Elon Musk vs OpenAI Trial, Google Cloud Surge, Meta’s Blocked Acquisition, Anthropic Winning

Summary

---

More or Less Podcast: "Elon Musk vs OpenAI Trial, Google Cloud Surge, Meta’s Blocked Acquisition, Anthropic Winning"

Hosts: Dave Morin, Jessica Lessin, Brit Morin, Sam Lessin
Date: May 1, 2026

---

Episode Overview

This episode is a characteristically vibrant, witty, and in-depth discussion among four Silicon Valley insiders and close friends, unpacking the week’s major tech news: the Elon Musk vs OpenAI trial, the meteoric rise of Google Cloud (with an Anthropic twist), Meta's blocked acquisition mired in geopolitical drama, and why Anthropic may just be quietly winning the “AI wars.” Interwoven throughout are their irreverent takes on corporate fashion, bets over Mars landings, YouTube gaming, and the relentless pursuit for both margins and memes in modern tech.
Tone: Candid, banter-heavy, sharp, and self-aware.

---

Sections & Highlights

1. Stagecoach, Jorts, and Opening Banter

00:00–03:35

• Brit talks about attending Stagecoach after Coachella and the speedy evacuation due to crazy tornado-level winds.
• Fashion digression: jorts (jean shorts) are in, per Brit, and “summer 2026 trends start at Stagecoach.”
• Dave: “Okay, we better get to the tech, you guys.”
• Light-hearted arguments about OpenAI and fashion — “OpenAI or jorts?” — set the irreverent tone.

2. Elon Musk vs. OpenAI Trial: Legal and Narrative Chess

03:39–12:56

Key Discussion Points:

• Jess’s previous prediction of rapprochement between Musk and Altman was off; trial is underway.
• Trial zeitgeist: Musk learned about key OpenAI fundraising from an Information article — and testified as much, leading to a shout out.
  • Jess: "Elon, thank you for the shout out." (05:13)
• Courtroom drama: Musk giving theatrical statements on the stand, judge's quips (“You don’t have to listen to him. That is not a legal opinion.”).
• Gag order on Altman, Musk, and Brockman using social media during trial (doubtful it holds).
• How long will this drag on? Roughly a month, with detailed debate on legal complexities, nonprofit/for-profit structures, and inside-baseball on OpenAI’s founding.
  • Sam: “It’s great WWE Entertainment, but if they can’t tweet about it, then… we’re litigating the number two AI player now.” (10:07)

Notable Quotes:

• Jess: “This is a case about whether you can steal a charity...” (07:31)
• Dave: “The facts that I've seen don't support that... at the very beginning, you know, should this be a for-profit? Should this be a non-profit? Should we do both in parallel? These are standard conversations...” (09:46)

3. The Real Business: AI, Anthropic, Open Source, and Margin Mayhem

12:56–33:04

Horse Race: OpenAI’s Stumbles, Anthropic’s Rise

• Wall Street panicking about OpenAI missing revenue/user targets; Jess credits The Information for breaking the story first.
• Sam: “I just don’t see the world in which there’s any differentiation on these things. So to me, the whole OpenAI thing is… just narrative weaving.” (14:10)
• Dave: On token economics: “The best open cloud implementations… are on the order of $1,000 or more per month… But that’s $36,000 a year, which seems quite cheap for, say, an assistant.” (15:22)
• OpenAI, Anthropic, and others fighting on price (and luxury) for high-end enterprise vs. open-source models thriving as cheaper alternatives.
• Sam: “Anthropic is really expensive relative to everything else. But… the open source model is getting really good. Really good.” (16:30)
• Growing consensus that AI model commoditization is inevitable; only “luxury token” markets retain margin.
• Tension: OpenAI/Anthropic’s best hopes may be regulatory capture — literally pushing to make open source illegal in the US.
• Dave: “What I'm actually hearing is that they're trying to make open source illegal at the federal level.” (25:39)

Open Source Threat & Policy Maneuvers

• Sam: “This is so much worse than a power company… you can’t ship power around the world globally, right? … create regulatory capture where… you can only use OpenAI in California because it’s the only safe model.” (26:04)
• Sam’s “South Park business model of AI: Start up, cash in, bro down.” (26:56)

OpenAI, Anthropic, and the Margin Problem

• Open models are “good, but not as good” — so the big labs must keep pushing ahead and raise the bar.
• Dave: “No great open source American model that is truly open source. Like Google's putting out a good one, but that's Google, right? ... All American models are massively for-profit. That's a serious community question." (30:56)
• Sam: “...the only answer is going to be algorithmic warfare.” (32:33)
• Discussion about Chinese innovation, open models, potential state-sponsored AI info-wars.

4. Meta’s Blocked Acquisition: Cross-Border Follies

34:31–38:23

• Meta’s attempted agent company acquisition (with a China/Singapore background) gets nixed due to Chinese regulatory intervention.
• Jess (on timing): “They summoned the team back from Singapore and then wouldn’t let them work… this is exactly at the height of TikTok…” (36:17)
• Unpacking the risks and the “lick-the-cookie-lap” investing strategy gone awry.

5. Google Cloud’s Surprise Outperformance: The Anthropic Effect

38:23–45:25

• Google Cloud beats expectations, largely attributed to surging demand from Anthropic (per Jess’s real-time newsroom updates and The Information reporting).
• Britt: “I've been long Google since the beginning. Just going to lick that cookie. Thank you.” (38:26)
• Sam: “Storage companies are going to crash… Energy and memory is where the money is made.” (43:10)
• Dave: “Storage up and down the stack is sold out worldwide.” (43:17)

6. Valuations, Narratives, and the “Fake” Token Economy

45:25–49:07

• Sam: “People are just investing in stories, they’re just narratives… Elon is the absolute king of this.”
• Paul Tudor Jones cited: at 22x earnings, the S&P 500 isn’t investable — so money flows into “narratives.”
• Jess: “How does that end?” Sam: “Badly. Unless it doesn’t.” (47:06)
• Debate about what is “investing” versus unquestioned story-chasing.

7. Anthropic as “Luxury Token Provider” & OpenAI's Strategic Fork

49:07–55:35

• Dave: “Anthropic is the luxury token provider, so…”
• Questioning of business models, regulatory tactics, and the inevitable AI commoditization.
• Anthropic’s business (margin-rich, but always at risk), regulatory capture, and continued “doomer” narrative compared to the Catholic Church's narrative longevity.
• Dave: “The anthropic doomer narrative basically uses the exact same fear trick in order to create a similar narrative. And that maybe they have actually chosen the correct narrative technology to win.” (48:52)

8. Agents, Consumer AI, and the Next Distribution Wave

• Sam shares his experience making custom podcasts for his kids using agent-powered tools: “Now whenever they want to listen to something in the car, I just email my bot… and I will tell you, I've really enjoyed them.” (51:17)
• Concept of future agents: competition, ranking, and “Gladiator: Agent Edition.” (22:36)

9. Pop Culture, Mars Bets, & YouTube Subscriber Hacks

52:34–59:53

• Brit and Sam’s million-dollar bet: first human on Mars by 2040.
  • Sam: "If Elon lands on Mars and hits that comp package, I will have made so much money that I won't even notice the million dollars that I owe..." (53:00)
• Met Gala drama, Michael Jackson biopic debates, family disputes over AI-powered likeness/image licensing.
• YouTube gaming: Subscriber numbers are meaningless — Sam bought tens of thousands for $200. “It’s so cheap… you just got to be every thumbnail.” (57:56)

10. Final Thoughts & Closing Banter

• Dave: “Don’t listen to us for business advice, but do listen to us for tech advice.” (59:35)
• Team jokes about never taking meetings to make money off the podcast; Sam admits he doesn’t know whose credit card is paying for YouTube bots.
• “See you next week!”

---

Notable Quotes, Attributed and Timestamped

• On the OpenAI-Musk trial:

  • Jess (05:13): “Elon, thank you for the shout out.”
  • Sam (10:07): “It’s great WWE Entertainment, but if they can’t tweet about it, then… we’re litigating the number two AI player now.”

• On AI commoditization and margin:

  • Sam (14:10): “I just don’t see the world in which there’s any differentiation on these things. So to me, the whole OpenAI thing is less about it has again, has nothing to do with reality. It's just narrative weaving.”
  • Dave (15:22): "The best open cloud implementations, for example, are on the order of $1,000 or more per month. ... That's $36,000 a year, which seems quite cheap for, say, an assistant or something like that."

• On Anthropic’s positioning:

  • Dave (18:07): “Anthropic is the luxury token provider, so…”

• On regulatory capture and open source:

  • Dave (25:39): “What I'm actually hearing is that they're trying to make open source illegal at the federal level.”
  • Sam (26:04): “...create regulatory capture where… you can only use OpenAI in California because it’s the only safe model.”

