A (1:51)

Yeah. Which is all the kids are back to. I'm stuck outside my office. I don't know how to get into my office. And that's just how it is. But someone will eventually let me in. The people who work in this office claim that they went for a walk. I think they're in their apartments, running to the office to let me in.

B (2:05)

Yeah, I think that was a work from home day until the boss showed up.

A (2:09)

And in fairness to them, it's a very rare occurrence. So they really. They're not wrong, they're just wrong.

B (2:14)

This one time they went for a walk. I love this.

A (2:17)

That's literally what they told me about.

B (2:18)

I just landed in Austin where, guys, there is no Waymo. And this is the first time I've experienced Waymo withdrawal. When I was flying here, I'm here for a conference. I was like on the plane googling, is there Waymo in Austin? I was like, phew, there is. It's through the Uber app. Great. Also love Uber. Open the Uber app. Did. No, I do.

A (2:37)

Wait, they can get you Waymos and Uber.

C (2:39)

Okay, first of all, you can only get Waymos through Uber in Austin, which is the weirdest thing ever.

B (2:44)

Guys, you obviously have not read any of the information's great reporting on Waymo. Waymo has a three pronged business strategy. I don't know, they have their own network in certain cities and then they're working with other rideshare providers on the back end. In other cases, like the Uber partnership in Austin, I am sorry to report this partnership does not appear alive and well. When I go to the Uber app, as I do, I do not see Waymo.

C (3:10)

It's because it's a roll of the dice. So the way it works in Austin, by the way, which is my hometown, is you pick a Waymo and. And it's a roll of the dice if you're going to get one or not through the Uber app. And. And so I actually am Waymo lucky. I end up getting Waymos a lot in Austin. I think right now there's a bunch of tech people come in Austin already and a bunch of them flying to Austin, namely my husband and you, Jess.

B (3:33)

Yes, we were both here. Yeah, they're.

C (3:36)

I think there's just higher demand.

B (3:37)

I am sorry, I cannot be a proponent of a service that gives me something based on a roll of the dice. This is not how I operate my life.

C (3:44)

Everything's a roll of the dice in technology. Jess.

A (3:47)

Here's the thing. I actually just think you need to get into the gatekeeper where they, where you get all the Waymos. Like this is like, remember how Uber very smart.

B (3:54)

Obviously I am one tweet away from that happening because I have become a Waymo influencer.

A (4:00)

Uber did a very smart thing which turned out to be, I guess, sexually quite somewhat illegal. I don't know. Remember they started making all the politicians never get Ubers because they didn't want to get regulated. They had like a gatekeeper which like kept out all the Bad people. Yeah.

C (4:13)

Can we get on the Waymo list?

A (4:14)

And now they just need to do the inverse. Wasn't that. Jess, you remember this? What did they get? They got. They got in trouble because they basically created a. Was it black ball? Was it called black ball?

B (4:23)

It was called gray. Gray balling gray.

A (4:27)

Totally reasonable. I mean, not nice, but totally reasonable. I just. You need white ball.

B (4:31)

Yeah, I just think I have a problem with like the Russian roulette approach to getting my way mos. It's like your hope is up.

A (4:39)

Yes, it's called reliability is kind of one of the key. No, the only reason I now use Waymo is like you actually can get them within 10 minutes wherever you want want them. If that weren't the case, you'd never open the app. But the fact that you can do

C (4:49)

that now, I mean, I get.

B (4:51)

I schedule ahead. They're on time to the minute. And I gotta say. And the Internet is skeptical of this claim, but as someone who's commuted to my office for a long time, twice a day, Waymos find faster routes every day. They are optimizing the Bay Area traffic with some AI stuff in ways that's faster. And it's. Anytime I get in another ride sharing service, the time to destination goes up. It starts at 24 minutes, it ends at 34 minutes. Every Waymo I've had in the last two days has been shorted.

C (5:23)

This is not just the future of autonomous driving of waymos. I think it's autonomous driving in general because, like, we're investors in Coco Robotics who have those little robots that take things to your door for like doordash and Uber eats and stuff. And they also get faster destination time because they're taking back alleys and they're taking optimal routes.

A (5:41)

To be clear, Uber could do that too. That's just a better product experience to like lowball people and then. And then improve. There's no reason that Uber human drivers.

B (5:49)

Well, that's what I was wondering. But that now you're. Yeah, but they're not lowballing, because I

A (5:55)

think they probably are if they're smart. Like, why wouldn't they lowball and make you exceed expectations?

C (5:59)

Speaking of Uber, can we talk about the Uber CTO news that you published this week?

A (6:02)

Wait, before we do that, can I make one other ode to Waymo, which is here's the reason I like it is the thing. I actually literally go to San Francisco more because of Waymo. Right?

B (6:12)

Because clearly you should get a key card.

A (6:15)

I'm really going to go to a Friend's office and I can't get in my own office. And luckily I have a friend around the corner.

B (6:20)

I have an office. You can go to my office.

A (6:23)

But here's the thing is it's consistent driving style. Smooth. So I don't get sick. My number one problem with Uber, I

B (6:29)

don't get carsick, I don't get nauseous.

