a lot of investing in the cycle is gonna come down to. We're all gonna sit here, and pontificate, and intellectually masturbate ourselves to some, you know, genius, you know, path that we think is gonna evolve. And at the end of the day, most of it won't turn out to be true and that path won't be real because this is such a dynamical system right now. There are so many feedback loops. One thing makes one step change and it changes every other step. But what we still know is that great leaders can lead, you know, and especially coming out of this, this Elon discussion and seeing the extraordinary achievements he, he's delivered, particularly today. I think that's maybe what a lot of early stage venture is gonna shift to in AI. It's really, you know, finding great people and, and, and I'll tell you one thing for sure, and I was kinda commenting on this earlier today, was I really think a lot of Series C and later companies, and I know we're gonna talk about this implosion discussion later, so many of those companies have a valuation that's less than their preference stack. And as a result, those founders that work there and those employees that are there are getting their equity wiped out. They'll have to get restructured.

   22. JCJason Calacanis48:25

       Could you do a favor, Freeberg? Can you just explain what that means technically to the audience, preference stock?

   23. DFDavid Friedberg48:28

       So when you, when you, when you raise money into a, a startup, the investors that give you that money, that invest that money, it's effectively a loan. You owe them that money back first before your shares get paid out in the, in the future. So if the company ends up being worth less than the money that they've invested, they get the money first.... and ultimately, if the company goes public or gets sold, they can convert their what are called preferred shares into common shares and participate. So even though you only, quote, "sold" 20% of your company for, let's say, $200 million, that $200 million actually has to get paid first. So now y- you've raised this $200 million. The investors... The company is now repriced because the market has come down by 80%, and investors are saying, "Hey, your company's now worth 175 million." Your company is now worth less than what you owe the pre, or the pr- the preferred investors. And if it's worth less than what you owe, which I think is the case for over 70 or 80% of Series C and later companies, you know-

   24. JCJason Calacanis49:29

       Oh, really? Wow.

   25. DFDavid Friedberg49:29

       ... we can kind of... Uh, yeah, I, and, and this-

   26. JCJason Calacanis49:32

       I didn't know that.

   27. DFDavid Friedberg49:32

       But this number comes from what I shared a few months ago, which is that 70% of publicly traded companies that went public in the last three years are trading below the cash that they've raised. So, if you translate that on a one-to-one basis to the private market, you know... And these, by the way, were the best companies. They actually got public. So, in the private market, you've got to assume that something in that late-stage market is on the order of 70, 80% tre- is worth less than their preference stack. So, a lot of those employees are gonna run. Those founders don't want to go work for the VCs when they get recapped and get offered a 4% equity grade.

   28. JCJason Calacanis50:02

       Where are they gonna go? Because, I imi- it looks like-

   29. DFDavid Friedberg50:04

       They're gonna start AI companies, and I think that's what-

   30. JCJason Calacanis50:06

       Fantastic, yeah.
4. 53:49 – 1:21:26

   Why AI is challenging/exciting to invest in right now, late-stage implosion, pay-to-play rounds

   1. DSDavid Sacks53:49

   2. JCJason Calacanis53:49

      Do you feel like this is a dust storm and it's murky and it's just hard to place bets as a capital allocator because w- something comes out the next day or 48 hours later or the next week that takes the previous idea and wipes it out and then, and how do you scale and be capital efficient?

   3. DSDavid Sacks54:08

      Well, we're at the early stages of a huge new wave and I think that creates a lot of opportunity. So yeah, you've got to basically separate what's really interesting from the fool's gold. There's definitely gonna be a lot of that. But, at least there's a reason now to believe that, say, dozens of unicorns could be created in the next couple of years. So before-

   4. JCJason Calacanis54:30

      Hm.

   5. DSDavid Sacks54:30

      ... we were getting kind of long in the tooth on some of these tech cycles. I mean, cloud, social, mobile, I mean, there was a reason to believe that those earlier waves had sort of played out, that the big winners had already been determined and maybe there wouldn't be too many more big winners in those spaces. But now we have a whole new catalyst for founders to do all sorts of new things and so I tend to think that's super exciting, you know? We're in the early stages and I do think there will be...... dozens of new unicorns minted in various aspects of AI. It could be in AI infrastructure, you know, whether you're seeing now there's a lot of funding that's gone into vector databases or platforms for creating agents. Or it could be in AI co-pilots, basically that tackle various professional categories and create a co-pilot for coders or co-pilot for doctors or lawyers, or architects. I think there's gonna be potentially multiple unicorns created in, um, in those categories. Uh, I think there's gonna be SaaS software products that were just good before but now will actually be great, because the incorporation of APIs from, you know, AI foundation models will- will just turbocharge the capabilities. And so there's a whole bunch of SaaS products that I think become newly interesting and- and better. They go from being vitamins to painkillers. So, you know, we're looking at all those categories and I think we'll end up making some bets. But there's also gonna be a lot of companies that are flashes in the pan or get undermined. You know, there'll be SaaS companies that actually become less attractive because of disruption from AI. But look, I think all of this- this maelstrom is great for an investor. I mean, if you're gonna spray and pray, it's not good. You got to be selective about where you take your shots. But I think this is the most exciting environment we've been in in a number of years.

   6. JCJason Calacanis56:27

      Right. Yeah.