• On the narrative economy:

  • Sam (45:25): “People are just investing in stories, they’re just narratives… Elon is the absolute king of this.”
  • Jess (47:06): “How does that end?”
    Sam: “Badly. Unless it doesn’t.”

• On Mars bets:

  • Sam (53:00): “If Elon lands on Mars and hits that comp package, I will have made so much money that I won't even notice the million dollars that I owe...”

---

Key Timestamps for Segments

| Segment | Timestamps | |--------------------------------------------|--------------| | Fashion/Festival banter | 00:00–03:35 | | OpenAI/Musk Trial recap | 03:39–12:56 | | AI Business/Anthropic vs OpenAI | 12:56–33:04 | | Meta Acquisition Blocked (China) | 34:31–38:23 | | Google Cloud/Anthropic surge | 38:23–45:25 | | Valuations & Narrative Investing | 45:25–49:07 | | Agents/Consumer AI Experiences | 49:07–55:35 | | Pop Culture/Mars Bets/YouTube Gaming | 52:34–59:53 | | Closing banter | 59:53–end |

---

Summary Takeaways

• The Musk vs OpenAI trial is pure spectacle: More about personal dynamics, legal technicalities, and public narrative than foundation-shaking impacts for AI's future.
• AI is already commodity: “Luxury models” (Anthropic, OpenAI, Gemini) may fight for enterprise, but open source models are improving rapidly and threatening the entire margin structure. Regulatory capture is the next defensive move for incumbents.
• Narrative rules the day: From corporate valuations to YouTube metrics, winning in tech is as much about weaving viral, defensible stories as core technology or financial fundamentals.
• Google Cloud’s growth is fueled by Anthropic, but also by infrastructure and storage demands, not just AI tokens.
• Meta’s cross-border ambitions continue to hit political brick walls.
• The “AI arms race” is as much about information warfare, legislative gamesmanship, and global policy as pure innovation.
• Bets on Mars, Met Gala tickets, and buying YouTube fame: This is real Silicon Valley banter — irreverent, sharp, and always considering the next hustle.

---

If you’re interested in AI, tech geopolitics, and the ever-shifting sands of Big Tech, this is an episode not to miss. For meme investors, policy wonks, or those just wondering what to do with their YouTube bots, “More or Less” continues to deliver both insight and insider gossip with style.

Transcript

[Britt]

I also want to call out Elon Musk's salary and comp package is now tied to the Mars colonization happening. So I feel like that Bones in my favor for the 2040 bet's just

[Dave]

worried about her bet with you, Sam. I'm counting on Sam's next secret project to pay for this bet.

[Sam]

Yeah, listen, Britt, here's the beautiful thing. If Elon lands on Mars and hits that comp package, I will have made so much money that I won't even notice the million dollars that I owe.

[Jess]

Great.

[Britt]

All I care is about the million dollars.

[Sam]

Well, that is a great emotional hedge to me.

[Britt]

I feel like you just conceded. I feel like you just conceded the bet to me. Dave and grip plus Salmon.

[Jess]

Jess put it all right to the test. More or Less.

[Britt]

Why?

[Jess]

Hello, friends. Welcome to More or less.

[Dave]

Hey, Jess.

[Jess]

What's cooking?

[Dave]

Morins had the best, best music weekend ever.

[Britt]

True. Stagecoach and we got evacuated.

[Dave]

We go to. We go to Indio, California for very different reasons than what?

[Jess]

I go for tennis.

[Dave]

No, Jess goes for tennis.

[Britt]

I know she goes to Coachella too. We're just the country crowd at Stagecoach. But it's not even just country anymore. It was a 90s reunion. We had third eye blind, Hootie and the Blowfish. Who else?

[Dave]

Wallflowers.

[Britt]

Wallflowers. Oh, my God. One headlight. So good journey. But didn't get to perform because of the evacuation and.

[Dave]

Ludicrous.

[Britt]

Ludicrous.

[Jess]

And.

[Britt]

And Pitbull. They stopped believing in Stagecoach.

[Jess]

No.

[Britt]

There were like 60 mile an hour winds. It was tornado style. And trees were falling on people. Some people got injured, some. So they just shut everything down. And then an hour later, the wind stopped. And so half the people left and didn't come back, but the rest of us did. And it was awesome. It was like Stagecoach with half the crowd. So that was really fun.

[Jess]

Brit, I say this with like, not as a leading question, but tell me about, like, how long did it take you to buy your Stagecoach outfits? Like, it just seemed like you were dressed to the nines.

[Britt]

Like I was.

[Jess]

Was this like a whole process?

[Britt]

Oh, it was a process. I consulted with many fashionable friends, very good outfits, but it's a thing at stage. It's the Coachella and Stagecoach. And actually all the fashion brands this week are doing like wrap up reports and leveraging Stagecoach and Coachella fashion in my feeds. They're tuning into new trends for summer based on Stagecoach and Coachella fashion that happen. So I might have been a Trendsetter for summer 2026.

[Jess]

Can you just Cut to the chase, then, because I don't think this pod audience is known as the most fashionable audience on the planet.

[Britt]

That's why I got to bring this flavor.

[Jess]

But tell us, what's going to be on trend this summer?

[Britt]

Well, you either go booty shorts or long jorts. Jorts are back.

[Dave]

There were a lot of jorts.

[Britt]

Yeah. Jean jeans that actually look like shorts.

[Jess]

So there's a jort hit, like mid calf. Where does a jort hit?

[Britt]

It's like, right above the knee. It's almost like a Bermuda short, but.

[Jess]

But that's a long short.

[Dave]

Okay, we better get to the tech, you guys.

[Britt]

Okay. Sorry. We're losing people.

[Jess]

No, I don't think so. I think there are plenty of people who want to know what to wear for summer. You guys are the worst. Open AI or jorts? Open AI or jorts? You want to talk about OpenAI? Let's go.

[Dave]

OpenAI or booty shorts? That's the real question.

[Britt]

Just saying. I could talk about all kinds of categories, so just throw it at me.

[Jess]

Just put it in the chat. I want to know. Summer packing season is here.

[Sam]

Please don't convince my wife to buy jorts.

[Jess]

I already bought some in Paris. I said.

[Britt]

I said booty shorts. Booty shorts or Jordans.

[Dave]

It's like, Sam, I saw you posting hot photos of your wife on Instagram, and we've got some good outfits for her.

[Sam]

I try to be the. The cheering squad here. I will say. Is that my most viral friends post ever? It was pretty funny.

[Britt]

It's cute how much you pose. Top photos of Jess. Sam, I have to say, listen, I

[Sam]

try to be supportive. I do think my wife's attractive.

[Jess]

And you were very supportive.

[Sam]

Sam, I don't know if that's kidding or not, but.

[Jess]

No, I don't know sarcasm.

[Dave]

That sounded sincere. Sam, I'll be your interpreter on this one.

[Britt]

I give it 50. 50.

[Jess]

Brit, zip it. You're not invited.

[Sam]

Brit might have better information than you. Dave, you and I are just like golden retriever puppy dogs. We don't know what the fuck is going on.

[Jess]

Okay, guys, I'm looking at the docket. Here's the docket. We've gone. We've gone old school. So I was wrong. I know it rarely happens. I predicted that there may be a rapprochement between Musk and Altman. I predict. I did. I'm going to own. I am going to own my mistakes. I make plenty of them.

[Dave]

I was wondering about that. I was like, Jess said this was going to be over.

[Jess]

I said I thought. I thought my guess would be and my guess was wrong. Good to know. Good thing I don't guess for a living. I report, but the trial is underway and we'll have to circle back to it because. Or we'll have to get into it because Musk talked about the information from the stand today. Damn. You know, it must be a hot topic when Musk is telling the jury that he learned about an important OpenAI fundraising from the information, not from Altman immediately texting him the article asking him, is this true? So, Elon, thank you for the shout out.

[Dave]

That was today?

[Jess]

That was today.

[Dave]

Wow. I wasn't sure. Was there going to be still, like, more, you know, stand talking and whatnot today or what the schedule is?

[Jess]

You mean testimony, Dave. The testimony.

[Dave]

Testimony, yes.

[Britt]

Stand talking.

[Dave]

I can't talk. I can't. I can't function. Guys, I haven't been to a music festival since before the pandemic, so just give me some grace.

[Jess]

Dave, it is midweek. You are still hungover from a midweek. We are too.

[Dave]

And I don't even drink. And I'm still strong over.

[Jess]

You know, I'm with you. I can. I can relate to that. But no. Well, let's catch everyone up on this trial. So it's midweek. So the trial's been underway since the start of the week. I was going to say the jury was impounded. Is that what you call a jury? You don't call it impounding. And paneled. And paneled. The jury was impaneled. Sam Altman and Greg Brockman attended jury selection. Highly unusual. No one ever attends jury selection. So that was very interesting.

[Dave]

Really.

[Sam]

It's not like they don't have anything to deal with on the home front.