C (6:30)

Same. That's.

A (6:31)

That's the thing. And the thing is that, that the average Uber driver is worse than the computer at driving. And then really every once in a while you'll get an Uber and it's like some guy in a rice rocket and he's just jamming on the brakes and you're just. I'm so sick right now.

C (6:44)

That's a right. Stop.

A (6:45)

You're making me physically ill. Look it up. It's probably not a term you're supposed to use anymore.

C (6:49)

I feel like that's maybe a derogatory term.

B (6:52)

No, I agree. Car sickness down. Efficiency up. Podcast listening up.

A (6:58)

I just emailed the entire way to the office. Hey, Larry, what's up? There we go. Someone in my office.

C (7:04)

Larry. Larry.

B (7:06)

Thank you for helping.

A (7:07)

Thanks, man. We'll catch up in a minute.

C (7:09)

Does he have a coffee in his hand, Sam?

A (7:10)

No, he doesn't. By the way, you want to see our. Here's our chief. Look at this. This is chief compliance. Be chief compliance officer.

B (7:16)

Okay. No one wants a tour of the Slow Ventures office. Very cute. Okay, so we. This is not brought to you by Waymo, but could be. Also guys, I love non San Franciscans who come to San Francisco for work and experience Waymos and then go like we do like full Instagram Waymo Life Waymo. Maxine and I also predict by the Olympics in la, Waymo is just gonna

C (7:38)

be like next up Joby will be flying everywhere.

A (7:41)

No, no, Brit, this is where we lose you. This is where you just like take it and go to the far.

B (7:45)

Guys, a lot of AI news and I do want. We have not talked enough about SpaceX. SpaceX by the month of June, but six weeks away is going to try to have the most epic IPO of all time. So we're going to talk about SpaceX and the information stunning revelations that they are losing billions of dollars. Specifically five but. And I'm going to make you guys do it like bull bear case on SpaceX valuation. So prepare for all of that. But first, we have some news. Sam, I'm going to give this to you because I believe you were an investor in the once known as All Company soon to have AI pivoted to a data center company.

C (8:21)

AI Birds. Is that the official name?

A (8:23)

This is Jonah Peretti all over again. When I looked this up two and a half years ago, Buzzfeed announced it was an AI company as a penny stock and fucking ripped for a good day, went back down. It's a great trade strategy.

B (8:34)

Yes, and it has not served poor Jonah well. But also, Jonah didn't then say he was going to get into the business of buying GPUs. He just wanted to do our surveys. Oh, I see.

C (8:46)

Jess, can you back up and just like present the.

B (8:49)

Of the.

C (8:49)

Of the story?

B (8:50)

Yes, I'll present some facts. I'll. I was hoping Sam would, but. But silly me. Okay, so this all came out through a filing because apparently Allbirds, which you can you. Everyone knows Allbirds. You know the like cloth shoe company. How else do you describe it?

C (9:03)

Well, Dave and Sam are like the first investors.

A (9:05)

I think in fairness, I will take zero credit for the. This is not a Sam investment. We were investors as a fund, but my God, this is not a Sam investment. I don't do well.

C (9:14)

It actually performed well for the fund because you sold at the right time.

A (9:17)

We sold well. We saw. I'll tell you exactly when we sold. We sold it like the Series A or B. I can't remember because you started looking at the numbers. It was so insane. We're like, the only way it makes

B (9:24)

sense to hold this selling at the Series A is.

C (9:26)

I don't think it was the A. No, it was later than that.

A (9:29)

I remember then it was the B. Then it was the B because it was getting. So it was great product. Like we love the product. Great, great products. But the math became simple, which is the only way it made sense to hold it is you believe that Allbirds was the next Nike. And we're like, we do not believe that Allbirds is the next Nike. And so we did start selling Allbirds,

B (9:47)

described by the Financial Times as the maker of wool trainers, which I love. Very British.

C (9:53)

Seriously, very comfortable. I have to give them that.

B (9:55)

Was once valued at more than 4 billion. Sold this month for about 40 million. And this shell company for its shell listing has informed people via a regulatory filing it is going to pivot its business to AI compute infrastructure with a long term vision to become a fully integrated GPU as a service AI native cloud solutions providers. It will be known as Newbird AI New Bird.

A (10:22)

It's just a cheaper way to like. This is the thing is like, it's still expensive to get Public and there's a bunch of people who want to be public for cheap access to capital. I've actually looked into this and like, honestly like buying a penny stock is like cheaper than the cost of lawyers and the bankers to go public in some cases. And so that's all this is.

B (10:39)

That's a great. So Sam, what is the temperature on these SPAC like shell listings? Because I don't think they've gone so well for the companies. But this is still a thing. And why is it a thing?

A (10:51)

Thematically, no one thinks they go well, but then everyone thinks there's an exception and maybe there are some excep to it. Right. Like, and the reality is these types of things specifically there's like some play that like requires cheap access to capital on a narrative. It's like, it's not a game I enjoy, but I understand that like there's some haranguings you can do with it. It's just funny that like who invests

C (11:10)

in these really weird hedge fund people?

