SPEAKER_00
I think we've repeatedly talked about this. AI business is a terrible business. Why is that? Every technology cycle has an infrastructure investment cycle. If you just look at the level of investment at the beginning of the web era, the way in the mobile era, the beginning of the cloud era, there's always this huge investment in infrastructure. And it's usually many years after the infrastructure investment that all of the value gets created. And we're just not even close to being there yet.
SPEAKER_00
These AI harnesses like Cloud Code, like OpenClaw, like Codex, they're very small, but what they look like to me is the future of what the applications on top of the AI infrastructure are going to look like in 10 years. [SPEAKER_02] Put it all right to the test, more or less. [SPEAKER_01] Hello, Morans. [SPEAKER_01] Welcome to More or Less. [SPEAKER_01] Jessica, Dave and Britt edition. Hey, Jess.
SPEAKER_02
[SPEAKER_03] I like that you introduced the Morans in your first line instead of the listeners.
SPEAKER_01
[SPEAKER_03] It's hello listeners, too, right? No, I'm saying hello to you. [SPEAKER_03] Okay, that's good.
SPEAKER_00
[SPEAKER_03] Let's get you back.
SPEAKER_03
[SPEAKER_00] Here we are. [SPEAKER_01] The listeners are here for me to say hello to you, for better or worse.
SPEAKER_01
That's what they're here for.
SPEAKER_03
I think what you're trying to point out is that there is no other lesson on the podcast. There's no other lesson.
SPEAKER_00
[SPEAKER_01] The other lesson is traveling.
SPEAKER_01
[SPEAKER_00] Doesn't matter. Which doesn't matter.
SPEAKER_03
[SPEAKER_01] It matters somewhat. [SPEAKER_01] But guys, it did leave me an opportunity to rejigger my entire closet to give more space for me, but also give more space from Sam.
SPEAKER_01
And viewers of this podcast on YouTube will notice my finger is taped because this went to a somewhat suburban weekend thing. All suburban weekends begin with, do I need to go to the ER? No, I'll go to Home Depot instead.
SPEAKER_00
[SPEAKER_01] So that's what I've been doing.
SPEAKER_01
Wait, wait, wait. The finger's good. [SPEAKER_03] Yeah. [SPEAKER_03] Time out. [SPEAKER_03] So first question, you removed clothes from Sam while he was gone to get more space, AKA his space in the closet? I don't think my marriage would survive that. No. What I did is he has a small portion of the closet. I decided to swap his small portion with another small portion so that all my things are together and all his things are together. You want continuity of your space.
SPEAKER_03
[SPEAKER_01] Yes. [SPEAKER_01] Okay. [SPEAKER_01] But in order to make the place I relocated him to hospitable, I decided to add more shelving to it so that I can argue he's getting more capacity.
SPEAKER_01
The old more shelving truck. Correct. [SPEAKER_03] And that's why you went to Home Depot. And that's why I went to Home Depot because I needed more shelves. [SPEAKER_00] I respect this, Jess. [SPEAKER_03] Me too. [SPEAKER_01] I'm a fan of a good DIY project, Jess. So right now we purchased the right melanite, but they would not cut it for me. So I have to go to the lumber yard sometime between now and Saturday. [SPEAKER_00] This is even more advanced. [SPEAKER_00] My respect is going up for you by the sentence right now. [SPEAKER_03] Is your AI agent teaching you what to do and walking you through every step? No. There's been no AI.
SPEAKER_03
[SPEAKER_01] I even used a real measuring tape, not my iPhone measure.
SPEAKER_01
I used measuring tape.
SPEAKER_00
I love everything about this.
SPEAKER_03
[SPEAKER_01] The finger's healing.
SPEAKER_01
And next week on the pod, we can ask Sam how he feels about his relocated closet. [SPEAKER_03] When Dave goes out of town, I actually use that opportunity to get rid of a few things that he won't remember. [SPEAKER_03] I have discarded no clothing.
SPEAKER_00
[SPEAKER_03] I have discarded no clothing. Yeah.
SPEAKER_03
[SPEAKER_00] You don't want to get rid of Sam's clothing because Sam's been wearing the same thing for 20 years and that could be catastrophic.
SPEAKER_01
[SPEAKER_03] No, no, no. [SPEAKER_03] Sam had a major style upgrade. [SPEAKER_03] I saw some photos recently. [SPEAKER_03] He's popping collars.
SPEAKER_00
[SPEAKER_03] He's all about the button ups these days.
SPEAKER_01
He's looking cheap. Did you guys see the double collar?
SPEAKER_03
[SPEAKER_01] He double collar. [SPEAKER_01] The world was confused. [SPEAKER_01] The world was confused. [SPEAKER_00] Well, anyway, he pulled a Sam.
SPEAKER_00
He pulled a Sam Altman. [SPEAKER_01] He did. [SPEAKER_01] Yes.
SPEAKER_03
[SPEAKER_01] You know the reference. [SPEAKER_00] Yeah, that's a deep guy. He pulled a Sam. I don't get why Sam Altman has double collars. [SPEAKER_00] No, it's an old WWDC, 20 years ago.
SPEAKER_01
[SPEAKER_00] I don't see the story. Yeah, that's a deep cut. [SPEAKER_00] Gotta be around. That's good. [SPEAKER_00] Let's get into this.
SPEAKER_00
Let's get into the current things. [SPEAKER_01] No, we will.
SPEAKER_01
We've got this trial, the trial of whatever raging ahead. Open AI, Elon. Juice here by the second.
SPEAKER_00
Is it really though?
SPEAKER_03
[SPEAKER_01] Yeah. [SPEAKER_00] Okay.
SPEAKER_00
All right. Well, what else are we going to talk about? I won't jump in.
SPEAKER_01
[SPEAKER_00] Tell us what else we're going to talk about. Well, we're going to jump into that.
SPEAKER_00
[SPEAKER_01] We also need to talk about the Met Gala, not from a fashion point of view, because that is not why you're tuned into this pod.
SPEAKER_01
[SPEAKER_00] Obviously.
SPEAKER_00
[SPEAKER_01] But for the anti-tech sentiment. Sometimes you guys have good fashion.
SPEAKER_01
[SPEAKER_03] I think Jess and I make good comments about fashion. [SPEAKER_00] You do, but it's not why people log into this pod. [SPEAKER_00] No, it's fine. So we really need to talk about the knives being out for all tech people, because I think it's important.
SPEAKER_00
[SPEAKER_01] I'm not trying to dismiss this.
SPEAKER_01
I think this is very important.
SPEAKER_00
[SPEAKER_01] Also, you know, guys, the enemy of my enemy is my cloud computing provider, because we have [SPEAKER_01] is not why you're tuned into this pod. Obviously. [SPEAKER_01] But for the anti-tech sentiment. Sometimes you guys have good fashion.
SPEAKER_01
[SPEAKER_03] I think Jess and I make good comments about fashion. [SPEAKER_00] You do, but it's not every. It's not why people log into this pod. [SPEAKER_00] No, it's fine.
SPEAKER_00
[SPEAKER_01] So we really need to talk about the knives being out for all tech people, because I think it's important.
SPEAKER_01
I'm not trying to dismiss this.
SPEAKER_00
[SPEAKER_01] I think this is very important.
SPEAKER_03
[SPEAKER_01] Also, guys, the enemy of my enemy is my cloud computing provider, because we have seen perhaps the strangest AI compute deal to date, which is Anthropic Colossus.
SPEAKER_00
[SPEAKER_01] So SpaceX, Elon Musk. [SPEAKER_01] What?
SPEAKER_01
Really? Have given over their first Colossus project. 100% of Colossus one. 100% of Colossus one. [SPEAKER_03] Not 40%. I know. [SPEAKER_03] Dave, did you wake up this morning? [SPEAKER_00] Oh no. [SPEAKER_00] I've been on calls all day. [SPEAKER_03] Dave's very busy right now. [SPEAKER_03] I can attest. And it is fascinating.
SPEAKER_03
[SPEAKER_01] And clearly all these other dynamics, including this trial, I think help explain how this happened.
SPEAKER_01
[SPEAKER_00] Yeah, no question.
SPEAKER_03
[SPEAKER_01] And we should talk about compute because we also published an interesting chart in the information that I think is my most commented on tweet ever in history.
SPEAKER_00
[SPEAKER_01] So compute, compute, compute. [SPEAKER_01] Let's start with Met Gala.
SPEAKER_03
[SPEAKER_01] Britt and Dave, what are your thoughts? I saw a lot of venture capitalists that I know there.
SPEAKER_01
[SPEAKER_03] And I was like, what's happening? [SPEAKER_00] And I'm just confused.
SPEAKER_00
What is the Met Gala for?
SPEAKER_01
[SPEAKER_03] Well, Lauren Sanchez Bezos and Jeff Bezos are now the key sponsors. [SPEAKER_03] I don't really care about any of this. [SPEAKER_03] Well, it's to support the Met. It's actually to support the Costume Institute of the Met. It's not even to support the core Met, guys.
SPEAKER_03
[SPEAKER_01] It is specifically to support the Costume Institute. [SPEAKER_01] And it's become both a fashion event and a socialite gathering of the year.
SPEAKER_00
[SPEAKER_01] And for years and years, there have been a parade of tech CEOs sponsoring tables, getting involved. [SPEAKER_01] TikTok sponsored, Instagram sponsored, right?