[Jess]

Exactly. That will be topic number two. Then we got to the judge putting a gag order on Musk, Altman and Brockman using social media during this trial. What's the over under on that holding?

[Sam]

Well, here's the thing is like this. This. This trial is not anyone on techmeme is not on Business Insider. No one cares about this trial.

[Britt]

Oh, it's on the Today Show. The. The general public care about the trial.

[Sam]

That's funny. Like, the tech world is like much more focused on other things. That's very funny to me that like, if you don't let them tweet, how will anyone have make money on the news media?

[Jess]

Well, information subscribers are getting their updates. But. But to your point, taken as briefs, not as major news as briefs.

[Dave]

I liked the quote from the judge, the. Do I have to put a gag order on you?

[Jess]

That's true, but I think she officially banned it. She also, like, Musk takes the stand and just starts talking, no questions. And he just looks at the jury and he's like this. First of all, whoever's making this movie should be, like, channeling me at this moment. But like, he was like, this is a case about whether you can steal a charity. And if this case is decided in their favor, every charity in America or something like that is vulnerable. And the judge just says, you don't have to listen to him. That is not a legal opinion.

[Britt]

That's so good.

[Jess]

So anyway, but we're on day, you know, a couple days into it, I think Musk is being cross examined. Brockman and Altman will ultimately take the stand. But still more witnesses on the Musk side. I don't know, guys. The legal experts think Musk has a tough road here. So really, why? Just because of the letter of the law? Not the spirit of it, but the letter of it.

[Dave]

Yes, it does come down to the letter.

[Britt]

How long is this case going to last? Do we know? Is this weeks, months? What are we thinking, days?

[Jess]

It's about. It's scheduled to last about a month, which I think maybe includes some of the expected jury deliberation. I'm actually not clear on that part.

[Dave]

Is this a month? Every day?

[Britt]

What if you're the person chosen for this jury? That'd be kind of sick.

[Sam]

I somehow doubt that Jessica would be chosen for this jury.

[Dave]

So wait, Jess, what is the letter here?

[Jess]

Like, they should just put arbitration lesson style.

[Dave]

Do you have a sense for what the letter of the law is?

[Jess]

Do I have a detailed sense of what the law is? No, to be honest. I mean, other than it's like. I mean, there are plenty of entities that have gone from being nonprofits to having a for profit structure. Remember, technically, this is a for profit nested under a governing nonprofit. If we really need to bring up the flowchart, that is not an unprecedented and I don't know legally how important this argument is, but OpenAI keeps on saying he wanted to convert it into a for profit that he controlled. And he's only. And. And because we wouldn't make this a for profit company that he could control that could potentially be merged with Tesla, he went off and started his own thing. So their point is that he was on board with this strategy.

[Dave]

The facts that I've seen don't support that. That was what. What was going on. There was a lot of, like, in the very beginning, you know, should this be a for profit? Should this be a nonprofit? Should we do both in parallel? These are like standard conversations that, like, literally everybody has at the beginning of any of these things.

[Jess]

But there were those conversations continued past the founding to this dispute.

[Dave]

I see.

[Sam]

I think this is a classic scenario. If you always end up in lawsuits for, like, the thing that no longer matters. It's like OpenAI isn't even the biggest AI company anymore. It's just a classic. The law is always so laggy that, like, we're litigating a thing where the gats out of the bag on all the data being stolen. Like, we're arguing about pennies in terms of what goes into a foundation or not. It's great WWE Entertainment, but if they can't tweet about it, then, like, the entertainment value goes down. And, like, we're litigating the number two AI player now. Like, but what are we doing?

[Dave]

So, Sam, I think it's interesting. The data question is clearly not even part of the narrative.

[Sam]

I know this is the spirit of the letter. Like, we're not even having the right lawsuit.

[Dave]

Yeah, totally.

[Sam]

It's fine.

[Dave]

It is what it is.

[Sam]

The right lawsuit will never happen.

[Jess]

There are some other lawsuits that maybe we should follow. Like the New York Times Be OpenAI, which you're reminding me, Sam. I should.

[Sam]

No, because they're just going to lose. It's like, it's one of the things where, like, this is the spirit of his level off.

[Dave]

But isn't that the one. But, Sam, isn't that the one that you care about? The New York Times one?

[Sam]

Yeah, but it's not gonna. But they're clearly gonna. I mean, like, look, that is the correct intellectual lawsuit. But that lawsuit, I just can't imagine that OpenAI, the cat's out of the bag. Like, it's just not gonna happen that that lawsuit will be successful. It's like a good. It's like standing on good moral ground. But it's like, that's never gonna happen. This is the wrong lawsuit. It's happening. Whatever. It's like it's kind of a sideshow. And then again, the backdrop is like, this is what always happens in the law is like, you're litigating last year's issue. Right? It's like, this is not even the right company to be going after, if you care. I mean, I don't want to go after any company just for the sake of going after them. But it's very funny to me that the trial starts the week that what happens is what we said was going to happen, which is that anthropic passed OpenAI.

[Britt]

But also you could argue that Elon is doing this as, as a way to put a road bump in front of OpenAI.

[Sam]

Yeah, but it's like you're road bumping the second place runner. Like, I don't care.

[Britt]

He doesn't care. He's trying to get XAI to be like a little bit further ahead and buy some Runway.

[Sam]

No, no, he's just. Elon's just flexing power because he needs to support the narrative that Elon always wins, which is, by the way, very smart for him. Like, I'm pro that, but I also

[Britt]

think he's trying to throw a little road bump in front of them. And meanwhile, OpenAI, as someone who follows like the general American, has his own red ones, who's also doing this whole press camp. I'm watching like the press campaigns around all of this because I'm so interested in the marketing and PR strategy everyone's running. And like OpenAI is like, Sam Altman also went on like the morning show circuit. He did in person, he did interviews. He's talking about like the future of AI and how good it can be for the world and curing Alzheimer's and the foundation and how much money they're pouring into curing diseases. And so I think it's just the timeliness of these things is super interesting when you look at the whole narrative strategy across the whole ecosystem right now. I agree.

[Jess]

So let's talk about the business horse race, because it's been all of six and a half days since we've last talked about this business horse race. But there have been some new developments. So the Wall Street Journal finally renewed their information subscription and read the articles that we have been writing over the last eight weeks, which included everything from missing targets of ChatGPT users to debates between the CFO and CEO over IPO timing. They added the information that which is pertinent, which would certainly be newsworthy on its own, that they have also missed revenue targets for ChatGPT recently, although we don't know by how much and we already knew that they missed the user targets, but revenue targets are important, so good detail, good detail. The markets have decided that this is sort of a catastrophe, I think. But Sam, what's your take on this and why are you very bullish on anthropic besides your own cloud code usage?

[Sam]

I mean, I'm not. I think I've been using anthropic a lot right now in cloud code. But like I actually just as I've

[Dave]

been saying all along, like, that's the car he's driving.

[Sam]

It's kind of drive, but I also don't care about the car.

[Jess]

It's not the Miata though. It's not the Miata.

[Sam]

It doesn't matter to me. Like, this is the thing is like, I think again, the story of AI that made these things such huge value nation dollar things was that there'd be some like race to AGI. Whoever wins is worth infinity. And that's clearly not true. And I think being, being the leader for so long and then flatlining revenue while someone else passes you just proves that that's not true. Google's earnings proves that's not true. Like it's just going to be a highly competitive space where like, I don't know where the margin comes from because like I do use cloud code a lot. I love it, it's great. I don't give a shit. Like I'd switch tomorrow. Like, it's all commodity to me. It's super useful. Like I have amazing tools running, but it's just cost per token and how good it is. Right?

[Jess]

Yeah.

[Sam]

And I just don't see the world in which there's any differentiation on these things. So to me, like the whole like OpenAI thing is less about it has again, has nothing to do with reality. It's just narrative weaving. Right. And like cap and cost of capital and like I just don't understand how if you once we saw, I mean, we talked about this weeks ago on the pod when the number you you on Jessica's chart, like, oh, I know how this works. Like Anthropic is clearly going to cross OpenAI. Right. And the point is not that like who's going for place one and going to be the Uber, it's just like. No, actually it's all commodity. Right?

[Jess]

Yeah.

[Dave]

One of the more interesting things that I've been thinking about on this front is that the tokens are so expensive when you're using them at the sort of maximum amount that, you know, people that are on the frontier using tokens to build things, it's really expensive. Like, I've been telling people a lot that the best open cloud implementations, for example, are on the order of $1,000

[Sam]

or more per month.

[Dave]

And you know, when you think about that as like a consumer product, that's wildly too expensive. Right. But when you think about it in terms of replacing actual labor or doing a job which is how a lot of people are thinking about these things. That's 36,000 a year, which actually seems quite cheap for say an assistant or something like that. And so what is anthropic's actual incentive to charge a lower and lower price when actually what they're competing against is the price of labor. Like don't they have an actual incentive to charge the maximum amount per token, not try to actually bring the price down? Which I think is kind of interesting. Like do they have more margin there than it seems? Sam?