A (11:12)

Well, it's not even weird hedge fund people. The reality is there's all sorts of like covenants and deals that go into these types of situations where like if you really want to be wonky about the finance of it, it's pretty hard to lose money and there's a ton of asymmetry on them. Right. And so like it's just a, it's a very bottom feeding me strategy. But the reality is, is like even if the stock trades down, like there's all sorts of ways that people make money on putting these things out or doing these types of structures. So they bunch of warrant coverage and you have all the covenants and blah blah, blah. It's totally not my world but like a rationale to it in terms of how it plays out. And again it's not as good for the companies, but people do make money on these things.

C (11:48)

At first when I saw the headline, I thought it was going to be something really cool and interesting where like everyone that's walking in their shoes is somehow creating more like there's something cool about it.

B (11:58)

That's a very bright brain, so optimistic. She.

A (12:02)

It's great. It's actually I need this in my life.

C (12:05)

This would tie it together for me.

A (12:07)

My favorite irony run right now, guys, is you guys remember Boom. Supersonic.

C (12:11)

Yes. Oh, they totally pivoted.

A (12:13)

Question, here's why. It's because it was the dumbest business. Like this was the company that was.

B (12:18)

Okay, so Boom was going to create supersonic airplanes, right?

A (12:22)

It was like an in terrible business. It's like a very nice vision in the Elon style. Like cool, like Concord 2.0. But like from a practical perspective, it's like, no way is this gonna work. No way is this gonna work. Right? And see how these investors marking this thing up on the story. Because it's cool people. It's a cool vision. You're like, this is just completely nonsense. Hilariously, it's gonna be worth a fortune. Because they couldn't get access. Why?

B (12:44)

What happened?

A (12:45)

Because basically they couldn't get. They couldn't make jet. Like they couldn't get access to jet engines. And there was like no way they were going to make their own jet engines that were. That, that would get certified and be able to fly. But they got far enough along and then I think AI probably helped them with this that they have a turbine they can use to run data centers. Power for data centers. So they're not a supersonic company, they're just a jet engine company because they couldn't get anyone to sell them jet engines, right? Or like make them for them. So like hilarious. Like you just have to be at the table story where it's like people who invested in it are actually going to probably going to make a ton of money, but completely for the wrong reasons, right? Or not because the idea was good, the idea was bad, right? But like they're going to get away with it and you kind of got to love it, right? Like the major swerve on it.

B (13:27)

We should probably cover more energy. Not as cover like to the degree this podcast covers anything but like the information has maybe 6 month old, 5 month old AI infrastructure and energy newsletter which is booming. And I was talking as we have two writers for it, Anissa and Anne. And I was talking to Ann yesterday because she was in San Francisco for an energy conference. And I don't know, maybe I don't know if she told me this on or off the record, but she thinks this whole like energy crisis is way overblown and that we are only using half of the energy we have in the.

A (14:04)

And that's before Venezuela today.

B (14:06)

And that's God, Sam.

C (14:08)

Oh God.

B (14:09)

And Hormuz. Anyway, so she and her sources, she's like an old school energy reporter and they are very skeptical of this energy scarcity narrative. But anyway, I thought that was interesting. Obviously we don't really know because we can't predict demand. But it makes sense to me that there's a rush into all of that energy stuff. So speaking of booms, let's talk about SpaceX, guys, okay, do we have bulls? Do we have bears? What are we thinking? And I'll also tee this up. I was kind of, I'm spending a lot of time talking to investors who, you know, big public market investors about how they're looking at possible anthropic SpaceX and OpenAI IPOs. And I gotta say, by and large, and it's anecdotal, but there are enough of them, they are very, very excited about these companies and they're selling down their Mag 7 positions a little bit or preparing to, to like go all in on these companies. So I'm sure there are other points of view, but I've just been struck.

A (15:05)

I always wonder, are these people tax free or taxable? Because this is the most interesting thing. Well, this is the thing I just think is so weird about how the public markets work in these big funds, right? Which is there's two very different types of investors that invest in these things. One are people who like, for whatever reason are tax free structure so they can switch what they own without paying taxes. And the other people get taxed when they sell.

B (15:26)

These are like the world's largest professional money managers. So they presumably aren't tax free.

A (15:32)

Well, it depends who they're managing money for. I mean it's complicated. But this is just like this thing which is like selling down supermarket is the answer is like selling down super high march positions. Right? We have to pay enormous taxes like to switch into something like an open air whatever.

B (15:47)

That's interesting.