SPEAKER_03
[SPEAKER_00] Yeah. But that's because those are those. [SPEAKER_00] I always viewed it as those that were relevant were involved in the, and there was some semblance of editorial and taste involved. [SPEAKER_00] And so, seeing our dear friend Adam Mosseri doing it, Instagram's directly in the middle of the cultural conversation, right?
SPEAKER_01
[SPEAKER_00] That made sense to me. [SPEAKER_00] It seems to me like there's a lot of just random. [SPEAKER_00] I don't know. [SPEAKER_00] Did the editorial piece come off? [SPEAKER_00] Is that not the case? [SPEAKER_00] Is it really a fundraiser, right? [SPEAKER_00] Is it just money now?
SPEAKER_00
[SPEAKER_01] I think it's always been just money, but with an Anna Wintour creation layer on top, a curation layer. [SPEAKER_01] But the big news this year was one of the honorary co-hosts were the Bezos's and specifically Lauren. [SPEAKER_01] And this led to protests, it seemed, from behind the scenes, just externally, and internally a lot of consternation, people refusing to come. [SPEAKER_01] This drumbeat of tech lash, is important to talk about. [SPEAKER_01] That's not tech lash, Jess. That's. I think it's tech lash. [SPEAKER_01] I agree with Jess. [SPEAKER_01] I think it's wealth lash.
SPEAKER_01
[SPEAKER_00] Well, I think it's all of it, Dave.
SPEAKER_00
[SPEAKER_01] And this is actually what I wanted to talk to you guys about, because I think we're seeing a crescendo and a convergence in a way that is probably quite consequential for these tech companies.
SPEAKER_01
So some of it's directly tied to layoffs and AI. And so you see this commentary, comparing the estimated price of Lauren's dress to how many Amazon workers' jobs could be saved, or something obviously ridiculous at that level, but gets at an issue in the public perception of tech in this moment that I think is very significant. So you've got sort of AI and job displacement, then you just have what to me seems like more standard hate the rich kind of things that you could lob against all billionaires.
SPEAKER_01
Standard, but I would argue is on the rise, especially with things like gas prices and everything else eating the economy right now, where I feel the bifurcation of the wealthy and unwealthy is continuing to spread.
SPEAKER_00
[SPEAKER_03] I think this is high level. [SPEAKER_01] But I just can't imagine.
SPEAKER_01
I mean, God forbid we're seeing violence against Sam Altman. I mean, I just think that this is boiling over in a big way.
SPEAKER_00
[SPEAKER_01] And it's got to be something that these tech companies are taking serious looks at.
SPEAKER_01
But at the same time, you mean big tech companies? [SPEAKER_00] Yes, big tech as opposed to startups or what do you? Yeah, I mean, the scale that we're dealing with. I mean, Jeff Bezos is 100 billionaire, right? [SPEAKER_00] This is a totally different scale. [SPEAKER_00] Elon Musk could be a trillionaire, guys.
SPEAKER_03
[SPEAKER_01] Yeah, these are very different scales.
SPEAKER_01
[SPEAKER_00] These are, you know, what is Amazon and Meta and Apple and the Mag Seven, Mag Eight, whatever, the big tech companies are very different animals than all the startups. [SPEAKER_00] All the even mid-tier tech stuff are just so much smaller. [SPEAKER_00] And so, Dave, what do you think though? I think we're at a point where, especially with AI, it could all get washed up in the anti-tech segment. I mean, the other major bit of news so far this week, I'm going to give you a take.
SPEAKER_00
[SPEAKER_01] I'm going to give you one more data point.
SPEAKER_01
This is what I do.
SPEAKER_00
[SPEAKER_01] The headlines. [SPEAKER_01] We are now seeing a Trump administration seriously considering essentially taking a Biden approach to AI regulation, which was much more hands on, which was contemplating reviewing models and training ahead of time.
SPEAKER_01
Now, you guys will remember that the Biden administration stance on that was what pushed Mark Andreessen, Ben Horowitz, along with crypto into the arms of Trump.
SPEAKER_00
[SPEAKER_01] The Trump position has been, we want to put a moratorium on AI regulation.
SPEAKER_00
[SPEAKER_01] The other major bit of news so far this week, I'm gonna give you a take. I'm gonna give you one more data point. This is what I do. The headlines. We are now seeing a Trump administration seriously considering essentially taking a Biden approach to AI regulation, which was much more hands on, which was contemplating reviewing models and training ahead of time. Now, you guys will remember that the Biden administration stance on that was what pushed Mark Andreessen, Ben Horowitz, along with crypto into the arms of Trump. The Trump position has been like, we want to put a moratorium on AI regulation. And now, in the snap of the fingers, we're seeing contemplated to be clear, not made into executive order territory yet, a whole set of AI regulations that would really represent a sea change in the approach of the administration. I think there are many reasons potentially behind that, but one of which has to be they're taking the temperature on the broader sentiment around AI and what they believe the public wants from a regulatory point of view. I think the technology itself is also shaping the change in that approach. But Dave, what's your take on this? I'm mostly hearing, at least in my back channels that I'm involved in, that that is fake news, that there's not really, you look at like Sachs and these folks that are working on this stuff in the White House.
SPEAKER_00
Like, this is not how they view things. Like, they view it as this is American capitalism, like we should. [SPEAKER_01] Of course, but Sachs has a different job now. He's still informally an advisor through PCAST, but he is not the AI czar anymore. There's no AI czar, right? Like, who is the AI czar? I think Sriram would say he's the AI czar. I don't know. I think that there's a lot of games being played that are unrelated to what the White House may or may not think.
SPEAKER_00
And that, I think Anthropic's out there, funding all kinds of things, OpenAI's. I mean, there's just a lot of narrative games going on that are self-interested. And, you know, I worry about this because America, what am I seeing on the open-source front? I'm going in reverse order on your things, Jess.
SPEAKER_01
There really is no order.
SPEAKER_01
[SPEAKER_00] Jess has absolutely no, or I'm sorry, America has absolutely no take on open source AI models. Okay? Like, we don't have them. Meanwhile, China is printing them on printing presses. Like, every single founder, every company I know, everyone I am talking to about open-source is trying to fix it. And then I'm trying to figure out how to use Chinese models to make their AI pipelines more efficient by the day. Because you know what? Like, I can now run, I have this amazing little Dell box on my desk now that runs 120 billion parameter models with a Blackwell that is actually pretty good. Like it runs the open-source like really, really well.
SPEAKER_01
Wait, I have to pause. You're using a Dell box instead of an Apple box and your name is Dave Morin? [SPEAKER_00] It was a gift. It was a gift. Apple doesn't make a, look at this. This thing is the size of a Mac Mini and it runs an Nvidia Blackwell, like a GB 10. Has 128 gigs of RAM. You can run pretty awesome inference loads on it. It's pretty amazing. Oh, I wish Sam were here to see this. The times have changed, Jess. I can't wait to tell. I mean, it's the thing we've been talking about on this pod for like two years.
SPEAKER_01
[SPEAKER_00] Like I now can run pretty powerful inference with pretty smart models, like on my desk and you know, this thing costs $5,000. Like it's not that out of reach. Now the point is, I think you've got a big battleground playing out, right? You've got Chinese models and open source coming really, really strong. The American answer to that has been, we are going to let American capitalism create Anthropic and OpenAI and maybe Grok. Looks like now Grok is out the door and maybe Colossus is really the only thing that comes of it. And Colossus will run Anthropic's workloads and that's great for xAI.
SPEAKER_01
No, they're getting xAI gets Colossus 2. There's two Colossus's now. So they're not giving up on xAI. [SPEAKER_00] Anthropic gets the hand-me-downs, you know, Anthropic's like the little brother getting the hand-me-downs. Yeah, I think Colossus 1 is fine, but yes, I don't think it's like. Okay.
SPEAKER_01
[SPEAKER_00] I don't know. I don't have much more to say on that, but I want to come back around to your first point, which is around this, you know, the inequality question. My new take on this is that there are around 50 million or so. I don't know if you go on Wikipedia or you look at any numbers, there's around 50 million software engineers in the world. And yet there are 6 billion people that use the internet that use software in the world. And so up until this point in time in human history, only 50 million people knew how to create software and therefore create software wealth. And what is coming with open-source and Cloud Code and Codex. These are tools that are only used today by single digit millions people, right. There are like 4 million people using Codex, some number similar of Cloud Code, probably about the same. And yet anyone can create software now. And I know I'm a broken record on this on the pod, but I think this is there's a likelihood that the inequality that has been driven by software entrepreneurship up until this point is that of a local maxima. And that because today anyone in the world can create software at a pace that was impossible six months ago, like you can speak English to create software and therefore create software style wealth. I think this is a totally new pathway for wealth and that we might actually see a reduction in inequality because of AI over the next 20 years. And that to me is the optimistic take.
SPEAKER_01
Dave, I think there's, yeah, I follow the logic of that take. I just don't see it happening. I think that, I mean, maybe not. I also think anyone, I think anyone who says they know what the future is, that's not, the future is unclear. But what I worry about is that there are these tools. [SPEAKER_00] And that because today anyone in the world can create software at a pace that was impossible six months ago. You can speak English to create software and therefore create software-style wealth. [SPEAKER_00] I think this is a totally new pathway for wealth and that we might actually see a reduction in inequality because of AI over the next 20 years.