[Sam]

I'd also say like the thing for me is very obvious is anthropic is really expensive relative everything else. Like you can have ways to make it more expensive.

[Dave]

That's what I'm saying. Like isn't there more elasticity here because people are thinking about it in terms of replacing labor? Not replace, not being like a consumer.

[Sam]

Well, I don't even think is that sophisticated. The open source model is getting really good. Really good.

[Dave]

Yeah, totally. I work with them every day.

[Jess]

What's the best one? What's the best one?

[Dave]

Gemma, do you think so?

[Sam]

I mean like I think we could debate which one of the best ones but like literally everyone.

[Dave]

Gemma. Minimax, Quen. But depends on your name.

[Sam]

Like with Quen people are saying the embeddings in Quen are better than they are in OpenAI. So this stuff is like moving so quickly and the open source is getting so good that if basically let's pretend that like there's going to be a general good, like I care what I what the cost of tokens is type of intelligence, an open route or whatever. Like that's most of what's going on. Most of what people need is super, super fancy calculators, right? It's not like the fucking. You don't need a PhD to be your assistant, right? And so the vast majority of tokens and spend in the market, the good market, the good marketing America audience is going to be going for cheap Quinn effectively over time and it's going to be because they care about the cost, right? Like that's going to be the.

[Dave]

Absolutely.

[Sam]

There will be assholes like me and Dave who are like actually I want PhD is doing all the work and I'm willing to pay five times as much because I don't care. Right?

[Dave]

And like oh, but we're building software and messing around.

[Sam]

I'm just saying like the point is like there's going to be a high end token market, right? Where you're like, yeah, like I just want marginally better. I'm not sure how much better. And I'm willing to pay five times as much. That might be the only part of the market that makes sense. And Anthropic seems to be very focused on that market.

[Dave]

Yeah, I mean, Anthropic is the luxury token provider, so.

[Sam]

Yeah, so you can be the luxury. I mean, there might be a unthinking luxury token provider that actually does generate margin and maybe that's what they're going for. But this idea of like, you want to.

[Dave]

I don't know. I don't think so, Sam, because I think they are today the luxury token provider. But like, I think you're right. And this is what I'm seeing in the openclaw ecosystem is that the open source models are replacing every single use case where they can, and they're doing it extremely quickly.

[Sam]

Right.

[Dave]

Because why would you pay? And so the, you know, and this is like left, right, all the way up and down the stack. Every entrepreneur in my portfolio is like swapping out open source models for everything. And of course you would. Right, because that's your margin.

[Sam]

Well, right. Right this second, I feel like there's still this world, though, of people who are like, like, I, I've heard this from a lot of entrepreneurs. They're like, in this moment, we're experimenting, this experimental budgets. We're spending way too much on Claude, but we don't care because later we'll just drop the cost. Right? And I think that's like the right mentality. And look, again, I go back to like imagine they were humans and let's pretend Claude is like a triple PhD. That's really expensive. Like you're gonna. But like you're not gonna. You might have them do a little bit of stuff, but it's not the vast majority of your work. Right? And so I think the reality is the vast, vast majority of work, everyone will just like go to cheaper labor and no one's gonna make money on it.

[Britt]

Even enterprise, though. Enterprise and like 10% of luxury consumers is kind of like the TAM for cloud.

[Dave]

But you wouldn't in the enterprise either, Brit. Like you'd want to, you know, you don't want to. You don't want to be this like, what, did I see the uber CFO or something a week ago? Being like, we blew through like 100% of our token budget in 15 days of this month for the entire year. Right. It's like no CFO is going to put up with that shit related to this.

[Jess]

Have you guys seen the. I think Altman said on some podcasts that they're looking at outcome. Outcome based pricing. What do you guys think of this? Is this sort of one of the. I'm not going to say the solves because that's really lame, but one of the solutions.

[Dave]

Wait, so what did he say? I haven't seen this.

[Jess]

I think he said they were just looking. But there's this concept, I've heard more and more people talk about it, which, and Sierra, for example, only charges its customers if AI succeeds successfully completes a customer service interaction.

[Dave]

That's smart.

[Jess]

And it's called like outcome based pricing.

[Dave]

It's not going to give you the best business model. Like right now what they're doing is charging luxury prices for luxury tokens and like getting away with like revenue murder, right?

[Jess]

Well, it's more than that, Dave, because they're also just charging for people who aren't using them because they're doing blanket seat licenses across companies when like 10% are active users and you know, 30% are mid users and no one else is using it.

[Sam]

It's CPM versus cpa, right? Like there'll be some CPA type exercises where you can get.

[Dave]

What do you mean by that, Sam? For everybody that doesn't know.

[Sam]

Well, like a cpm, like you charge for a thousand impressions. Early Internet, hard to instrument. You just charge for a thousand tokens wherever the fog. Right? CPA is like you start saying, no, no, no, I only want to pay for people who like sign up for my thing. And you pay more for the cpa, but like it's more aligned. Right. And like I, like, I think that's right actually.

[Dave]

And I don't know, this is the first time I've heard this, this episode. I think that's probably right that like they'll be, I just made it up. Well, I know, but I think that's probably right. Like what Jess is saying, like people paying for the outcome, like incentivize. Maybe that's like the new agent as a service model where it's like, you know, you pay me based on accomplishing the job I actually tell you I'm doing rather than being like, you're going to pay me some marginal price on top of the tokens.

[Sam]

It's just a race to the bottom though.

[Britt]

I would pay more for the, for the less revs I have to do to get to agent to do what I want it to do too. So more efficient paths to the outcome.

[Dave]

Really?

[Britt]

Yeah. It's so annoying to babysit these agents and be like, no, I asked for this I needed it to do these three things.

[Jess]

Your agent is supposed to babysit the agents.

[Britt]

I know, but like, I. My agent was making me a model today, and then it didn't like, embed the formulas, so I had to go back and.

[Dave]

Yeah, Brit, you need Sam and I's agents.

[Britt]

I have good agents guys and I have like a lot of them, so. Because they mostly get it right.

[Dave]

By the way, this is one of my favorite things that Peter Steinberger always says to me, which is that he would rather that pull requests be prompt requests, because his agents are so good they can just handle any prompt you throw at them. And like, mine are pretty good. Sam's pretty good. Like, it's an interesting new idea.

[Britt]

I want to rank all of our. How do we rank our agents on whose is best? It's like, I want like Gladiator Agent Edition, and I want to know who wins in the end.

[Sam]

Tell them to hack each other and steal each other's token.

[Britt]

I think this is the next big entertainment thing on the Internet.

[Sam]

Oh my God, I hope not. If that's what we're left with. I mean, look, look, I do think

[Britt]

Elon's definitely gonna go there.

[Sam]

I just. Look, I. The smart people I talk to are all like, look, this is classic, you know, people, People always argue. Like people are too obsessed with Adam Smith and the Invisible Hand. They're like, no, you can never make money on anything because efficiency. That's obviously not true. But I would say I think this one's tending towards the cost of energy much faster than anyone wants to admit. And look, that's why, again, like, the cynical take on an open air, some of these people who have been like raising gajillions of dollars and then plowing

[Dave]

it back, it's not just open. I'm anthropic too. I think anthropics peril is actually much more obvious.

[Sam]

Well, they're both. But the basic point is like these companies that went out with these narratives and raised very high priced money to then buy 20 cents of infrastructure. Like, the infrastructure is worth not zero. It is infrastructure. Like you're gonna run something in it, right? But like, it is. It's a commodity business, right? And it's a great one. I really enjoy it. But it's definitely like commodity business with tech multiples. It.

[Dave]

It is interesting. One of the things that I've been thinking about a lot is how much all of these things are just abstractions for inference. Like the more the sort of deeper you get into building harnesses and building you know, stuff that kind of sits right in front of the inference layer, the more you realize that all things are just like, terminal UIs and things for, like, pure inference costs, right?

[Sam]

In the end of the day, we're just multiplying big numbers, man. It's just a big calculator.

[Jess]

So if you're OpenAI and anthropic, what is your pricing strategy in the enterprise? Because you know that open source is getting to parody on many things.

[Dave]

It's not clear. That's obvious, Jess. Like, I actually think this is not being talked about enough.

[Jess]

But didn't you just say open source? But wait, I thought you said open source is good.

[Dave]

It is, but what I'm saying is that, like, I don't know. I don't know if anyone talk. I don't see this as in, like, the public narrative and conversation right now. Like, what are they going to do? Right? Like. But I'm sitting there.

[Jess]

Oh, no, but that's why we're having it, because we're. We're ahead of.

[Britt]

We're ahead of the times.