A (15:47)

You basically aren't betting, you're betting that whatever you're selling out of is going to underperform everything else by 50. Right? And that's this like crazy hurdle to get over emotionally and mentally where you're not actually saying like I'll give you the SpaceX, like fortunately, because like I have challenges. Elon. Elon's a complicated character, pros and cons. But like SpaceX is so cool that like I invested a bunch of money in Space X a long time ago. I would have an enormous mark on that when it goes out, regardless of where it goes out. Good for me. But here's the problem. The question of what I do with that is really a challenge because I don't want to pay 50% taxes or 40, whatever it ends up being in California to like rotate out. I'd rather just be long it and hold it because you'd have to believe not only is SpaceX overvalued or you have to have the opinion that it's so Bad. Right. That it's going to underperform the rest of the market by 50% from where it is for it to be worth taking the tax burden of selling it. And I think a lot of what you see in these high marked positions, that's a good questions from my perspective especially for people who like if you're sophisticated you can borrow against your positions, et cetera is like they get super marked up when they're right and then you really can't sell them. And so that creates this floor on the pricing where like no one wants to rotate out of them. Like it has to be really painful to rotate out of them. Right. And that is a really interesting dynamic that you know is part of why the market is so disconnected from the economy from my perspective. Right, right.

C (17:08)

Because the S1 is saying what they're going to go out at 1.75 trillion as the valuation. And then I'm confused by the numbers too just because the information reported something different from Reuters. Right. So there's no why.

B (17:20)

I mean please.

C (17:21)

I'm just, I mean obviously I know who's right and just making sure that the numbers are all a little wonky.

B (17:28)

The valuation stuff is all pure rumor now. No, no one knows what I mean that, that isn't to me it's just kind of irrelevant. But let's take these one at a time here. Let's take these one time first. Sam, that, that is a very good point. And I'm. But I assume there's some tax charming.

A (17:44)

Not if you're like if you're a university. Yes. And that's the thing that's really interesting is like universities or people who are not taxable get to do way better moves in the market than anyone who's taxable. Right. And this is a weird thing which is very bad I think for the economy long term that all the money and returns will pull up in nonprofits because for profits get so whacked. Act on making smart allocations. But it's funny because for me you're like do you hold or sell SpaceX? It's actually not a question is it worth $1.5 trillion? You have to ask yourself is it worth 750 million? Right. And that's like a very different question but from an.

B (18:14)

So that's an important consideration in these funds allocations. But also I think you look at the Mag 7 and or whatever approximates it and like you know, is there a 50x story for those companies? Like you're also thinking about that if you're an investor?

A (18:31)

No, totally. But, but, but there' with that is 50 is a big number. But here's the thing. It's the same problem recursively, which is how big can companies get? Right. When Facebook was trading at 100 billion, that was a big number. People were like a trillion dollar company. That's insane. Right? And what's happened is there's a lot of money that flowed into the system. Everything rises, blah, blah, blah. The market is not the economy. And so all of a sudden the 10x case has nothing to do with a thing you're projecting about the business. It's just like a bunch of like the world changes. And so for me the question of like is there a 50x on any of these things has so much less to do with relative value than just like where like how much money gets printed and like a bunch of other factors like that. How big can things get? How big number go up? So then you're like, yeah, like I guess it can over a certain period. But it's, it's more of a bet on the secular like economy than it is a specific company.

C (19:22)

And so is the bet Sam, then that you'd think SpaceX would have to get like 5 to 10 trillion in value to make this reasonable. If it's going out at like one and a half, like it's at least a 3x.

A (19:32)

I think the basic point is like if you have high March positions from zero, which SpaceX will be for everyone? Everyone, right. Anyone who's in SpaceX is in it a lot lower than it's going out in theory, I think. Right. There's this thing which is like you really have to not believe to sell it because the default course in speed is that all the big companies will kind of pace each other because people have money and have to stuff it places. Right. And so it has very little to do. Like the question of Is there a 10x in SpaceX? There could be a zero if it falls off a cliff. It feels unlikely. The question is more like you have a tax free marked up thing. At what point are you willing to take a 50% haircut to get out of it? You don't have to believe it massively underpaces everything else you could invest in.

B (20:11)

So that makes sense for the existing investors though. But this performance of the stock is going to also be largely shaped by new investors coming in. And that's the whole idea of going public access to new capital, right?

C (20:22)

Well that's what I mean. That's what I mean. I mean investing As a new investor at the current valuation, you have to believe this is a 5 to $10

A (20:29)

trillion company, I think maybe. Or you just have to believe. Here's the reality, which is like, not enough people, like, people hold it long term, right. Just as they hold other things long term for the. For tax reasons as well as other things. So there's not that much. There's going to be less supply than you think of shares. Right. And then the demand for shares will have much less to do with like a specific thesis that will be like, yeah, you kind of have to buy it because it's huge, right? Like if you're indexing a bunch of stuff and it's in a bunch of indices and like you just. It gets bought and like. And then like you have this thing where it kind of floats up with everything else, then the retail cult on top of it is a. Is a bonus.

B (21:03)

Yeah, that makes sense. There, there are. There is money flowing out of these really big tech companies into like Asian chip companies, though. So it's some point, like point standing that you have to tolerate your taxes. But I'm not talking about investors who were in early and have huge markups. I'm just talking about the world's largest money managers who were buying and selling constantly.

A (21:25)

It's It. Yeah. That's just an interesting question of who actually is trading this and for what purposes and what their tax structure is around it. Because it is. You say, hey, I have fresh money to deploy. Where do I put it? That's one thing. But what the recycling is just like killer. Like you have to have a very small.