SPEAKER_01
[SPEAKER_00] And that to me is the optimistic take.
SPEAKER_00
Dave, I think there's, yeah, I follow the logic of that take. I just don't see it happening. [SPEAKER_01] I think that maybe not.
SPEAKER_01
I also think anyone who says they know what the future looks like, the future is unclear. But what I worry about is that there are these tools. You were just talking about open source going to democratize the tools, but not the economics that is inherently deflationary. You may, and also markets tend towards consolidation. Tech has a sort of consolidating power as well. It has a decentralizing power, but look at how many search engines there were. Look at how many search engines there are now.
SPEAKER_00
[SPEAKER_01] Look at how many social networks there were. [SPEAKER_01] Look at how many social networks there are now. [SPEAKER_01] I completely follow that the AI revolution is different and is fundamentally unleashing very powerful tools far wider than some of these other things. [SPEAKER_01] But I think technology does end up consolidating both the economics, the control and other things as well. [SPEAKER_01] And that's what we've seen in other waves. [SPEAKER_01] And the reality is I'm not focused on anything other than the software engineering piece. The idea that only a small. Look, you used to be able to make a lot of money relatively by being a journalist.
SPEAKER_00
And now everyone's a content creator.
SPEAKER_01
But journalists overall, except at the extremes, are not very well paid.
SPEAKER_00
[SPEAKER_01] I don't know that there are ever any super rich journalists. [SPEAKER_01] But just because you've democratized access to the medium, it doesn't mean you've democratized access to the economics. [SPEAKER_01] Yeah, but I think that the economics were completely unattainable unless you could speak this esoteric language, these esoteric software languages. [SPEAKER_01] And you had to also be able to speak the language of capital and raise huge amounts of capital to put enough engineers in a room to build a software system that could capture any economics.
SPEAKER_00
Right. And so I guess that's the thing here. The Doomer narrative, I refuse to let the Doomer narrative take over our podcast because I believe that the Doomer narrative comes from a specific subset of Silicon Valley. I think there's an optimistic take here too, which is that now everyone can make software. And I think there's a widening. [SPEAKER_01] It's hard for me to see anything but a widening inequality from this vantage point, knowing that that's such a slim lens because so much is unclear. But Britt, what's your take on this? [SPEAKER_01] Well, I continue to believe that what they've seen is true.
SPEAKER_00
[SPEAKER_01] And if you use the analogy that Jess, you and I went through, which is media companies to creator economy. [SPEAKER_03] Still going through it.
SPEAKER_01
[SPEAKER_03] Still here. [SPEAKER_00] Media companies to creator economy. [SPEAKER_03] Fine. [SPEAKER_00] Jess and I both have run media companies.
SPEAKER_00
[SPEAKER_03] And I still believe that there are great content people in this world, despite the fact Vox is now getting acquired by Rupert Murdoch. [SPEAKER_03] We can talk about that later. [SPEAKER_03] Not by Rupert Murdoch. [SPEAKER_03] Did that happen today? [SPEAKER_03] Guys, you're...
SPEAKER_01
I thought there was a Rupert... What was the Rupert news? Guys, James Murdoch.
SPEAKER_00
[SPEAKER_03] Kendall Roy. [SPEAKER_03] Oh, a different Murdoch. [SPEAKER_03] Kendall Roy. [SPEAKER_03] You've got the wrong Murdoch. [SPEAKER_03] We'll go back and talk about the Vox news. [SPEAKER_03] I clearly read a headline and not the story. [SPEAKER_03] And honestly, this Murdoch believes in climate change and actually is devoted to important things. [SPEAKER_03] Okay, so we're happy.
SPEAKER_01
[SPEAKER_03] Great. We're very happy. We're so excited about this.
SPEAKER_00
[SPEAKER_01] I hope the deal gets done.
SPEAKER_01
There's still a high level of quality that can get paid and have a great job in this industry.
SPEAKER_00
[SPEAKER_01] And I think the same will be true for software engineers. [SPEAKER_03] At the same time, there's now a new trillion dollar economy for everyone else that wasn't a content creator before and didn't know how to write an article or make a video and now can and does and has profited off of that. [SPEAKER_03] And so I think to your point, I think you're both right. [SPEAKER_03] Yes, but to what level of profiting, right? [SPEAKER_01] I'm not saying they're going to have venture-backed companies, but they're going to make maybe hundreds, thousands, tens of thousands, millions of dollars, some of them on a spectrum, right?
SPEAKER_00
[SPEAKER_03] And so I think that's net positive and optimistic for people across the world to think about who might never have known how to write software or to turn an idea into a company or an app. [SPEAKER_03] They've been extracted from by people who could create software. [SPEAKER_03] I think this is really important to just point out. [SPEAKER_03] Every other week we have this debate. I want to have the same debate, but in the context of public sentiment towards all of this right now, right? [SPEAKER_01] So take away what tech people are saying, Dave, and the Doomerism and all of that with its multi-agenda. [SPEAKER_01] But it plays into it. [SPEAKER_01] Right?
SPEAKER_00
[SPEAKER_01] The Doomer agenda has become popular culture. [SPEAKER_01] Okay. [SPEAKER_01] But what do we do about that?
SPEAKER_01
[SPEAKER_00] Or what do you guys do about that? Because that's a reality today. Once people start realizing, once the everyday person in Texas that I grew up with is like, Whoa, I made $1,000 from this random AI thing I created, which I don't think is going to happen today, but maybe a year from now. [SPEAKER_03] And that's going to spread that sentiment and people are going to be like, Whoa. [SPEAKER_03] And by the way, Dave and I did an hour-long meeting with our head of school this week. [SPEAKER_03] Every time we do conversations with normies, with the head of school or whatever, they're like, What?
SPEAKER_01
[SPEAKER_03] I can do all this stuff, but all the parents hated it and everyone's disputing how we should use AI and how we should do all these things. Whoa, I made $1,000 from this random AI thing I created, which I don't think is going to happen today, but maybe a year from now.
SPEAKER_01
[SPEAKER_03] And that's going to spread that sentiment and people are going to be like, Whoa. And by the way, Dave and I did an hour long meeting with our head of school this week. It was just like, every time we do conversations with normies, the head of school or whatever, they're like, what? I can do all this stuff, but all the parents hated it and everyone's disputing how we should use AI and how we should do all these things. But also I could run my school more efficiently. So you think the path out is education?
SPEAKER_01
[SPEAKER_03] Yes. I'm just saying people have not even started using the most powerful parts of AI yet. And once they have, and once they realize how much more efficient they can make their business, or they can start a new business, or they can make their life, I think the sentiment will naturally change. [SPEAKER_03] Yeah, not just education. We actually have to have better products, right? [SPEAKER_03] I keep telling everyone it's 1984 and the computer was just invented. [SPEAKER_00] It's 1997 and websites are a thing. [SPEAKER_00] It's 2007 and mobile apps are now a thing, right?
SPEAKER_01
[SPEAKER_00] And at the beginning of all of these things, nobody had the slightest clue what all of those new objects were going to be. It turns out that we're going to have billions of computers made. We're going to have billions of websites. We're going to have billions of apps. And we're right there right now on AI agents. And nobody has the slightest clue, even though everybody in town keeps talking about how the future is already here. Everybody's already saying AGI is already happening. It's like, dude, no one is using this stuff yet. Single digit millions of people are using this stuff. And the only thing between here and there is more people using this, which is going to be better user experiences, better design, better products. We haven't cracked it yet. ChatGPT itself still isn't beyond 800 million people and it's declining in growth right now. And so we have to make better products in order to get this vision to the rest of the world. ChatGPT is declining in growth. It's actually losing users. On MAU.
SPEAKER_00
[SPEAKER_01] Or it's growing less slowly. MAU looks like it's growing slower on all the charts I've looked at. But growing slower isn't losing users. It's just growing less slowly. It's still growing.
SPEAKER_01
Growing slower. I don't know. Declining growth. I don't think it's losing users. [SPEAKER_03] I think it's growing less slowly. Look, I think this will be a very interesting conversation to come back to in 24 months. Because I think that again, everything you guys are saying makes sense and I can chart the logic in it. I also think people aren't going to buy it. And remember layoffs are a big reason. You mean the public sentiment thing? [SPEAKER_00] Yeah, the public sentiment. And I think layoffs, look, Coinbase just had layoffs. You know, Meta has had recent layoffs.
SPEAKER_01
Meta had recent layoffs and came out and said, we can't confirm that there won't be more. And we don't actually know the optimal size of this company. That is corporate messaging that is telling people to batten down the hatches for more layoffs, right? By the time you're saying that publicly.
SPEAKER_00
[SPEAKER_01] Yeah, but this is all post pandemic. Jess, you and I are in similar circles. Everybody is doing this because they over hired in the pandemic. AI is a convenient excuse. Now they're doing it because they're spending $200 billion on compute as well. I think they're doing it because they also can do things more efficiently for real. They can do things way more efficiently. [SPEAKER_01] I think it's both, right? Dave, think of the things you could do with your agents, right? It doesn't mean you're going to stop hiring people, but it does mean that there are certain people.
SPEAKER_01
I have to hire more people. That's what I don't understand about this whole narrative is everyone I know working in this stuff is more busy than they've ever been. That's part of what drives my whole passion around this is that everyone I know at the heart of the AI conversation and at the heart of AI innovation is working harder and more than they've ever worked in their entire life to the point where they're not sleeping well.