[Dave]

I'm not totally sure. I mean, I sat there. I sat there at this. I got invited to a founder event last week to give a talk, and. And then I watched Lightning talks from like 10 to 20 founders in a row about their AI implementations. And like, these are like, very sophisticated businesses. We're talking businesses on the scale of, you know, anywhere from tens of millions to hundreds of millions in revenue per month. And everyone is trying to switch out as many open source models as possible across their entire stack. Like, that's like the active conversation switch

[Sam]

in as open source models. Yeah, the Quinn. Yeah, Quinn's on the rise.

[Jess]

No, no, I think you misheard me. If. So you're anthropic and OpenAI and open source is a real threat. So what do you do from a pricing standpoint?

[Dave]

Well, what I'm actually hearing is that they're trying to make open source illegal at the federal level.

[Jess]

Oh, my goodness, Dave, you just broke big news on the pod.

[Sam]

The classic regulatory capture move.

[Britt]

More litigation.

[Dave]

I'm not. This. I'm not telling. This is not. This is being talked about on Twitter. This is not just me.

[Sam]

There's like two models. One is if you have a utility.

[Britt]

Jess, break it.

[Jess]

I just did.

[Sam]

I mean, if you have a low cost, if you have a competitive utility, you can, as I've pointed out and we've talked about before, is like, this is so much worse than a power company. At least with a power company, you can't ship power around the world globally, right? This is like being in an unregulated utility business where you have no regulatory barriers and no physical barriers. You meet with everyone globally, you fucked. Right? And so, yeah, of course, one is create regulatory capture where we say okay instead of PG and E. It's like, you can only use OpenAI in California because the only safe model, right, like that's option one.

[Dave]

Exactly. And this is very real, by the way, in the United States.

[Sam]

So. So it's that or if you have a utility that you convinced everyone was worth tech multiples, there's the south park answer, which is, what is it? You start up, you cash in, and you bro down. Like, you got to get out.

[Jess]

Okay, stop splitting that a little bit more, Sam. Give me a little bit more.

[Sam]

This is the south park model of business building. You, you start up, you cash in, and you bro down. And like, these guys are moving from cash in to bro down right now because, like, they're.

[Britt]

What is that? What is bro down mean?

[Jess]

What can you Nobody. I not. Can you say it again? Can you draw your analogy more clearly, guys?

[Dave]

One of the ways. One of the ways that I know how things are going in the city with the cashing in part is what kind of cars I see on Highway 1 every day.

[Jess]

Just way most. Not on Highway 1.

[Dave]

No, no, no. I saw a, like, diamond encrusted supercar on Highway 1 this morning. Like, it was literally full of gems.

[Sam]

Yeah. Like, this is called Miami Crypto 2000. Whatever.

[Dave]

I mean, I've never seen anything like this. I literally was like, what was that? It was so bright.

[Sam]

The problem is. The problem is with all this. This is like the thing that kind of blows everyone's brains, is two things can be true.

[Dave]

When did this become L. A?

[Sam]

Two things can be true at the same time. Thing one is, this stuff is so fun to build software with, and I have better software for myself than I've ever had, and I'm loving it. Right. Two, it's actually a bad business. Right? And, like, that's very hard for people to wrap their heads around that both things can be true at the same time.

[Dave]

I mean, at least we're talking about both things being true at the same time now, Sam. Because we know people are tired of us telling them that it's a bad business. So we have to figure out how to get from here to there.

[Britt]

You've been saying this every episode. I don't get what's new about any of this.

[Jess]

It's a little more nuanced because I think it's very possible that anthropic and OpenAI could prove to be bad businesses. And by bad businesses, that means very low margins, not that much revenue. But this era is. We. If this were, you know, a year before the META IPO, or a couple years before the META ipo, we'd be talking about whether META could ever make any money.

[Sam]

That's not true at all. META made money the whole way through. That section was. META was highly profitable from like almost year two.

[Jess]

Okay. We would be talking about a scale of numbers. Fair enough. That was much smaller than 20 billion, 30 billion in ARR. So I do think you have to give these. You can look at the existential landscape and the competition and everything and say it is still an open question how good these businesses are. I think that's true. But you also have to give them credit for the business they built.

[Sam]

Can I sell you a dollar for 90 cents? I mean, the problem is, is that,

[Jess]

like, well, saying that the valuations are the problem.

[Dave]

But, Sam, something's happening. You know, I think Jess is right. There's like something happening worth analyzing more deeply. And like, look, here's the reality of the open models, right? Like, they're not as good. Right. Like under 70 billion parameters.

[Jess]

You guys just said you named three that were good.

[Dave]

Let me finish. Under 70 billion parameters. You know, they don't. It's very hard to get them to talk for any number of turns. It's just uninteresting to talk to for a long period of time. Their tool calling is like, it's okay,

[Sam]

you don't get there.

[Dave]

It'll get there. But like, you have. Sam, each time the big models come out, all of our expectations raise. Right. For what's possible at the high end. And so you have this thing where open source is also always going to be lagging the frontier. And so it's not going to be like a total replacement. And so just that's the answer to your question, which is what are you thinking at the, you know, at the strategic level if you're a frontier lab? Well, you're trying to make the next model happen, Right. Because that raises the bar for everybody in the whole world as to what they expect is possible. And then on the open source side, you're trying to do regulatory capture, at least in the United States, right?

[Jess]

Yes. Which is where we were starting this. Yep.

[Dave]

Yeah. And I think that's like, actually a really interesting question because my experience from working on an open claw is that the Chinese innovation in Open source models is really rapid and really impressive. Like these guys are doing extremely good work extremely quickly. And by the way, there's no great open source American model that is truly open source. Like Google's putting out a good one, but that's Google, right? And so I think it's an interesting like foreign policy problem that like we have no American models and all of the American models are massively for profit. And so I think that like, that's like a really serious community question we should be answering. And people will dive into this and be like, oh, this is about ideology and safety. Like there should be no American model because it's not safe. Right? And so there's like all these aspects of this that I don't know, kind of aren't being talked about and are the real things.

[Sam]

You know what we should do? We should do what Europe did to us on the Internet. We should put some GDPR shit all over the Chinese models. You know, some things to click on.

[Jess]

Okay, so do we think that Anthropic is going to try and block open source in the US or that that will even happen? I think it's going to be.

[Britt]

That has. Is, is there precedent for this in any other like industry?

[Dave]

There's literally bills that have been proposed

[Sam]

as we can't even block TikTok in the U.S. how do you think we're block this?

[Jess]

Right? It's like we haven't talked about TikTok in a while.

[Sam]

It's like, I just. I hear you. I actually, look, I actually could get behind the theory that like there are dangerous models and like what sandbox. I have no idea how you actually like implement any of that, but it is an interesting.

[Dave]

That's the problem, Sam, is it's impossible to implement and you and I both know it. Like when you can just download the weights for these things onto your laptop, like it's impossible, right? And so you have to think about policy from a totally different angle. Like it just doesn't.

[Sam]

I actually think you don't think about policy. You know what I think you probably think about just in terms of how this plays out is like assume you just, you're not gonna get borders for this type of stuff. The only logical conclusion is AI to AI warfare, right? Where you're like, we need to launch like Anthropic launches a full assault on destroying the open source models and tricking them and just. And making them worse, right? Like there has to. Like that's actually the logical conclusion of this, is that you can never put up Regulation will never work, but war does.

[Dave]

Interesting. I'm gonna have to think on that one.

[Sam]

Like, I think that's actually the thing about it. Like this. The Neal Stephenson book version of this would be that it's like, we can't defend ourselves from cheap open models and they're subverting our country. Forget like, you think. You think that the Chinese social engineering budget, like, around the US Is subverting the US Wait till they get into the models, right? You're like, okay, our only logical solution is to assault.

[Dave]

That's what they are doing. I mean, that is what they're doing, right? Like the benchmark. There were a bunch of benchmarks that came out two, three weeks ago that show that Quinn specifically, which everyone is saying is the best. The best. The best I'm using it, is by far the most biased. Like, the. The model is extremely biased.

[Sam]

Yeah, I get it, I get it, I get it. I just think it's a really, like. It's kind of like the algorithmic social media problem on, like, massive steroids, right? That's distributed so you can't even stop it. Right. And so it's like, someone told me this week, I don't know if this is true or not, but it was an interesting fact. I was having a meeting. They're saying that the Chinese are running an $80 billion a year social engineering budget for. Against the US right now, right? So the US has like an $8 billion.

[Britt]

Like, what are they doing? Like, TikTok.

[Sam]

Oh, so like, we're running like an eight, like two orders of magnitude less in terms of whatever the relevant unit is. That's like, again, like, who knows if this is true or not of the details of it, but the idea is like, this is very organized and at scale. And like, I'm like, oh, this is an even bigger battlefield. And you're like, you just can't. There are no borders. And so the only answer is going to be algorithmic warfare.

[Britt]

Does this. Does this play into the Manus block at all that came out this week?

[Jess]

I'm sorry, but this was the most obvious news.

[Dave]

It was, but it took a while. It's surprising how long it took, right? Like, because presumably they would have done some amount of integration already. And so how do you unwind that?