B (21:39)

There's not a lot of new LP money going into the system. So I think a lot of it's recycling. Let me explain this thing. I, I mean, I think this is why you. I hear a lot of hedge funds targeting retail now because they think there's certain. Maybe. I mean, I. The top tech hedge funds have all told me this. They think that the institutional or stubborn wealth is sort of fixed and they'll fight for a slightly more share of that, but they have to go to retail or other things. So.

A (22:09)

But on that note. Bye, ladies.

B (22:11)

Oh, my God.

C (22:12)

Wait, what just happened?

B (22:13)

He dashed and died. It is the bottom of the hour meeting it as possible seem as a meeting I just don't like.

C (22:21)

Is there any heads up about this? Oh my God.

B (22:23)

Maybe. Honestly. Okay.

C (22:26)

Wow.

B (22:26)

It's the Brit and Jess show.

C (22:28)

Thanks, Sam.

B (22:30)

We don't need them. Okay, but. But we have a picture.

C (22:31)

Everyone should comment about this and give Sam shit on Twitter.

B (22:35)

I actually, I didn't even give him a chance. Sam has this weekly email, as many of our listeners will know, and he wrote a particularly introspective one about AI and the crisis of meaning that it is precipitating and this being a sort of key issue that has gone undiscussed in society and in the AI productivity even. I got feedback on this email which does it or this post, which. Which he also shared online and has apparently spiked his newsletter subscription. So I was going to let him talk about it, but now I wasn't.

C (23:05)

Positive feedback or negative feedback?

B (23:07)

Well, I got two types, but most had an element of if Sam is on something, please get him off of it. That was like the prevailing, you know, like implying that mushrooms were a boss. Because he was very philosophical. Very philosophical.

C (23:21)

Philosophical, okay, philosophical.

B (23:23)

Just a little, Little loopy. Little loopy. I blame the jet lag. Okay. But SpaceX's business. Turns out it's not space, folks. The star. Well, I guess it's kind of obvious. Starlink three times as big almost as space launch.

C (23:37)

We were just talking about data centers compute. Like this is everything, right? And I think this is like Elon's way of getting data centers in space and everything that we need to power the AI future that we have. The space launch here that you're showing in this chart is. Is 4 billion out of 19, right? So it's like less. Yeah, it's about less than a quarter.

B (23:57)

This is something I wonder and I don't know an answer to this but like how long space economy do you have to be to be bullish? SpaceX, right? Like, because they. And I'm interested as they really embark on their roadshow in earnest. Like how much part of their thesis is. I mean, I'm sure it's a part, but how much of it is like. And the space business of launching things into space will be projected to go to whatever because we're going to have bases on the moon and you know.

C (24:25)

Yeah, Elon's moved a lot of his interest from Mars to the moon recently.

B (24:31)

It seems like, yes, they pivoted to

C (24:33)

the moon, which doesn't bode well for my bet with Sam about.

B (24:37)

I was going to point out that

C (24:39)

stepping on the moon by 2040, but that wasn't my fault. That was Elon changing his strategy. So if he would have, why Sam

B (24:46)

and I took the other position. But yes, you know, we'll see.

C (24:50)

Well, I'm still not giving up my bet yet. 2040 is a long ways away but

B (24:53)

what any other SpaceX. I mean, the other thing I hear that wasn't obvious to me is how much this will be seen, this IPO will be seen as a bellwether for Anthropic and OpenAI, even though there are parts of their business that are competitive parts, not just the idea of. And the test case of a mega,

C (25:13)

mega companies that like are losing, losing revenue and going out at like 90x or a hundred x their actual revenue on the market. And Anthropic this week I heard, is also now raising at an 800 billion.

B (25:27)

Well, they're getting offers. I mean this is like I, one of the, one of the hardest things. It's not one of the hardest things I do. One difficult thing is when a fundraising story becomes a story because, you know, hot companies, people are like sending them emails and writing letters at all, you know, moments. Right. And so at some point the company decides to do a round or not.

C (25:48)

Yeah, but don't you think it has to do with all of the momentum in the last month and sort of the Open AI versus Anthropic everything. And I mean Anthropic's last valuation on paper was 380 or 360, something like that, I believe. And so now they're trying to anchor with OpenAI, I think.

B (26:04)

And I, we showed a couple weeks ago the revenue chart the information had in the lines. Like we sort of talked like when will they cross? Right. Or will they cross? And so I think it's, it's still unclear. I mean, at the same time Anthropic has major capacity issues.

C (26:20)

So many. Yeah, COMPUTE has been insane with Anthropic

B (26:23)

and we had a story this week about them changing their pricing in partial response to that. And so I, I don't know where these things stand.

C (26:32)

They, they, I think everyone has COMPUTE issues though. It's not just Anthropic.

B (26:35)

Yeah, but I think Anthropic is more, I mean the, the sense of word on the street. I mean you can, it, it's, you can see it either way.

C (26:43)

Right.