SPEAKER_00
Their health isn't as good.
SPEAKER_01
[SPEAKER_00] It's like this stuff is not making us more efficient. [SPEAKER_00] Well, I also think there's a dopamine hit this thing gives you and they are obsessed with it. [SPEAKER_00] Yeah, look, the token maxing population is not the population we're talking about right now, right?
SPEAKER_03
And some version of that's coming. Everybody else is also going to experience this, right? [SPEAKER_01] Tokens are going to become part of your job. They're going to experience something, but I think it's really hard to know exactly what. To use a cheesy word, what the diffusion of all of this is going to look like and how. I think it's so early that we cannot. It's not easy to extrapolate how this is. [SPEAKER_00] And this is why you see wildly different projections. And I'm open to that. Maybe I'm totally disconnected and maybe I'm.
SPEAKER_00
[SPEAKER_01] Yeah, maybe.
SPEAKER_03
[SPEAKER_00] Maybe I have no idea. Maybe I'm absolutely. To the listener, hit me on Twitter if we sound completely disconnected from what's actually going on.
SPEAKER_00
Don't worry. People will tell us in the YouTube comments if we're totally disconnected and how we're totally disconnected. But I think you know, I've talked about starting to get some policy proposals around how to handle this out there. I think when OpenAI bought TPPN, that, whatever acquisition you would fill in the blank from two months ago, it was like changing the public perception of AI was one of the stated reasons.
SPEAKER_03
[SPEAKER_01] Yeah. Tell better stories. These guys have only used the Doomer kind. We've talked about it at length on this pod. It's a Christian theological Doomer. We are creating a God. [SPEAKER_00] Don't worry. People will tell us in the YouTube comments if we're totally disconnected. But I think you know, I've talked about starting to get some policy proposals around how to handle this out there. I think when Open AI bought TPPN, whatever that acquisition was two months ago, it was like to change the public perception of AI was one of the stated reasons.
SPEAKER_03
[SPEAKER_01] Yeah. Tell better stories. These guys have only used the Doomer kind of we've talked about it at length on this pod.
SPEAKER_01
It's a Christian theological Doomer. We are creating a God. You should fear it.
SPEAKER_03
[SPEAKER_00] They've used this one version of a message. It's actually ancient spiritual technology. They use literally Christian theology to sell this thing to the world and scare everyone. And now they're stuck. They're like, ah, AI is less popular than ice. What do I do? I have to tell a much better story. Right? I have to give people the optimistic take. And look, if people are going to feel like it's coming for their jobs or their kids' jobs, it's going to be very hard to come up with a narrative that's going to create that.
SPEAKER_01
Now, Dave, you're right. In the past, the product itself has closed the gap. Right. And Google ate all the world's information until it was super convenient and then it was Google and you could never live without it. Right.
SPEAKER_03
Or name your privacy concern that went away with some new convenience. But it's crescendoing. I feel like I'm the voice of Sam here on this or something like that. It's we're the product. No, the policy thing. You just mentioned policy, right? Policies are so far behind. We can't count on policy to help with any of this. If we're staying up in this, what is the popular conversation amongst all of the citizens of the United States and the world?
SPEAKER_01
[SPEAKER_00] Policy is not going to help.
SPEAKER_03
Right. The only thing that's going to help is better products. Right. That do better at what we're doing. And then I guess that's the breakdown in ideologies. Right. Some people believe you need to slow it down and stop it and be Bernie or whatever in policy. Some people are technology maximalists and they think if we can make better technology, then maybe that will solve the problem.
SPEAKER_00
And that's the great debate. But don't you think speed is the biggest part of this? Because during the search era and the mobile era, those were both radical changes, even the PC era for people, for their jobs, for how they live, for how they worked.
SPEAKER_01
But they happened over so many years that I think organically people adapted and companies adapted. And right now we're in this crunch period where everyone's racing at the speed of light to get to the next model. And so everyone needs to keep up and every company needs to keep up and jobs are losing faster and everyone needs to pivot faster. And there's this amount of stress and anxiety. [SPEAKER_03] Who is everyone? Companies across the country, schools, literally. No, I mean, this isn't happening across the country.
SPEAKER_00
It is.
SPEAKER_01
[SPEAKER_00] The literal negative sentiment is because people are afraid of what AI means for them, for their kids, for their jobs, for all of this stuff and how it's going to take over their lives. [SPEAKER_03] And I wonder if it was slowed down and all of this was happening over a five or ten year time horizon instead of a one year time horizon. I don't think you can slow this stuff down.
SPEAKER_03
But I think that's what the government's trying to do. No, they're not, Britt. Nobody's trying to do anything. Guys, DeepSeek is raising a gargantuan amount of money. Well, they're from China. Our government is like, maybe we should put the brakes on some of this stuff. No, I know. [SPEAKER_00] But I just think that the human sentiment of change is not meant to move this fast. Even high growth tech startups are freaking out. Oh, we have to change. We have to do layoffs. Humans don't like change in general. I think it's possible.
SPEAKER_00
I'm going to try this out.
SPEAKER_03
I don't know if I believe it. I think it's possible. It's not actually moving this fast.
SPEAKER_00
That's what I'm saying. [SPEAKER_01] It's not. The narrative, including the role of the press broadly and the role of commentators and everyone is, you're right, Dave, they're piling into the Doomer narrative or they're piling into the diffusion is here tomorrow narrative. Right. And what's happening, but it's a hard line because you do need to evangelize this stuff to get experimentation. Of course. Right. [SPEAKER_01] And so, but then it's like, you've been AI pilled. No, I haven't been AI pilled, but I did like, you got to learn about this. Yeah, totally. But look, the numbers. Just look at the numbers.
SPEAKER_00
That's what I'm trying to do. In every conversation I'm in, I'm just grounding people in the numbers. There are only 4 million people using Codex. That's no one, literally no one is using AI engineering tools. [SPEAKER_01] No one in the grand context of the world. It's got to be the same number, right? If you just look at the numbers, there's around 10 million software engineers that build iOS and Android apps. So the population of developers in the world is somewhere in the ballpark of 10 million. And Codex's last number was 4 million. My guess is that Claude's is similar, single digit million. And so that's awesome. I love this stuff.
SPEAKER_01
I'm using it every day. It's so much fun.
It's a new paintbrush that you can paint amazing new paintings with. Right. And it's so fun and it's great new technology.
SPEAKER_00
But the numbers are so tiny, so tiny.
SPEAKER_01
And I get it. [SPEAKER_00] So the population of developers in the world is somewhere in the ballpark of 10 million. [SPEAKER_00] And Codex's last number was 4 million. [SPEAKER_00] My guess is that Claude's is the same, it's single digit million.
SPEAKER_03
[SPEAKER_00] And so that's awesome. [SPEAKER_00] I love this stuff.
SPEAKER_01
[SPEAKER_00] I'm using it every day. [SPEAKER_00] It's so much fun. [SPEAKER_00] It's a new paintbrush that you can paint amazing new paintings with. [SPEAKER_00] Right. [SPEAKER_00] And it's so fun and it's a great new technology. [SPEAKER_00] But the numbers are so tiny, so tiny.
And I get it. [SPEAKER_00] The revenue numbers are enormous, but the numbers are still so small in terms of how far this is diffused into society. It's so small. [SPEAKER_00] I want to transition this to the compute conversation, just some new developments there, because this is the other side of the coin of this. [SPEAKER_00] But Dave, I also think it's interesting to point out that you're making this point that it's so small that so few people are using it that we don't know. [SPEAKER_00] You're also saying that you and everyone you know have to hire more people or going crazy. [SPEAKER_00] I've never heard comments like that have people extrapolating to.
SPEAKER_01
[SPEAKER_00] Well, that's what I think. [SPEAKER_00] I think we've well articulated it. Perhaps we tend to articulate it on the pod that there's no good business in AI. I think people are sick of us talking about it this way on the pod that I don't think we all think there's no good business in there. Well, I mean, I think we've repeatedly talked about this, that the AI business is a terrible business.
SPEAKER_00
[SPEAKER_01] Why is that? Every technology cycle has an infrastructure investment cycle. If you just look at the level of investment at the beginning of the web era, at the beginning of the mobile era, at the beginning of the cloud era, there's always this huge investment in infrastructure. Right. And in this case, it's been in compute, which is a great thing to switch to next. But it wasn't until the application era started long.
SPEAKER_01
[SPEAKER_00] And it's usually many years after the infrastructure investment that all of the value gets created. [SPEAKER_00] And we're just not even close to being there yet. [SPEAKER_00] Right. That's what I'm trying to point out is that these AI harnesses like Cloud Code, like Open Claw, like Codex. [SPEAKER_00] They're very small, but what they look like to me is the future of what the applications on top of the AI infrastructure are going to look like in 10 years. [SPEAKER_00] And nobody's started building them yet. [SPEAKER_00] It's all just pretty new. [SPEAKER_00] And we're figuring this out.
SPEAKER_01
[SPEAKER_00] And there's a huge amount of business to be made around how do you build stuff to use the AI infrastructure better and make it more usable to people?
SPEAKER_00
Because it's still very hard to use AI inference in your job. There has to be a tool for you to use the inference to do the job that you're doing. Right. And that requires user interface and design and well thought out tools. And a lot of that is all just really new. Most people just use AI to get information today.
SPEAKER_03
[SPEAKER_00] They don't use the inference to do their job.