[Jess]

They were trying, but they summoned the team back from Singapore and then wouldn't let them work. I mean, for people who are following

[Sam]

suit, they actually went back from Singapore to mainland China and they went.

[Britt]

They had to, they had to, they had to.

[Dave]

This was like Spy game stuff, Sam.

[Jess]

So for people who haven't been following Meta acquired. Well, let's go back. I remember Benchmark invested in this Singapore based but Chinese run and did a

[Dave]

big lick the cookie lap on it agent company.

[Jess]

Right. And they were like, they were. So they did a lick the cookie lap and it was like they were like, oh, guys, like, cross border investing's back. And this is the height of. That's how benchmarks be.

[Sam]

No, I don't know.

[Dave]

It's actually, you took a risk, you lose.

[Jess]

And this is exactly at the height of like TikTok. Like Cross border investing wasn't back. There was no window. Like, not at all.

[Dave]

No, it wasn't. It was extremely risky.

[Sam]

It's only back of Elon says it's back.

[Dave]

This. This investment was extremely risky. Like at the time, this was extremely risky.

[Jess]

But of course they made it and

[Dave]

then they like, it was a hail Mary and now it. It fully missed.

[Jess]

Yeah, it has. Yeah. And. But anyway. And then the company chugs along. Then you have Meta and we'll get to earnings later because I know you guys listen to us for earnings is the only reason. But you have Meta.

[Dave]

I mean, I sure do.

[Jess]

Is how Dave knows what companies are doing. No. So Meta is on the hunt for a new agent strategy, a new business model. Just like new, new, new, new things. It's talking to a lot of places about acquisitions. They get very excited about Manus. Okay, interesting deal. You're watching this and you're like setting your clock for when China is going to interfere. And of course they interfered and they did quick, soon after in terms of sort of challenging the deal, calling back some of the key team to China. So they were in clear integration limbo. And now that the deal is off and they're sort of unwinding it, but

[Sam]

they did integrate it already, though, which is kind of wild.

[Jess]

I think some of it. I think some of it.

[Britt]

What does that mean, though? Like, does money go back to investors? Are they like, yes, code go back. How do you unwind it?

[Jess]

These are great questions. I don't know the answers to all of them. I've seen some headlines.

[Sam]

Brit, did you change your hair color?

[Britt]

I did. Thank you. Back to my roots.

[Jess]

Yes. Sam, we talked about this earlier. Well, let's circle back to that topic.

[Sam]

I like it.

[Jess]

Let's circle back. It looks great. Brit, because of you, I'm gonna start brushing. I hurt my shoulder, so I can't brush my hair. So that's why I have to apologize to everyone for my appearance.

[Britt]

Just wing it. It's fine.

[Jess]

Britt is just like crushing it.

[Britt]

This will convince people to tune into the video stream who are only on audio.

[Jess]

Guys, you can just, you can rate Britt's versus my hair every Wednesday and we'll see one of these days I'll get it done. I don't know the particulars of it and you've got to give meta points for trying. I mean you gotta give points for trying. But I do think back to being told that cross border investing was back just as Nvidia chips were getting banned in China and deep Seq researchers were having their passports confiscated as reported by the information. And here we are. So we're back to exactly where we knew we were. Guys, Google. This is when we turned to the pod ripped on an earnings report of greater than expected profits.

[Sam]

I can't believe I don't know Google stock.

[Britt]

It's my alma mater and I've been long Google since the beginning. Just going to lick that cookie. Thank you.

[Dave]

The good news is every time Sam told said he loved Google, I bought Good Dave.

[Jess]

Way to go smart man.

[Sam]

At least one of us is making money. But it's like there's like two things I regret in life. One is I remember like early on financially in the public markets recently, early on, do you remember like when there was this really funny situation where Nancy Pelosi's husband right around New Year's bought like an insane number of Nvidia calls and there's this like joke about how Nancy's Pelosi husbands is like the greatest trader in the history of the world. And I was like, that's actually pretty good. I don't own anybody's stuff. I should probably copy the Nancy Pelosi tweet. And I really thought about it and like I did the math on it's insane. Like I think he probably made like $50 million on like a million 2 million bucks out. Like good for him. And the second is this fucking Alphabet thing where I've been calling forever that people are miscounting Alphabet and yet I don't pull the trigger because I'm bad at this.

[Britt]

By the way, Google IO is coming up very soon and I think, I think there's going to be a ton of new Gemini stuff that's coming and we haven't talked about Gemini in a long time. I know we talked about Gemma just

[Dave]

now, but what a formidable company. It's such a formidable company. It's craz.

[Britt]

I think by the end of this year Gemini is, is Much higher in the rankings.

[Dave]

Even just the returns that they're going to see from their SpaceX and anthropic investment is insane.

[Jess]

But you do, because you see an AI overview or two if you're still ever.

[Sam]

No, to be clear, I just, I use Google. It's funny to me, like the entire chatbot paradigm is just so uninteresting. Like I don't use any of them anymore. Right. Like I'm deep in other stuff. But, but it does seem like from a Good Morning America perspective, Gemini is in a good place.

[Jess]

Good Morning America.

[Dave]

Which by the way, Sam, I think that's one of the big, that nobody's paying attention to right now, which is that there's only something like, I think I saw it was like 3 to 4 million Codex users, which means there's also in, you know, single digit millions cloud code users. Yet everybody in town thinks that this stuff is broadly out there and that everyone's realizing what's going on.

[Britt]

It's game over. Anthropic's winning everything. Nope.

[Dave]

There are like 6 billion people that use software in the world and there's only 50 million total software engineers, single digit millions doing all this stuff. Like there's literally like a 5 billion person market out there for figuring out how to bring agents to the normal person.

[Sam]

Or maybe you don't experience them and don't use them that much.

[Britt]

Also, by the way, I don't know, there's probably Claude actually projects. There's 1.8 to 2 1/2 billion people with Google accounts in the world. So like 30% of the global population have Google accounts.

[Sam]

I just asked, asked, I just asked, what's the. So basically 50 to 100 million people use terminal or CLI interfaces at least occasionally. So maybe it's 1 in 10 right

[Jess]

now, Dave, but what, let's see. So the reason Google, Google's earnings outperformed is really cloud growth. And the reason I think Microsoft that had decent numbers but the stock is down is because cloud growth I think was around expectations. But like expectations are so high these days that like if you just hit them, investors aren't happy. But like I, you know, have a team of 50 reporters who could presumably help answering this question. But I still don't understand how Google cloud is growing so fast. Like are they. But like what are they bundling? Is it the TPU moment? Is it, yeah, it must be the Gemini moment.

[Sam]

Or is it just honestly, or is it just like, you know, the funniest things. I looked through this recently, you know My number one line item changes in terms of my costs of like the infrastructure I run for myself. Since all this agenda. It's actually curious. Dave, you're the only one with the prayer of getting this right. What do you think I spend Digital Ocean.

[Britt]

Thanks. Thanks Sam.

[Sam]

But what it's. It's in digitalocean. Because I love Digital Ocean.

[Jess]

I got it right.

[Sam]

No, no, no, but it's a specific thing in digitalocean. Like where am I actually spending money?

[Jess]

I got the biller on the credit card.

[Britt]

It's not a token.

[Sam]

Where am I actually spending money? It's not. It's definitely not tokens.

[Jess]

Memory storage.

[Sam]

Yeah. Basically Jess, I will give it to

[Jess]

you two four two.

[Sam]

The answer is it's postgres. My postgres bill is the thing that's gone up the most. It's databases. So that's the irony of this whole thing is like. Like it's not the actual place that someone like me, and again, I'm a pretty aggressive user and I've built a lot of apps and like I think it's great but I use it in a very specific way. The thing I'm actually spending a ton more money on, which the Wall Street's already figured out because they priced this in, is actually storage and database. It's not the actual agents because like all of a sudden all this data that used to be not that useful is super useful. And so I'm like, okay, well now I have to save everything. And like that's the place where there's actually leverage in the system. So you know, I know they've already run up a bunch. Another mistrading opportunity, if you understood is like it's the storage companies. Storage companies are going to crash.

[Dave]

Not just pure storage, but I mean you're seeing this, right? Like storage all the way up and down the stack is like sold out worldwide.

[Sam]

It's Seagate. It's like people like that like that

[Jess]

are really by the way, as a. No. So as the like dumbest user of cloud code in the category of least experienced. Technically, that is the thing that constantly surprises me about these services. Like the thing that like my brain thinks they should do automatically that they don't is remember things. Like it is counterintuitive to me that I actually have to tell them that because no other service that I use operates that way. Like I type in a Google Doc and like the concept of saving anything was erased a decade ago. But like we're back to it again and I forget to do it a lot and it's really annoying. Can you ask it to save everything it does automatically?