B (26:43)

Then there are also critics of OpenAI who felt like they overbuilt. And in fact the information has story in the last week about some tensions between Sam Altman and Sarah Fryer, OpenAI's CFO on IPO timing, a lot of which came down to, you know, her concerns about these COMPUTE commitments and what OpenAI could potentially be left with. So, you know, but at the same time there was a memo that was leaked from OpenAI, an OpenAI memo that was leaked. I think it was in OpenAI's interest for this memo to be out there. I also think it was written in a way where they wanted it to be out there, where they came after Anthropic, both for how it calculates ARR and also on this point of having underinvested in compute. So it is a problem for Anthropic.

C (27:26)

And it was not the same memo that was also leaked today about OpenAI's focus on the enterprise moat and how it's not about the model, it's about the entire system and, you know, obviously condensing everything into Codex plus chat and everything else. I think it leads to what I'm seeing, which is just like we've talked about this on the pod every episode. All of this is electricity. How do you get someone locked in? Because you could switch one model out for another. And if Claude's COMPUTE isn't holding up, I'm just going to go over to OpenAI and everyone at the Frontier Labs right now, I guess seem to think that owning the enterprise and just getting people super reliant on your stack is the only way to stay relevant and to win in the differentiation war. Google today launched Gemini on desktop. So Google's coming out hot with their response. And obviously we've got all the developer conferences coming up in May and June.

B (28:22)

This is what I said in my team starting like March. I'm like, guys, developer conference season, what's coming out, what's coming out? It's going to be like Alphabet soup of announcements and code names.

C (28:33)

But I just think it's still hard to call any of this. I've been in board meetings a lot recently with companies who are like, yes, we're fully moved over to Anthropic. And it's like, like they show the, the chart of when everything moved over. And it was like in December when I'm like, of course everyone moved over to Anthropic in December, but it's April and who knows what like April 2027 will bring? You know, everyone is going to be launching new things all the time functionality, like.

B (29:02)

And I, I mean, we have a tiny, you know, tiny engineering team relative to all these companies information. But my conversations with our team are like, okay, what do you think based on like roadmap and what you think now? What's our best structure for this? And is it cursor? Is it Claude, Is it what? You know what? Whatever it is, I get an answer and then three weeks later they're like, answer changed. We think this approach is better. And then all of a sudden like I have an idea for something, I'm like, I want to build that. And they're like, well if you want to build that then the answer has to be that. So of course it's optionality. And I think the now obviously big companies, you are doing these big enterprise deployments. I mean, how ironic that it's like the SaaS strategy when everyone's trying to rip out SaaS. But I also think at thing will play a way, way bigger role in this. And these total anthropic and OpenAI bills, like don't acknowledge that for many, many things you're going to use cheap models. I mean we have, I mean Meta's AI works well. I mean it's a consumer facing but like it's better at giving me what's happening with tennis than anthropic.

C (30:04)

And well, just depends on the task that you're doing. Right. So every company is going to have to orchestrate their like local cheap model. They're super like luxury fancy but quick and smart model. And it's not going to be about tokens per engineer or whatever like Uber and Meta are actually doing on leaderboards right now in their engineering department. Yeah, because that's not the actual best way to see if people are leveraging AI. Right. When did they use the right amount of tokens to get the right amount of success? And I don't know how we're going to measure that, but that seems to be what every CEO is thinking about right now.

B (30:37)

Do you have companies, Brett, that are token maxing and what do you tell them? And for people who don't, I'm sure our listeners know, but it is kind of wild. Like their big companies are incentivizing ranking people based on using the most, which equates also to the most expense, which is not historically how PNLs have been optimized.

C (30:56)

I think the only facet of that that I like is that right now we're in the part of the curve where you just need everyone in the company to be using this. And so if there's a way to encourage. It's almost like a habit your brain has to learn. Like there was a CEO the other day who was, who was like asking me, okay, I'm ready to try open claw. I'm going to get my CTO to help me set it up. And I was like, no, you should just go to cloud or GPT and ask it how to have you. You set it up and it's like learning to think that instead of asking the person, I'm going to ask the agent is a habit your brain has to learn and switch into. And for people in engineering, I think that's more native right now or product. But for people in sales or HR or you know, even some just non technical roles, it's not native. And so the first step is just getting everyone to think AI natively. The second step is then to think about like okay, now how do I use these tokens more efficiently? And Jess, let me tell you, as a parent of an 11 year old that I know we've talked about on the pod a lot, who is like token maxing. We looked at our bill, our token bill for the last month for Ansel, our 11 year old. Guess, guess how much money we had to pay for tokens in the last month.

B (32:03)

But 10 years.

C (32:05)

Okay, less than that. But more than a thousand.

B (32:08)

Wow. Thousands of dollars.

C (32:09)

Fifteen hundred dollars of tokens that my 11 year old has inadvertently.

B (32:12)

Well, not inadvertently. He's got a business now. You just have to get him selling.

C (32:15)

I know we gotta monetize this thing because like the Morin bank account can't sustain 1500amonth for each child.