SPEAKER_01
[SPEAKER_00] And that's going to be the next 20 years. [SPEAKER_00] Right. And that takes a long time because a tool needs to be made for their job or their new job.
SPEAKER_00
And then they need to learn how to do it. And humans don't metabolize things very fast.
SPEAKER_01
[SPEAKER_00] And so I buy your take that it's much slower than people think. [SPEAKER_00] And the narrative is way out in front. [SPEAKER_00] Which is also I so we don't have to belabor compute, which I know is not everyone's favorite topic, but two important things happen. [SPEAKER_00] One, let me just share this chart.
SPEAKER_00
So The Information had some reporting about Anthropic, which has been behind in compute capacity. Yeah, it's clear they made a massive strategic mistake. They made a massive strategic mistake. They seem to have remedied it pretty quickly. I think they've remedied it really quickly.
SPEAKER_01
[SPEAKER_00] And at great expense. [SPEAKER_00] I don't know about that. [SPEAKER_00] What are the actual numbers? Yeah, I mean, they Google and now Colossus, right?
SPEAKER_00
[SPEAKER_01] Yeah, but Brit, this is probably costing them a fortune. Like this is a massive mistake. [SPEAKER_01] I mean, 200 billion. [SPEAKER_01] Whoever is the CFO or did the planning is I don't know. [SPEAKER_03] It's just not. [SPEAKER_03] Krishna Rao, smart guy. They made a mistake. Look. Yeah, but this was a serious mistake. [SPEAKER_03] It's a serious mistake. It will cost them. I think Dario said at their conference today that their revenues up 80x where they thought it would be. [SPEAKER_03] They're doing just fine. Okay.
SPEAKER_01
[SPEAKER_00] So, but here is an interesting thing. So The Information broke the news about the size of Anthropic's latest Google Cloud commitment at 200 billion. And then we charted the backlogs in deals for Microsoft, Oracle, Google, and Amazon from their earnings. What percentage are OpenAI's commitment and Anthropic's commitment? [SPEAKER_00] It's together. They're about half. Okay. On one hand, this is not surprising, obviously, who else are the really big customers of compute at this moment.
SPEAKER_00
[SPEAKER_01] But it's also really striking how much growth that's happening in these multi-trillion dollar hyperscalers is coming from the spending commitments of these two companies. [SPEAKER_01] From the private markets, basically. [SPEAKER_01] From all venture capital. [SPEAKER_01] Right? [SPEAKER_01] Yeah, but am I reading this right? [SPEAKER_01] That OpenAI was just so much smarter. [SPEAKER_01] The scale of what they were doing was so far beyond Anthropic. [SPEAKER_01] And why hasn't OpenAI invested in Google? [SPEAKER_01] Well, so my other take is when they, these guys will, they just view themselves as too competitive.
SPEAKER_00
[SPEAKER_01] But when they put that aside and OpenAI said. [SPEAKER_01] Yeah, but how's Anthropic not competitive to Google? [SPEAKER_01] Well, OpenAI thinks it's competing with everyone. [SPEAKER_01] But the other thing, by the way, Dave, I'll circle back to your question is this chart just shows the money is invested, right? [SPEAKER_01] So this is those who believe that this is a circular economy.
SPEAKER_03
[SPEAKER_00] So Microsoft is investing in Anthropic's. [SPEAKER_00] You have to believe it. [SPEAKER_00] It just is. It's going right back up there. [SPEAKER_01] We should have put this in one chart team. [SPEAKER_01] Yeah.
SPEAKER_00
[SPEAKER_03] We have it in other graphics right back up here.
SPEAKER_03
[SPEAKER_01] But when they put that aside and OpenAI said. [SPEAKER_01] Yeah, but how's Anthropic not competitive to Google? [SPEAKER_01] Well, OpenAI thinks it's competing with everyone. [SPEAKER_01] But the other thing, by the way, Dave, I'll circle back to your question is this chart just shows the money invested, right? [SPEAKER_01] So this is those who believe that this is a circular economy. [SPEAKER_00] So Microsoft is investing in Anthropic. [SPEAKER_00] You have to believe it. [SPEAKER_00] It just is. It's going right back up there. [SPEAKER_01] We should have put this in one chart team.
SPEAKER_00
[SPEAKER_01] Yeah. [SPEAKER_03] We have it other graphics right back up here.
SPEAKER_01
So Dave, I think you're, I think, how do I stop sharing?
SPEAKER_03
[SPEAKER_01] Okay. [SPEAKER_01] Yeah. [SPEAKER_00] Anthropic screwed up for this compute, but also, and I think they will fix it. [SPEAKER_00] It's at a great cost. [SPEAKER_00] But after striking Colossus, they instantly rolled back some of the limits. [SPEAKER_01] They instantly upped the capacity.
SPEAKER_00
[SPEAKER_01] I mean, I actually think they resolved it pretty quickly.
SPEAKER_01
And I don't think you can fault companies for having even more exponential growth than the exponential growth that they thought. I mean, they were obviously too cautious in compute. And that was obviously a mistake, but that's going to be nowhere near existential for them. And I think it's hilarious that Dario, Elon Musk, are the inverse of each other as humans. Like, they really are not similar on any metric, except their hatred of Sam Altman right now.
SPEAKER_00
[SPEAKER_01] They're very different. [SPEAKER_01] Like, Dario, he's starting to be more outspoken, especially around his Doomer narratives.
SPEAKER_01
But he has been—it's a good point, Dave, how to describe it. But he articulated very different principles around safety, articulated very different principles around capital than Musk. I would say, while he says some very bold narratives on certain topics, he's an understated human being in many ways, minus pushing those narratives. He's built a culture at Anthropic where absolutely no one leaves.
SPEAKER_00
[SPEAKER_01] XAI is a revolving door of people. [SPEAKER_01] They're just very different type of leaders. [SPEAKER_01] And they've had very different approaches to regulation.
SPEAKER_01
Elon has been back off.
SPEAKER_00
[SPEAKER_01] Dario's been full-court press. [SPEAKER_01] So, I would just say there have not been many common threads on how they operate their companies, what they've articulated as important, that I've seen, having followed this space pretty clearly. [SPEAKER_01] But now we have this mega deal where SpaceX is selling or renting computer to Anthropic. [SPEAKER_01] Elon is tweeting up a storm about all the time he spent with the Anthropic team and what great humans they are and how aligned they are. The elephant in the room is just OpenAI. [SPEAKER_01] And Musk wanting to stick it to them every way possible.
SPEAKER_00
[SPEAKER_01] I mean, as this trial has gone on, and the SpaceX IPO, we're showing any customer demand for your compute. [SPEAKER_01] And by the way, I don't think it's a coincidence that Anthropic said it would be potentially interested in orbital data centers, which of course is the driving narrative of the SpaceX IPO and which no one in the industry at all believes is actually a near-term thing. [SPEAKER_01] So, it seems like they both got some extra juice out of this agreement, but I mean, it makes a lot of sense.
SPEAKER_00
[SPEAKER_01] I think it's a very smart move from Elon's perspective and probably a prescient one on his part to have gone full steam ahead and building all these data centers, knowing that this opportunity would present itself at some point. [SPEAKER_01] But it is wild. They're very odd bedfellows is my take. I don't know. What do you guys think? [SPEAKER_01] I can't take Elon seriously anymore. And I feel he's taking the Trump pill of being best friends with someone one day and not the next. [SPEAKER_01] And so I don't actually trust any of these deals, that they're all going to happen, go through, that he's on one side or the other.
SPEAKER_00
[SPEAKER_01] I feel no one—these are all characters as we've talked about before masquerading and one month is very different from the next. [SPEAKER_01] And I think you're even seeing the same thing. We've been so hot on Anthropic for the last few months, but also ChatGPT's new model is really awesome. [SPEAKER_01] And now there's a lot of people starting to use that again. [SPEAKER_01] Did you guys get invited to this party? I didn't get invited to the party. There was a 5-5 party or something? [SPEAKER_01] On 5-5. Yeah. Yeah. [SPEAKER_01] Companies now parties for model releases. Oh, please God.
SPEAKER_00
[SPEAKER_01] This seems very something. Okay. Sorry, Brent. It's swinging. So you like the new ChatGPT model? [SPEAKER_01] Yes. And I still use all of them. And I'm still like, I just think we're so early in this. [SPEAKER_01] And what you know, Sam would say they're all just going to go to the same place and be commoditized. [SPEAKER_01] But I actually think we're going to be flip flopping back and forth for the next many months.
SPEAKER_01
And we'll see who wins out, but everyone's just getting their compute infrastructure. [SPEAKER_03] Everything in place and trying to build as fast as they can. [SPEAKER_03] Meanwhile, Google has 3 billion users with registered accounts. [SPEAKER_03] So I'm still long Google here.
SPEAKER_00
[SPEAKER_03] This is a cycle where Google goes slightly quiet and then it's IO, and we'll see. [SPEAKER_03] Yeah. Two weeks from now, Google is going to come out with a bang. [SPEAKER_03] And we just don't know. Have you guys been following the trial? [SPEAKER_03] No. Can you fill us in? [SPEAKER_01] Well, so it was really recent days have been Brockman's turn on the stand. [SPEAKER_01] And I think as we've talked about too, Greg diary, the entire founding of OpenAI, his heart's wants, desires, financial goals, everything. [SPEAKER_01] And that is all in evidence. So he's been asked a lot of questions by that.