[Sam]

Yeah, I do that. I have it log everything. I also have like every single request I make, I have it log all this data. I actually hilariously I Briefly was spending $100 a day on the Twitter API by ACC because I was having it log everything and then I had to like be like oh you don't need to do quite this much. But yeah, I mean like, like look, the reality is it's kind of wild that Western Digital Seagate, these stocks are ripping. They're up like you know, thousand percent. Right? Not 1000% but they're up 100x100 101x in like the last month. Right. Because everyone's kind of coming to Jesus that like actually this it's going to be energy and memory. Energy and memory is where the money is made.

[Jess]

Oh guys, here's some real time analysis from the information team. Instead of just shooting my mouth up, apparently Google Cloud is up because of Anthrop. There you go. So maybe this is all a little more circular than we thought.

[Britt]

Well and they just did a big. They just put 40 billion into anthropic too.

[Jess]

They're just servicing them or

[Dave]

they're serving TPU's into their cloud basically.

[Jess]

Yeah. So they also are investing more as is Amazon and we have to talk about my favorite battle of the week guys. But what's also happening and I don't know if this is the case with the Google Cloud deal but Anthropic is so underserved in compute that they're buying like in last minute deals and on the spot market and all that kind of stuff is bad for their margin but is good for the provider. So and I don't know what percentage of Google's so called beat is, is that, is that. But if, if there it could be some anthropic overage in there which is really interesting.

[Sam]

Well this is like the big thing about all these prices and like you know, these values. Like I do just think you to understand that people, this has been true for a while but like there's just, people are just investing in stories, they're just narratives and how strong the narrative is. They have no financial backing like to reality in most cases and like Elon is the absolute king of this in terms of like what' on. But it's like there was a, did you see the Paul Tudor Jones thing today that was floating on the Internet?

[Dave]

I did see that.

[Sam]

I haven't actually watched the whole thing Yet, But I watched part of it looks excellent and he's basically like, you can't invest in the S&P 500. Right.

[Dave]

Like his basic thing is, I wanted to ask you what you thought of this, Sam, and if you saw it,

[Sam]

it makes total sense. He's like, look, you know, yes, on a 100 year basis, like the S&P 500 has been a good bet, but at 22 times earnings or whatever, again, I don't want to misquote it.

[Dave]

Although I'll misquote it like it was 21 or 22.

[Sam]

It's like insane. Like, it's like we're so far off center that you need the six to get to a positive. You need to be at six times every once in a while. And like we're so out of whack. So then what happens? Everyone's like, well, I have a bunch of money, it has to be stuffed somewhere and I can't put it into the economy because it doesn't make any sense. So I'm just going to store it in stories. Right. And the way you store it in stories is with Elon or with the anthropic or whatever. So number.

[Dave]

But that's also not good investing, which I think is perhaps what's dangerous right now. Right.

[Sam]

Like the question is what's dressing, what's investing and what's trading anymore.

[Dave]

If you can't be long the S&P 500 right now, then like think that's

[Sam]

the thing is that you can't forget value investing. Like you can't be long financial returns in any logical sense. And so as this thing spins out of control, you're just long narrative. Right. And like everyone's a trader.

[Jess]

Yes. How does that end?

[Dave]

It doesn't seem good.

[Sam]

Yeah, badly. Unless it doesn't. Like that's kind of the thing. Right?

[Jess]

Like badly unless it doesn't.

[Dave]

I think that is the truth is that people are betting to the upside of that trade.

[Jess]

Is this what we would call a risk on market? What would we call this? Does my using.

[Dave]

No, it's not risk on. Yeah.

[Sam]

Okay.

[Dave]

Because the risk on the fundamentals aren't there. Yeah, the fundamentals aren't there. And so the only possible explanation is that people have convinced themselves that the demand for tokens is infinity and will continue to go up and so put all money in that narrative. Like that's basically what's going on.

[Sam]

Well, and the reality is the price of any of these things you're investing in has nothing to do with reality. It only has to do with the price that it already was and then where it's going. And therefore it's all just on story. Right. Which is really scary because there's no global liquid fundamentals against that. But I also. Some of these stories, like the Catholic Church did okay for a few hundred years on stories like, you can have these like story bubbles.

[Dave]

Yeah.

[Jess]

We had. My colleagues, we hosted an event on the New York Stock Exchange on Monday all about where we are in. And it's specifically on financing, AI and data center. So wonky stuff. Good sentiment stuff. And everyone's like about to burst, but we don't know when.

[Dave]

Sam. I hadn't. Where did Sam go?

[Britt]

Sam is. There he is.

[Dave]

I had not considered the idea that the Catholic Church's narrative that was so successful for multiple whatever for many hundreds of years was a doomer narrative. And that the anthropic doomer narrative basically uses the exact same fear trick in order to create a similar narrative. And that maybe they have actually chosen the correct narrative technology to win.

[Jess]

Dave, I'm going to make an observation. You really want to talk about anthropic?

[Dave]

I do.

[Jess]

Is there more you want to say right now?

[Dave]

No.

[Jess]

Okay. No, it's good. But like, I think it's reflecting some sentiment that is broadening in the industry.

[Dave]

I think it's just good to be balanced on all these things.

[Jess]

I see the good old balance hack. Yes. Okay. We got through earnings, we talked AI pricing, what we got, our trial. What else is happening in the world right now? I will say for tennis fans. I interviewed Maria Sharapova this week, five time Grand Slam champion, and she didn't want to talk to me about. I wanted to talk to her about tennis and the status of Alcaraz's wrist. She laughed at me. She does not care about the status of Alcaraz's wrist. Hopefully it'll be fine for Wimbledon.

[Britt]

I don't even know what that means.

[Dave]

Alcaraz's. Alcaraz is hurt, Brett. He's been out.

[Jess]

He's out. He's out for Roland Garros. But anyway, she has a new podcast that we should plug here. It's called Pretty Tough and I think it's pretty good. It is, yes. A sort of girl bossy women in business podcast. I'll tell you why in a second. But it's. I listened to the first interview and it did not feel like.

[Dave]

Why do athletes need podcasts?

[Britt]

Everyone's training. Everyone's doing a podcast.

[Sam]

Listen to all of them.

[Britt]

Dave, I don't have Time for that many podcasts.

[Jess]

I used to be. I, I probably listen to three podcasts a week and I just rotate.

[Britt]

I want an agent that can like take the 20 podcasts I care about and like cut the best parts of them into one good podcast.

[Jess]

But that's not why I listen to podcasts though. I listen to them to kill time. I listen to them for entertainment. No information.

[Britt]

Yeah, but you just. There's so many.

[Sam]

I gotta say, you should for your boys. You guys, actually, I'll send you the link. LessonBoyPodcast. AI. I actually now I've listened to more podcasts. I have a long time because I just decided to make them for our kids. So now whenever they want to listen to something in the car, I just email my bot. I say make a podcast on how Minecraft was made or how the history of Nintendo or how EA Sports FC Mobile 26 was made, or Jetpack Joyride.

[Jess]

It's pretty good guys, I gotta say.

[Sam]

I've learned.

[Jess]

Would you rather listen to like the same, you know, shitty kids podcast or be like, no, I do want to learn about this thing. It's really good.

[Sam]

How baseball 9 was built is so fascinating. And like I would never.

[Britt]

It's kind of like the acquired podcast, but lesson boy style.

[Sam]

So I just basically. But I was like, I got sick of listening to our kids stupid stuff. So I like literally now we just get in the car, I'm like, what do you want to learn about? And then I just prompt it. Make a podcast on blank and within five minutes it's boom. And I will tell you, I've really enjoyed them.

[Britt]

I like the one about Tetris, the inventor of Tetris. I want to do that one.

[Sam]

It's great. They're all like five to 10 minutes and they're really interesting.

[Jess]

Sam used 11 Labs voice, so it's sort of him but with some modulation. And I actually think putting some in our kids voice would also be cute for them.

[Britt]

Like they're like, that's a good idea.

[Jess]

They want to hear.

[Britt]

And then your kids voices are going to be uploaded to the Internet forever.

[Sam]

Yeah, I don't think we want that for their banking security, but okay, sorry, we don't.

[Jess]

No, that's true. I didn't think about that.

[Britt]

But just Sam's. Sam's voice for his banking security is fine though.

[Jess]

I mean his voice is on the Internet. That's like unfortunately.

[Britt]

But like speaking of media and content, can we talk about the New York Times did the 30 Greatest Living American Songwriters Roundup.

[Jess]

I Watched all the Taylor videos. They were excellent.

[Britt]

I. Me too. We also watched Jay Z. I also want to shout out Dolly Parton, Paul Simon, Bob Dylan. There's so many good Bad Bunny. I thought that it was a really great sort of collection and package of content that I highly recommend people check out.

[Jess]

Completely agree. I thought it was great. And I. You know Taylor Slade, like, she was so interesting.

[Britt]

Yeah, she did good. I also want to call out Elon Musk. Is his salary and comp package is now tied to the Mars colonization happening. So I feel like that bodes well in my favor for the 2040 bet.