B (32:22)

Yeah, no, that Sam also got the boys building games which is totally like. They have just totally subverted all screen time rules by like convincing Sam this is in a different category. He doesn't quite see it yet, but it, it's quite funny to watch. So what else? We didn't talk. Did you and Dave talk about the open AI New Yorker profile last week by chance?

C (32:45)

Well, we had a long laundry list. I don't know if you, you. I know you were in Italy and you didn't catch get to catch up because we had Guillermo from Vercel on the episode last week, which was fascinating because he's truly at the center. He's like the Switzerland across all the big labs and was telling us all the tea. It was on the laundry list of news topics but we didn't get into it. But I, I think that. Yeah, I mean also Sam Altman's house got like attacked twice in the last week.

B (33:11)

Yeah.

C (33:12)

So Sam Altman and OpenAI have definitely been through the thick of things in the last seven to ten days.

B (33:17)

Yeah. No, obviously the b like tragic and scary just for a second of media criticism. This was so Ronan Farrow had been working on a profile of Sam Altman for I think he said 18 months. I caught wind of it a little over a year ago just because he was calling everyone and the Word on the street was that he was, like, probing some really serious allegations, like, none of which were in the story. And actually, you know, came out as sort of defending Sam in a lot of contexts. So it's interesting because I think there's. There's some group of readers. You know, it was. It was a very tough look at his management style and personality. And what. For anyone who followed the saga of Sam being ousted as CEO, the piece will be familiar to you because it retread a lot of that ground. But for other people, like the. Like, my inbox is full of people who are saying, like, oh, that could have been much worse. Like, not much worse because they know of bad things, but just like, there was so much anticipation that this was gonna be like this real hit piece and it kind of coming out. And we know, like, Sam was very, very anxious about the piece. The board was anxious about the piece. So, yeah, from my point of view, it was sort of a nothing burger. But I say that as probably the person in the media who follows this story the most closely so well.

C (34:33)

I think, though, the thing that a lot of people aren't understanding right now is that there's so much competitive pressure amongst all these Frontier Labs and Max7 companies that, like, there's truly oppositional research teams inside each company that I've now heard about for probably five to ten times, like, frames different stories and companies that have been planted by the other. And it's like the nastiest PR setting I've seen in all of tech. Like, I have not seen people do this kind of, like, drastic PR strategy since the early days.

B (35:10)

You know, these companies are also these Personas and personalities, so have all this history, but. And so I actually think for Ronan Farrow, who's the reporter who took down Harvey Weinstein, like, to come out and not, you know, after 18 months of, like, following what all those competitors were saying, not find anything publishable or, you know, it's sort of in some way a relief to people. But, yeah, it's nasty out there. I mean, when we published some of Dario's anthropic's internal messages where he talked about why in the wake of having lost the Department of War contract, you know, it was really a window into his thinking about, you know, he was throwing. Doing a lot of shit with OpenAI and that. And now we're seeing it. So it's memo war. It's the era of the memos. We've reached that point in the cycle. But, yeah, to be continued. I'm sure There'll be a lot more OpenAI news to come. Britt, what else is cooking? I've got a book recommendation. But what.

C (36:05)

Hold on, I need to talk about this Carta news that's coming out because, yeah, I can give you like, what's happening in venture capital land, which is, I know we've been saying we're in the bubble bubble for a long time, like a year or two years, but truly we are in the bubble of all bubbles. And I don't know how like, sustainable this is. So Carta announced the seed valuations are skyrocketing more than ever before. So between Q4 and Q1, things went completely vertical. Now the top 5% of seed rounds are topping $175 million valuations. Oh my God, 3x over the last year.

B (36:40)

Wow.

C (36:40)

And even in the 75th percentile or 50th percentile, we're talking about like $45 million seed rounds, which is insane. Like a typical seed round used to be $15 million, maybe 20, you're by 10% for a million and a half, $2 million. And now if you're doing that, you're actually under the 25th percentile. So it's totally crazy. And you might think, wow, this, this sounds a lot like 2021. And actually it's double 2021. So it's like twice as worse as 2021. Like if there's a graph that I can share that, that Carta also announced today and it is like the top 10% of startup valuations for series D, C, B, et cetera are now like 200 to 300% higher than 2021. I really hope you know we're going to see a lot of trillion dollar companies.

B (37:40)

As an investor, like, what is the takeaway? One should, it's making this.

C (37:45)

Well, so like we're about to go out and raise our next fund.

B (37:48)

Can you say that? Haven't you just violated some major.

C (37:51)

Okay, maybe I should take that. Okay, let me, let me start that again.

B (37:54)

You're fine. You said you're about to, you haven't.