SPEAKER_00
[SPEAKER_01] I mean, look, Musk's lawyers are pit bulls and I think they're actually very good at what they do. [SPEAKER_01] I was catching up with my lawyer yesterday and I was just like, what do you think's going to happen in the trial? [SPEAKER_01] And he was like, Musk has great lawyers, you know? So I don't know. [SPEAKER_01] But basically it was a lot of going back to the earliest days of OpenAI with Elon, with Greg, with Sam. [SPEAKER_03] It was revealed that Altman, because it was a nonprofit back then, was paid Greg $10 million out of his own family office and kept it from Elon.
SPEAKER_00
[SPEAKER_03] And so all these alliances were being formed beneath the surface and it just seemed like a complete tinderbox ready to explode, which it did. [SPEAKER_01] And you know, it also came out that Brockman's stake is worth about $30 billion, which is a point that he was just grilled on and asked, well, why haven't you donated it? [SPEAKER_01] And he was like, Musk has great lawyers. So I don't know. [SPEAKER_01] But it was a lot of going back to the earliest days of OpenAI with Elon, with Greg, with Sam. [SPEAKER_03] It was revealed that Altman, because it was a nonprofit back then, paid Greg $10 million out of his own family office and kept it from Elon.
SPEAKER_00
[SPEAKER_03] And so all these alliances were being formed beneath the surface and it just seemed like a complete tinderbox ready to explode, which it did. [SPEAKER_01] And it also came out that Brockman's stake is worth about $30 billion, which is a point that he was just grilled on and asked, well, why haven't you donated it? [SPEAKER_01] And all these questions made me realize, I do think Altman, who's very shrewd, was really prescient in not having equity in OpenAI. [SPEAKER_03] And it can get paid in a lot of different ways. [SPEAKER_03] But I was actually the first to report that he didn't way back when in all these funding rounds.
SPEAKER_00
[SPEAKER_03] And it was peculiar to me, but the sense was that he really felt that it would be better for everyone, including not coming back to bite him for having personally made so much money on a particular technology. [SPEAKER_03] So Mira Murati has been testifying or her deposition is being played. [SPEAKER_03] And so questions about Sam's leadership are coming out and a little bit more TMZ than substance in some regards. [SPEAKER_03] Yeah, I guess that's been my read is that everybody thought this was going to be, I don't know which side thought the other was going to be the pariah in the eyes of the public, but it's all a thud in my view.
SPEAKER_00
[SPEAKER_03] It's wealthy people bantering. And it's unfortunate. It doesn't seem to serve the public interest and maybe I'm misreading it, but it looks to me like a dud trial. [SPEAKER_03] But I guess we'll see. [SPEAKER_03] Well, I don't know. [SPEAKER_03] I mean, I think ultimately Musk has to show that OpenAI violated a contract or committed fraud. [SPEAKER_01] And I do think these lawyers are smart and all of these things do potentially weigh towards that, but it's a little circuitous. [SPEAKER_01] And the jury obviously has a lot of power in this case, too. [SPEAKER_03] And so you're just trying to convince jurors.
SPEAKER_00
[SPEAKER_01] Polymarket gives Elon a 41% chance of winning the case.
SPEAKER_01
Has it changed at all? I haven't looked at the... It's been pretty stable.
SPEAKER_00
[SPEAKER_01] It started at 56%.
SPEAKER_01
So he's losing? Yeah, but you have to invert these things, right?
SPEAKER_03
[SPEAKER_01] Because the betting pool is on the flip side, right? [SPEAKER_01] So yes.
SPEAKER_00
[SPEAKER_01] The title is, will Elon Musk win his case against Sam Altman? [SPEAKER_01] And it dropped. [SPEAKER_01] Yeah.
SPEAKER_03
[SPEAKER_01] Polymarket's betting against him winning.
SPEAKER_00
[SPEAKER_01] It started out at 55 and we are at 41 right now. [SPEAKER_01] It's been all down since it started really.
SPEAKER_03
[SPEAKER_01] But there's been some bobbling in the last week for sure.
SPEAKER_00
[SPEAKER_01] OpenAI also revealed in a filing that Musk had approached Brockman about potential settlement two days before. [SPEAKER_01] That doesn't really help his case.
SPEAKER_01
Although I think that was more body language. Wait, sorry, sorry. What happened?
SPEAKER_00
So OpenAI put out a filing saying that a couple of days before the trial at the Breakthrough Prizes, Musk had approached Brockman about potentially figuring out a solution to this or something.
SPEAKER_01
[SPEAKER_00] So OpenAI put it in... I think the reason they issued a filing is they wanted to be able to ask Brockman about it. So they had to submit that as testimony of some nature. But it also made it, apparently, according to this OpenAI filing, when Brockman rebuffed him, Musk said in a couple of weeks, you two are going to be the most hated people on the planet. And that hasn't happened. So it shows how disconnected they are. [SPEAKER_03] All of this is from reality, right? [SPEAKER_00] I don't know that anyone cares about this, even people in town. It's interesting TMZ stuff. [SPEAKER_03] Well, I know.
SPEAKER_01
[SPEAKER_03] I mean, not everyone has to care an equal amount about everything, but there are things at stake here. [SPEAKER_03] I mean, OpenAI is important. [SPEAKER_03] OpenAI's IPO is something that's going to be widely watched and a lot of people are going to have a clear stake. [SPEAKER_03] And so I don't think we have to overindex towards it for drama's sake, although please someone make the movie. [SPEAKER_00] It is happening. [SPEAKER_00] It is news. Guys, side news on, you can actually bet in $6 million of betting on Polymarket on the number of times Elon Musk will tweet in a given week.
SPEAKER_00
[SPEAKER_01] And for this week between May 1st and 8th, there's $6 million that's gone into it. The current 48% is 160 to 179, but the numbers go all the way up to 500 plus. This is a hilarious Polymarket.
SPEAKER_03
On this point, this is a fun pop culture thing.
SPEAKER_01
One of the things I've been hearing a lot from smart people running interesting things with Claude is people are doing Polymarket trading operations using several Claude or several agents. And the best trading market is Twitter mentions, whether or not something's going to be mentioned on Twitter.
SPEAKER_03
[SPEAKER_01] So a trade like that.
SPEAKER_01
[SPEAKER_00] It can't be very easy to manipulate. [SPEAKER_00] Didn't Brian Armstrong like this? [SPEAKER_00] I think that's why it's so good.
SPEAKER_00
[SPEAKER_01] It has cameo-esque vibes. [SPEAKER_01] You're just getting people to say things. [SPEAKER_01] I don't know.
SPEAKER_01
Yeah. Cameo. [SPEAKER_00] We had a Polymarket story today about the slow start for their US business. [SPEAKER_00] So worth checking out for people interested in such topics. [SPEAKER_00] What else? [SPEAKER_00] And Steve, you want to give us the Claude update? [SPEAKER_00] You're in meetings all day. [SPEAKER_00] You're not even looking at your information news feeds. You must be swamped. What's wrong with you? His Claude is delivering the news nightly. That's fine.
SPEAKER_01
Are you getting MCP'd in? I don't know. Yeah. Cameo. [SPEAKER_00] We had a Polymarket story today and the information about the slow start for their US business. [SPEAKER_00] So worth checking out for people interested in such topics. [SPEAKER_00] What else? [SPEAKER_00] And Steve, you want to give us the OpenClause update? [SPEAKER_00] You're in meetings all day. [SPEAKER_00] You're not even looking at your information news feeds. You must be swamped.
SPEAKER_01
What's wrong with you? His claw is delivering the news nightly at night. That's fine. Are you getting MCP'd in? You want an MCP deal? I got you. [SPEAKER_03] We got you.
SPEAKER_01
[SPEAKER_03] We're way more into CLIs at OpenClause. [SPEAKER_03] Yeah. [SPEAKER_03] Can you explain that for me? [SPEAKER_00] What's a CLI? [SPEAKER_00] Command line interface. [SPEAKER_00] Oh. [SPEAKER_00] Yeah. [SPEAKER_00] Yeah. But you can still MCP. Those are two different things.
SPEAKER_03
[SPEAKER_00] Similar. [SPEAKER_01] The main OpenClause update is that we have had a lot of bugs and stability to work on for the last few weeks. [SPEAKER_01] And there's a lot that has happened and a lot of people using this thing at very large scale. [SPEAKER_01] And so we've been very focused on stabilizing the core, reducing what OpenClaw is to only what matters and then putting everything that is third party things that a lot of people use into a plug in architecture so that the central core is safe, secure, fast, and as bug free as possible. [SPEAKER_00] It's definitely taken us longer than we expected.
SPEAKER_03
And there was a rough week last week, but we're at a good place now. [SPEAKER_01] Things are speeding up.
SPEAKER_01
[SPEAKER_02] People are happy, maybe turning the corner. It's been a lot of work, though, doing a release almost every two days right now, sometimes multiple times a day. Dave, is OpenAI helping with this, too? Or is this really just the open source network as it's been growing separate from them? I know they have some involvement. [SPEAKER_03] It's the open source network. [SPEAKER_03] They actually have a couple of people contributing from the OpenAI engineering team who have been contributing as open source maintainers.
SPEAKER_03
Obviously, Peter is the largest contributor and he technically works at OpenAI. And then we're in conversations about the other big update, which is the foundation.