[Dave]

Britt's just worried about her bet with you, Sam.

[Britt]

20:40 Bet one human steps on Mars is the bet. Million dollars.

[Dave]

I'm counting on Sam's next secret project to pay for this bet the other direction.

[Sam]

Listen, Britt, here's the beautiful thing. If Elon lands on Mars and hits that comp package, I will have made so much money that I won't even notice the million dollars that I owe you. So, like, we're all good. We're. That is a great emotional hedge to me.

[Britt]

I feel like you just conceded. I feel like you just conceded the bet to me, so I'll take it.

[Sam]

No, it's an incredible emotional hedge. I'm either getting paid need a million dollars by Brit, or I'm massively rich by 2040.

[Jess]

I just really don't. Sometimes I don't have things to say, so maybe I shouldn't say them.

[Sam]

Well, you can say mean things. I can take it.

[Jess]

No, no, it's not. It's not about that. I have more incoming reporting confirming this really is an anthropic bounce.

[Dave]

Bounce. Oh, on Google.

[Sam]

For Google, bounce would imply was down. I'd say it's a moon.

[Jess]

It's a anthropic. Okay, well, what. What do you guys morons have cooking this weekend? Like, what. What's happening? Any more concerts?

[Sam]

Wait, before we do that, can I tell you one other fun anthropic story?

[Jess]

Sure.

[Sam]

I had to text. I recently texted someone who used to work for me because someone was trying to get something fixed at Anthropic. I was like, oh, yeah, this person works at Anthropic. And I texted him, this is a person who worked for me a long time ago. And when I pulled up the text thread, it was hilarious because the last text that I had with him were like 18 months to two years ago. And he was deciding between two offers, one between Anthropic and the other, like, it was a Smaller startup. And he's like comparing the comp packages and what they were. The valuation of anthropic topic. So I have all this data from him and it's really funny because I'm of course like, well, I think they're both overvalued, blah blah, blah. And then I'm doing the math. I'm like, good for you. You just made a lot of money.

[Jess]

I like that. Anything else cooking? We've got birthday parties and a tennis tournament over here.

[Britt]

We've got birthday parties and the Devil Wears product too. And yeah, maybe I'll go see that.

[Jess]

Yeah, maybe I'll go see that. I think I've seen and the Michael

[Britt]

movie, by the way, the Michael movie is apparently crushing. Even though the ratings are. It did like $100 million in opening weekend.

[Sam]

What is a Michael movie?

[Britt]

The Michael movie is about Michael Jackson and his nephew or cousin or somebody actually plays him. But the whole Jackson estate is like split on this because you know the narrative. They actually stopped the narrative by the 90s. So they didn't go.

[Dave]

No way. They are not split.

[Britt]

They didn't go into the craziness about what happened later.

[Dave]

They are super excited about how much money they're making.

[Britt]

No, there's some of the Jackson family that claimed they were molested by Michael like his, his nieces and nephews and things. And so they're like this. The other part of the family is just cashing in and this guy was so messed up and. But then they're just telling the good part of the story. So anyway, I want to see that one too.

[Dave]

Who actually owns the ip? I mean like all of the, the Jackson estate. Do they really? Yeah, there's no private equity or no, no big, no large scale family offices involved in this capital stack. Like I don't believe that the Val

[Britt]

Kilmer estate is the first to legally license his, his name and likeness into AI and he's going to be in an upcoming movie.

[Jess]

Oh, guys, the Met Gal is next Monday. Have you heard the controversy?

[Britt]

No.

[Jess]

The Met Gal is no longer cool. I read it.

[Britt]

Why? Because Anna's out.

[Jess]

Well, Anna's like a toe out the door. I don't think she's really out the door, but there are a lot of people upset that Lauren Sanchez and Jeff Bezos or Lauren Sanchez Bezos and Jeff Bezos are hosting this. And there would be a lot of protests, a lot of anti tech sentiment and apparently they're not selling tickets. I would still buy a ticket for the record, if allowed to buy a Ticket.

[Sam]

You still sell tickets to this?

[Jess]

Yes. You pay a lot of money to go.

[Sam]

I thought you paid for a table.

[Jess]

You can do either. Yes. They're not selling tables or tickets.

[Sam]

What are they selling?

[Jess]

They're not sell. They're not hitting their goals. They're not hitting their goals, according to this article.

[Sam]

Oh, I see. I see. People aren't buying. They're for sale. I'm like, what is the business model?

[Jess]

It is. It is a fundraiser for the Costume Institute. Anyway, I will be on the Instagram looking at the outfit, so I'm doing just fine with it.

[Britt]

I'll be on.

[Jess]

I just would like, you know, just to catch everyone up on that fact, so. Okay. Well, it's nice chatting with everybody. We have a whole quad here for the whole episode.

[Sam]

Do you think Jason's son will go to the net gala next?

[Jess]

Jason's son?

[Sam]

The Tron guy?

[Britt]

I don't know what you're talking about.

[Jess]

Oh, that guy. I don't know.

[Sam]

Justin son. Yeah. He backed the Tron, and then he bought all the Trump coins, and he, like, there's all. Well, I just feel like he'll be the next person.

[Britt]

I like when Sam does pop culture predictions. Probably. Yeah, Sam. That's probably right.

[Jess]

Well, dear listeners, thank you for joining us for another episode. We would love your feedback. May take it, may not. But we appreciate it because that makes

[Sam]

the number go up.

[Jess]

And make number go up, by the way.

[Sam]

You guys see that I bought us a bunch of subscribers on YouTube.

[Dave]

Wait, you did what?

[Sam]

We have, like, 30,000 followers or more. Maybe 40 now.

[Britt]

No, we have 57,000 subscribers.

[Sam]

And it's like, what I figured out.

[Jess]

How did you follow that?

[Sam]

It's like, welcome, random $300. Here's the thing. It turns out I tested this. If you put up, we should probably just buy, like, a few million. It's so cheap.

[Britt]

Well, then we can charge for ads.

[Sam]

Yeah, exactly.

[Dave]

That actually makes me wonder about all of you to.

[Sam]

No, no. The subscriber number on YouTube is clearly fake because I just, like, literally paid 300 bucks now, Dave, I did figure something out important, which is I just tested four. I just. All I did is hit the promote button on four videos. Unsurprisingly, Brit and Jess massively outperform us on clicks.

[Dave]

Perfect.

[Sam]

That makes sense.

[Dave]

We're putting you guys in front more 100.

[Sam]

Like, you guys got to be every thumbnail. And, like, we can just have whatever number of subscribers we want to invent on YouTube.

[Britt]

It turns out that makes me sad for YouTube.

[Dave]

Yeah. I thought YouTube was less gamble than this.

[Britt]

Well, but let's game it while we can and then we'll figure out monetization later.

[Jess]

I just also want to say I'm impressed with this game. We do business by talking. The only times we talk about business are in one or two text messages a month where we ponder whether we should build one. Or at the end of our episodes, to which I sometimes respond to the text message, who wants to put 15 minutes on the calendar to talk about making money for the podcast? And everyone thumbs it up and then never responds.

[Britt]

So, you know, I'll get on a call with you to talk about that.

[Jess]

Brent proposes things we should spend money on and I propose that we should make money before we start.

[Britt]

I like to make money too, for what it's worth.

[Sam]

I also like when I hit the button to promote these things. I actually have no idea whose credit card is connected to it. It's probably mine. I literally have no idea. I think it's mine.

[Britt]

Who made the channel that's through the credit card?

[Sam]

I don't know, but whoever. Like, you're gonna have a few hundred bucks of like tens of thousands of subscriber buys.

[Britt]

It's less than all of our toke. And so it's better.

[Jess]

You know, that's an even better business model, is not knowing who's paying with that. Don't listen to us for business advice, but do listen to us for tech advice because most of the time we know what we're talking about. And we'll see you back here next week.

[Sam]

Guys, we have ended. 200 bucks. 10,000 subscribers is unbelievable.

[Jess]

Okay, I'm gonna end before we reveal some fraudulent behavior going on. Goodbye. Okay, bye, guys.

[Britt]

If you enjoyed this show, please leave us a virtual high five by rating it and reviewing it on 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 Esson lesson. And as for me, I'm Brit. See you guys next time. I also want to call out Elon Musk. Is his salary and comp package is now tied to the Mars colonization happening. So I feel like that Bones narrative narrative narrative in my favor for the 2040 bet.

[Dave]

Let's just worried about her bet with you, Sam. I'm counting on Sam's next secret project to pay for this bet.

[Sam]

Listen, Britt, here's the beautiful thing. If Elon lands on Mars and hits that comp package, I will have made so much money that I won't even notice the million dollars that I owe.

[Britt]

All I care is about the million dollars.

[Sam]

Well, that is a great emotional hedge to me.

[Britt]

I feel like you just conceded I feel like you just conceded the bet to me.