C (37:56)

We're about to. We're not in processing something. I'm not soliciting investors, but we are going to raise our next fund soon. And I'm doing a lot of like, data gathering about portfolio construction and all these things. And like, when you run a portfolio construction model as a venture capitalist, you're saying, okay, cool, I've got a hundred or $200 million fund size. I'm going to make X many bets at this Amount of dollar per check in this fund. And typically that would be very easy to forecast because I'm going to make, you know, $32 million checks and this many angel checks, et cetera. And this much for follow on now, I'm like, like, am I going to make $32 million checks or am I going to make 10, $10 million checks or am I going to blend? Is the, like, where is this? Are you following where inflation's going? And if so, do I need to project that out, out over the next two to three years? Like, is the AI bubble going to keep going? Is it going to fall dramatically? You know, there's all these questions you're asking yourself that you then need to like communicate to LPs who are going to be like a lot of them, not, not the most sophisticated ones, but a lot of them are like, well this doesn't compute in my spreadsheet, like we want 10% in every million. And, and so, you know, I think that it's just this active conversation ongoing between GPs and LPs right now of like, where do we even mark these things? And at the end of the day you don't want to be in the like lower quartile of companies because honestly the ones that are able to command the higher valuations tend to be statistically the most successful. So it's this like blessing and curse where it's like, should I pay up and how much of my fund do I put in in at a $200 million valuation for a seed round?

B (39:37)

I mean that's what's going to return your fund unless it doesn't collapse at

C (39:41)

the end of the day. Like what I've come to is like it just has to be what percentage of the fund is going in rather than what percentage of equity you're getting out of these companies. Right? Because like if I'm putting in a 5% bet from the fund, that's like a 5 million check on a hundred million dollar fund, like that's significant, right? And if you can, if you can hundred up checks that 5% check, like you're in great territory. So anyway, I think fund math is changing and no one really knows how long we're going to be in this bubble. But if you look at that same chart I just showed of what happened in 2021, I mean the bubble only lasted for, you know, a year and a half or so.

B (40:17)

But this is AI.

C (40:18)

People are saying AI is going to last longer, it's going to last forever.

B (40:21)

I mean, what is AI even What is a bubble?

C (40:24)

What is a bubble?

B (40:25)

I do have to bounce in a second to do my, my, my real job, but I gotta go full millennial and plug Lena Dunham's new book, Fame Sick.

C (40:35)

So old school Lena Dunham. I haven't like paid attention to what she's doing in a lot of years. Since Girls on hbo.

B (40:42)

Since Girl. Well, I think this is the her journey which has had ups and downs and horrors and show like some shows, some creative pursuits. But I obviously loved Girls when it came out as a 20 then 20 something in New York. And anyway, I'm 10, 15% into the book, but I'm loving every word. She's like an exquisite writer and I'm just like, I'm like excited to find time to go keep reading it. So for, for probably our female audience out there, but I'll give another female

C (41:13)

audience media plug and then I have a really interesting like airline news announcement and then we can wrap. But okay. Margo's got money problems. New show on Apple TV just started last night. It's Michelle Pfeiffer, old school and Elle Fanning and Dakota Fanning in a new series that I find I watched the first episode and really liked and then

B (41:32)

I'm gonna write that down.

C (41:33)

In other news, airlines, you know, United is now launching this. Like you could get the full row and coach and it like expands into a bed. This was in United's big news. Now Air New Zealand and other airlines are starting to prototype bunk beds and economy class. Class meaning literally triple stacked rows vertically on each side that you, you actually don't sleep in the whole flight. You sit in a chair in the regular coach. And for like 500 extra bucks on your long haul flight, you get four hours in the bunk bed to have a nap. And I don't know how I feel about sleeping.

B (42:14)

How do you wake someone up? I have a lot of triple decker,

C (42:17)

you know, and plus like the snoring, I just can't. I'm like, I'm not gonna be in the bunk I next to the snoring dude. And obviously you want the top bunk. There's like so much strategy that goes into this, but I'm interested in all the innovation happening in airline configurations and customer appeal right now. So.

B (42:34)

Fun fact. Yeah, that is a fun fact. Yeah, it's. I think that family row in United is cool. Although I did today. I fly a lot of United, but not enough. I asked United, I asked, I gave my stats to ChatGPT and I'm like, am I close to 1K. And they're like, no. I was like, oh. Woke up 9 Global Surgery. So I'm feeling a little. A little down, but okay, Brit, well, it's been a pleasure. Thanks.

C (42:57)

I know.

B (42:57)

I mean, staying on for the allotted commitments that was made to you that you agreed to in your calendar. But we, I guess, have to thank the dear listeners and viewers who stuck with us this far for the Brit and Jess show. We should have a quorum next week, I think.

C (43:14)

Yes.

B (43:15)

But time will tell. Sam may decide to go to the office again, which will leave us all stunned.

C (43:20)

Yeah, no more Sam going to the office. I prefer him in the pool house.

B (43:22)

I mean, obviously we have to revisit the fact that, like, no one in his team was there and then they were pretending they were going for a walk. I think that's wonderful. So with that, we'll say toodaloo and see you back here next week for another episode of More or Less. Bye.

C (43:37)

Bye, guys. If you enjoyed this show, please leave us a virtual high five by rating it and reviewing it on Apple, Apple Podcast, Spotify, YouTube, or wherever you get your podcast. Find more information about each episode in the show notes and follow us on social media by searching for more or less avemorin, essenlesson. And as for me, I'm Brit. See you guys next time.