   7. DSDavid Sacks56:27

      I mean, it makes me want to go to work every day and see-

   8. JCJason Calacanis56:31

      It's so funny you say that, Sax, because I literally am looking for an office space in- in San Mateo to start, like, doing the Incubator in person again. And on Monday, I'm having 60 companies come to San Francisco, will be at my attorney's office, and we're having, like, a Founder University with just all these new startups that we invested in to just hang out for a day. The enthusiasm right now is amazing, and what's really unique is the developers who had three out of seven companies they interviewed with offer them 150 or 250K packages, RSUs, whatever. Now, there's no offer from Facebook, there's no Apple, there's no Twitter, there's no Google or Microsoft offer coming in to be the backstop against starting a company. So what are they doing? They're saying, "You know what? I got two friends who got laid off. I got one friend who's halfway out the door. Let's just start something. Let's just start something. Who can give me 100K? Who can give me 500K?" And it's, uh, it's- it's so invigorating to see the talented people, not people who've learned how to, you know, hack a pitch de- together and tell a story, but people who are actually coding and making MVPs. It's- it's truly exhilarating right now the amount of two and three-person startups I'm seeing.

   9. DSDavid Sacks57:47

      Yeah.

   10. JCJason Calacanis57:47

       And so while you'll have this... And it- and it's incredible. I've never seen this amount of destruction and creation occurring simultaneously. I love this zombie porn's concept. You have one half of your portfolio coming apart at the seams, layoffs, reducing their targets, while people are coming in the door with products that are absolutely awe-inspiring. I- I'll just give one example. I- I had a company come out of our Founder University. I gave them $25,000 to incorporate, a developer and his brother who's a screenplay writer. They're taking screenplay writing software, Sax, and you'll appreciate this, having produced two amazing movies, Thank You For Smoking and the Dolly film is called?

   11. DSDavid Sacks58:24

       Dollyland.

   12. JCJason Calacanis58:26

       Dollyland?

   13. DSDavid Sacks58:27

       Yeah. Coming out in two months.

   14. JCJason Calacanis58:28

       They are making... Coming out in two months. Congratulations to Emmy Award-winning, Oscar-winning producer. He's gonna win an Oscar this time.

   15. DSDavid Sacks58:34

       (laughs)

   16. JCJason Calacanis58:34

       He's... You know- you know the screenplay writing tools that have existed-

   17. DSDavid Sacks58:39

       Mm-hmm.

   18. JCJason Calacanis58:39

       ...that are like word processors with formatting? What they're doing is they're saying, "Hey, write some dialogue," and then you can have dialogue and say, "Hey, make it a little snappier, make it a little Tarantino-ish, make it a little more, you know, Sorkin-ish, and then make a storyboard with, you know, stable diffusion." And I was like, "Well, this is the genius idea."

   19. DSDavid Sacks59:00

       Right.

   20. JCJason Calacanis59:00

       "I mean, it's unbelievable. Of course, I'll give you $25,000 for your incorporation." And then they're coming to the ... I'll give another 100K." And- and every single piece of software- Will that company in success ever raise 25 or $30 million, do you think? No, I think there'll be 12 people. I think it'll be 12 people. I'll- I'll give them either 25K, 100K, and then a million. But this is my problem, ... our industry raises $100 billion a year on the premise that each company, before they become a unicorn, will absorb between 500 and a billion dollars. Yeah, no. I'm gonna own 20% ... 10 to 20% of the company for low millions and- and we'll see. Midjourney is 12 people and no, it's totally bootstrapped. Like, I guess what I'm saying is, because in the world of AI, so much work is done for you for free, this is why I'm asking maybe we will have to change how we do business.

   21. DSDavid Sacks59:50

       On the Hollywood example, there's about to be a Writers Guild strike.

   22. JCJason Calacanis59:53

       (gasps)

   23. DSDavid Sacks59:54

       And they may want to think twice about that, because this is not the time where you want to be encouraging the industry to find alternatives to writers.

   24. JCJason Calacanis1:00:02

       You want to get back in the office and you want to say, "Hey, can I ... Is there any other work I can do this weekend, boss?"

   25. DSDavid Sacks1:00:08

       (laughs)

   26. JCJason Calacanis1:00:08

       Did you see the news about BuzzFeed today? They had won a Pulitzer. They just shut down BuzzFeed News, the entire news division gone, another 15% gone. The layoffs, and then there was a report this week that Zuckerberg is doing his third round of layoffs, another 4000 people. That puts him at 24,000-

   27. DSDavid Sacks1:00:26

       Yeah.

   28. JCJason Calacanis1:00:26

       ... with a hiring freeze.

   29. DSDavid Sacks1:00:28

       So the way I describe the current funding environment is it's a tale of two cities. It's the best of times, it's the worst of times. If you're a hot AI startup that's able to tap into the zeitgeist, that's doing something that's perceived as cutting edge or relevant, there's a strong why now and you're early, you know, you're early stage, you're able to raise money for that. The spigot is turned back on. There's a lot of funding for those types of early stage startups. But if you're a Series C stage star- you're a late stage startup with a ... it's called a pre-AI model, the spigot is just turned off completely. I mean, let's look at that chart from-... crunch base, where the amount of Series C funding has gone from something like 10 billion a quarter last year to, like, zero. I mean, they're all-

   30. JCJason Calacanis1:01:16

       This chart is-

Episode duration: 1:39:11