SPEAKER_01
[SPEAKER_03] I've already mentioned it publicly several times, but the 501c3 is set up. [SPEAKER_03] We actually have capital coming in the door now.
SPEAKER_03
We've actually begun onboarding a full time team to work on the open source core at the foundation, which is great. [SPEAKER_01] That's a really important piece of doing a good job on this project, given how large the scale is at this point. [SPEAKER_00] So we've been working through that. [SPEAKER_00] And then the conversation with OpenAI, we're still figuring out what level they're going to support us. [SPEAKER_01] We burn an enormous amount of tokens on this project. [SPEAKER_01] We haven't disclosed publicly how big it is, but the number is very large. [SPEAKER_00] And so the help of token providers of various kinds is very important to us.
SPEAKER_03
[SPEAKER_01] We have ongoing conversations with many different partners right now. [SPEAKER_00] And we're going to get a blog post out in the next couple of weeks talking through a lot of that. [SPEAKER_01] So I'll save that for then. [SPEAKER_01] Did you see the Meta news and the information a little bit more about their personal agent plans?
SPEAKER_01
I think this was out there to some degree, but yeah, at OpenClause we've been aware for some time that they have a fork of OpenClause and we expect they'll do something pretty big. [SPEAKER_00] So that's cool. It's good to hear they're doing that.
SPEAKER_03
[SPEAKER_00] And yeah, it's great.
SPEAKER_01
[SPEAKER_00] It says end of June, I think, right? [SPEAKER_00] We see the same thing. And in China, you've got huge deployments on ByteDance, Baidu, Alibaba. They all have forks. [SPEAKER_00] They're all deploying at large scale. So there's a lot of this going on all over the world. [SPEAKER_00] My big AI news of the week is I connected my granola to my cloud code. [SPEAKER_00] You're late to that. [SPEAKER_00] Look at me. [SPEAKER_00] Yes, I did that six months ago. That's great. You know, what's cool about, oh, you can do some damage with that. Tell me, give me two examples of things now that I've made this connection I can do.
SPEAKER_01
[SPEAKER_00] Okay, one thing that's obvious, there's a partnerships business that you run at the information and probably hot leads. You have CRM of some sort. [SPEAKER_00] I don't know what software you're using for your CRM. [SPEAKER_00] But literally, today, within 30 minutes, I imported all my granola notes for the last year and a half, two years, and all of them to populate a new CRM for me on the fundraising side of things. [SPEAKER_00] And I can now rank them. You had cloud, using anecdote as a technical term or you're saying you added notes, we've had a sales meeting, you're pitching me on, you know. [SPEAKER_00] Okay, you had a sales meeting. [SPEAKER_00] Okay.
SPEAKER_01
[SPEAKER_00] No, one of your sellers talked to Sony a year ago, and they were like, out of the side of their mouth, they were like, Yeah, we really don't have budget this year. But probably mid 26 is when we're going to look at these opportunities again. You forgot about it because they weren't on your hot leads list. [SPEAKER_00] Guess what pops up in your CRM now with a flame icon? Sony because it's mid 2026. And you know you need to prioritize that and you have the exact anecdotes from those. [SPEAKER_00] I think we pay Salesforce a ton of money to do that. [SPEAKER_00] Well, you don't need to anymore. [SPEAKER_00] Bye Salesforce.
SPEAKER_01
[SPEAKER_00] No, all of this is happening. Now you can custom create it, you can feed in everyone's granolas. Every time you have a new meeting with Sony, now you're enriching the data over and over again. It's good.
SPEAKER_00
Yeah, I don't know. I'm interested in what my team thinks about sharing granolas. They don't have to share all granolas. You can auto save granolas to certain folders. So all of my fundraising meetings go to a folder and that folder can be shared. There's a lot of different ways to do this. Shout out to Gary Tan. He's been doing a bunch of really interesting innovation around this with his G stack and G brain project. Jess, you should check out G brain and I can take this offline with you. But there's all. Yeah, I'm interested in what my team thinks about sharing granolas.
SPEAKER_01
[SPEAKER_00] They don't have to share all granolas. You can auto save granolas to certain folders. So all of my fundraising meetings go to a folder and that folder can be shared.
SPEAKER_01
[SPEAKER_00] There's a lot of different ways to do this. Shout out to Gary Tan. He's been doing a bunch of really interesting innovation around this with his G stack and G brain project. Jess, you should check out G brain and I can take this offline with you. But there's also a bunch of cool ways you can embed all of your transcripts and whatever other data you need for workflows like what Britt was talking about into vector databases and then have your Claude or your cloud code be able to recall things like memory in a way that's really powerful. I keep telling it to remember things because Sam taught me that. [SPEAKER_00] Yeah, yeah.
SPEAKER_01
[SPEAKER_00] I like asking it what I did today.
SPEAKER_03
[SPEAKER_00] Yeah, no, that's one of the best use cases for what it's worth: synthesize my day.
SPEAKER_00
And what should I do tomorrow based on what I did today? You know how everyone's getting their agent to send them a daily briefing in the morning with the top news and the weather and all the things.
SPEAKER_01
[SPEAKER_03] Well, someone connected this to a receipt machine to make a physical paper.
SPEAKER_03
[SPEAKER_00] They can leave for their kid in the morning. [SPEAKER_00] On the kitchen table to be like, remember, it's cloudy today. [SPEAKER_00] Pack your jacket. [SPEAKER_00] Here are the three things happening this afternoon. [SPEAKER_01] Make sure your math test is tomorrow.
SPEAKER_00
[SPEAKER_01] I'm also in favor of talking to children. [SPEAKER_01] I also believe that.
SPEAKER_01
[SPEAKER_03] I know. But let's say you're traveling. And you know, you don't want to be the nagging mom on text message, reminding your nanny to then tell your kids. [SPEAKER_03] Yeah, it's a novelty. I got it.
SPEAKER_00
[SPEAKER_03] Okay, that's good. [SPEAKER_03] This is good. I have my AI homework.
SPEAKER_03
[SPEAKER_01] Britt, did you see Devil Wears Prada 2?
SPEAKER_01
A lot of journalism shout outs. [SPEAKER_03] I can't. I invited eight women to go with me and everyone flaked. [SPEAKER_03] I feel we live in flake culture now. [SPEAKER_03] Well, Britt, we are ships passing in the night because last Thursday I had a dinner date with one of our friends who stood me up, but she had a good reason.
SPEAKER_00
[SPEAKER_03] Look at this. What is with you guys and all your flake friends? [SPEAKER_03] Flake culture. Flake culture. [SPEAKER_03] Everyone has a different opportunity.
SPEAKER_01
[SPEAKER_03] She had a good reason. And she's an avid pod listener. So there's no resentment over this. [SPEAKER_03] I know which one it is. [SPEAKER_03] I had a babysitter and I thought, what am I going to do with this free time? [SPEAKER_03] Because it was the only night I had a babysitter for like ten days. [SPEAKER_03] And because we had recorded the pod and you said you were going to go, I had been seeing all the promo, but I had no idea where, when this movie hit theaters.
SPEAKER_00
[SPEAKER_03] I booked myself at the San Mateo Cinemark. [SPEAKER_03] Solo movie date? [SPEAKER_01] Solo. [SPEAKER_01] I think this is the only time I've actually seen a movie alone in my life, which is odd. [SPEAKER_01] I can't think of another one. [SPEAKER_03] I had the best time.
SPEAKER_01
[SPEAKER_03] The theme of the demise of journalism in the layoffs across the industry is the backdrop to this whole thing. [SPEAKER_00] Yes, it's corny. [SPEAKER_00] Yes, it's silly, but it was a delight. [SPEAKER_00] I loved every minute. [SPEAKER_00] I'm thinking of doing this for Mother's Day solo. [SPEAKER_00] You should go. [SPEAKER_00] I had a great time.
SPEAKER_03
[SPEAKER_01] I really enjoyed it. [SPEAKER_00] It's a nice two hours. You know, these movie theaters are really, you've probably been in a while. I don't even know. I've maybe seen you outside of F1.
SPEAKER_00
[SPEAKER_03] I see movies a lot. [SPEAKER_03] No, yeah, they're great. [SPEAKER_03] They have recliners now. [SPEAKER_03] Yeah, it's great. [SPEAKER_03] It was a great experience, the San Mateo Cinemark.
SPEAKER_01
And I have to thank you for the idea. I would have just gone home and missed out on my evening out.
SPEAKER_00
[SPEAKER_01] And so highly recommend. [SPEAKER_01] Dave, what's on your list?
SPEAKER_01
What's on your pop culture viewing? Ella Langley, and she is the first country artist since Taylor Swift to cross over. Her song, Choosing Texas, is number one on the country charts and the Hot 100.
SPEAKER_00
[SPEAKER_01] Yeah, she's a big deal.
SPEAKER_03
[SPEAKER_01] She's great.
SPEAKER_01
You know, she wasn't the best that we saw at Stagecoach a couple of weeks ago.
SPEAKER_02
[SPEAKER_01] She was amazing, but she's clearly young and doesn't have the stage presence of, say, a Lainey Wilson or a Post Malone.
SPEAKER_01
[SPEAKER_03] She's a rookie. [SPEAKER_03] But I think that's pretty rad that she's got this crossover thing happening. Yeah, which is cool. [SPEAKER_00] Yeah.
SPEAKER_03
[SPEAKER_00] And she's a great musician. [SPEAKER_01] Super great musician. [SPEAKER_01] I was also just so wildly impressed with Teddy Swims a couple of weeks ago. [SPEAKER_01] I've always liked his music, but he was unbelievable in person. [SPEAKER_00] Is Taylor Swift doing a Toy Story song? [SPEAKER_01] Yeah. [SPEAKER_01] Yeah.
SPEAKER_01
Yeah.
SPEAKER_00
[SPEAKER_01] Oh, her team messed up. [SPEAKER_01] They accidentally leaked an Easter egg, then they retracted it, but it's obvious.
SPEAKER_01
She has a song. She now has a credit on the Toy Story.
SPEAKER_00
[SPEAKER_01] It's coming out June 19th or 20th, something like that.
SPEAKER_01
Oh, so that countdown was early.
SPEAKER_00
[SPEAKER_01] Yeah.
Yeah. But now we know. I bet everyone's thrilled about that. [SPEAKER_03] Now we know. [SPEAKER_01] I wouldn't have talked. [SPEAKER_01] Yeah, you're welcome. [SPEAKER_01] I wouldn't have talked. [SPEAKER_01] What Toy Story are we on? [SPEAKER_01] Toy Story 5. [SPEAKER_01] And Frozen 3 is happening later this year too. [SPEAKER_01] You know, I have watched and enjoyed the first two Frozens back when. [SPEAKER_01] But you know what's not going well in the spirit of May the 4th be with you? [SPEAKER_01] Star Wars.
SPEAKER_01
Yeah. Yeah. But now we know.
SPEAKER_00
[SPEAKER_01] I bet everyone's thrilled about that. [SPEAKER_03] Now we know. [SPEAKER_03] I wouldn't have talked. [SPEAKER_01] Yeah, you're welcome. I wouldn't have talked. What Toy Story are we on? [SPEAKER_01] 11? [SPEAKER_01] Toy Story 5. [SPEAKER_01] And Frozen 3 is happening later this year too. [SPEAKER_01] I have watched and enjoyed the first two Frozens back when. [SPEAKER_01] But what's not going well in the spirit of May the 4th be with you?
SPEAKER_00
[SPEAKER_01] Star Wars. [SPEAKER_03] Star Wars is not going well. [SPEAKER_01] What do you mean not going well? [SPEAKER_01] It's Star Wars. [SPEAKER_01] What are you talking about?
SPEAKER_01
No.
SPEAKER_00
[SPEAKER_01] All the newer movies are not, the new characters aren't really sticking with people. [SPEAKER_01] So the Disney people are having to revive it all. [SPEAKER_01] Yeah, there's a rumor that they might split the three sequels off into their own timeline or something because it's been going so badly. And the Mandalorian movie bombed in the box office. And so there's potentially a new storyline that they have to generate because everybody just wants Luke Skywalker and the original characters.
SPEAKER_03
They don't want all these Reys and all these new things.
SPEAKER_00
There's a limit for how long you can stretch a franchise, but it's not good. If Star Wars can't make it work, I don't know what the prospects for the rest of Hollywood are. Yeah, I mean, it's a bummer. I thought that this Mandalorian, I don't know, I'll probably still go see it with the kids, but I thought maybe it was going to be good, but it looks like it's bombing pretty badly.
SPEAKER_01
[SPEAKER_00] There you go. [SPEAKER_00] Okay. Dear friends.
SPEAKER_03
[SPEAKER_00] Well, it's always a pleasure chatting about exactly the same topics with you every week. [SPEAKER_00] Having exactly the same argument. [SPEAKER_00] It's a Twilight Zone.
SPEAKER_01
[SPEAKER_00] What's going to happen to me?
SPEAKER_03
[SPEAKER_01] Job displacement. [SPEAKER_00] Just kidding. [SPEAKER_00] We're moving the ball forward.
SPEAKER_01
[SPEAKER_00] People, we're grateful for all you listeners out there. [SPEAKER_03] Please do.
SPEAKER_03
We do want to hear from you, especially I think on calibrating the tech sentiment and what to do about it. So with that, we will say farewell from now. Thanks for listening and watching, and we'll see you back here next week for another episode of More or Less. [SPEAKER_00] Bye. Bye guys. [SPEAKER_01] See you later. [SPEAKER_01] If you enjoyed this show, please leave us a virtual high five by rating and reviewing it on Apple Podcasts, Spotify, YouTube, or wherever you get your podcasts.
SPEAKER_01
Find more information about each episode in the show notes and follow us on social media by searching for at More or Less, at Dave Morin, at Lesson, at Jay Lesson.
SPEAKER_03
[SPEAKER_01] And as for me, I'm at Brit. [SPEAKER_01] See you guys next time. Because it was the only night I had a babysitter for like 10 days. And for some, because we had recorded the pod and you said you were going to go, I had been seeing all the promo, but I had no idea where, where, when this movie hit theaters. I booked myself at the San Mateo Cinemark. Solo movie date?
SPEAKER_01
Solo. I think this is the only time I've actually seen a movie alone in my life, which is odd. I can't think of another one.
SPEAKER_03
I had the best time. The theme of the demise of journalism in the layoffs across the industry is like the backdrop to this whole thing.
SPEAKER_00
Yes, it's corny. Yes, it's silly, but it was a delight. I loved every minute. I'm thinking of doing this for Mother's Day solo. You should go. I had a great time.
SPEAKER_01
I really enjoyed it.
SPEAKER_00
It's a nice two hours.
SPEAKER_03
You know, these movie theaters are really, you've probably been in a while. I don't even know. I've maybe, I see you outside of F1. Oh, I got a movies a lot. No, yeah, they're great. They have recliners now. Yeah, it's great. It was a great experience, the San Mateo Cinemark.
SPEAKER_01
And I have to, I have to thank you for the idea. I would have just gone home and, you know, missed out on my evening out. And so highly recommend. Dave, what's on your list? What's on your pop culture viewing? Uh... Step away from the claw. Ella Langley, and that she is the first country artist since Taylor Swift to cross over. Like her, her song, Choosing Texas, is number one on the country charts and the, and the Hot 100. Yeah, she's a big deal. She's great. You know, she wasn't the best that we saw at Stagecoach a couple of weeks ago. She was amazing, but she's clearly young and doesn't have the stage presence of, say, like,
SPEAKER_03
a Lainey Wilson or a Post Malone. Uh, she's a rookie. But, you know, I think that's pretty rad that she's, she's kind of got this crossover thing happening.
SPEAKER_01
Yeah, which is cool.
SPEAKER_00
Yeah. And she's a great musician.
SPEAKER_01
Super great musician. I was also just so wildly impressed with Teddy Swims a couple of weeks ago. Like, I, I've always liked his music, but he was unbelievable in person. Like, just kind of the, the, the...
SPEAKER_00
Is Taylor Swift doing this, a Toy Story song?
SPEAKER_01
Yeah. Yeah. Yeah. This got, oh, her team messed up. They accidentally, like, leaked an Easter egg, then they retracted it, but it's obvious. She's, has a song. She's, and now she has a credit on the Toy Story. It's coming out June 19th or 20th, something like that. Oh, so that countdown was early. Yeah. Yeah. But now we know. I bet everyone's thrilled about that.
SPEAKER_03
Now we know. I wouldn't have talked.
SPEAKER_01
Yeah, you're welcome. I wouldn't have talked. What Toy Story are we on? 11? Toys 5. And Frozen 3 is happening later this year too. You know, I have watched and enjoyed the first two Frosens back when. But you know what's not going well in the spirit of May the 4th be with you? Star Wars.
SPEAKER_03
Star Wars is not going well.
SPEAKER_01
What do you mean not going well? It's Star Wars. What are you talking about? No. Like all the newer movies are like not, the new characters aren't really sticking with people. So like the Disney people are having to revive it all. Yeah, there's a rumor that they might split the three sequels off into their own timeline
SPEAKER_00
or something like that because it's been going so badly.
SPEAKER_01
And like the Mandalorian movie bombed in the box office.
SPEAKER_00
And so there's like this potentially new storyline that they have to generate because everybody just wants Luke Skywalker and the original characters.
SPEAKER_03
They don't want all these rays and all these new things.
SPEAKER_00
You know, there's a limit for how long you can stretch a franchise, but it's not good. If Star Wars can't make it work, I don't know what the prospects for the rest of Hollywood Yeah, I mean, it's kind of a bummer. I thought that this Mandalorian, I don't know, I'll probably still go see it with the kids, but the, I thought maybe it was going to be good, but it looks like it's bombing pretty badly. There you go. Okay. Dear friends. Well, Morin's it's always a pleasure chatting exactly the same topics with you every week. Having exactly the same argument. It's like a twilight zone. What's going to happen to me?
SPEAKER_01
A job displacement.
SPEAKER_00
Just kidding. We're moving the ball forward. People were grateful for all you listeners out there.
SPEAKER_03
Please do. We do want to hear from you, especially I think on the calibrating the tech sentiment and what to do about it. So, but with that, we will say farewell from now. Thanks for listening and watching, and we'll see you back here next week for another episode
SPEAKER_01
of more or less.
SPEAKER_00
Bye.
SPEAKER_03
Bye guys.
SPEAKER_01
See you later. 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 at more or less at Dave Morin at lesson at Jay lesson. And as for me, I'm at Brit. See you guys next time.