Arnold Ventures Co-Chair @JohnArnold on what’s stopping compute from becoming a real commodity market: “Compute is not a commodity where one data center is providing the same product as another data center.” “You either have to deliver into a certain product that has a lot of buyers and a lot of sellers… or you need some type of index that the industry trusts.” “There just aren’t that many short-term deals. Transparency into those is difficult.” “The distinctions between the chips and the design of the data center and all the specific needs is just different.” “Trying to either come up with the physical delivery mechanism or the index that everybody trusts to be right is really hard.”

SemiAnalysis ClusterMAX Lead @JordanNanos explains how Nebius picked up demand CoreWeave couldn’t handle and then executed well enough to reach the Platinum tier of its GPU cloud rankings: “CoreWeave has set the standard. I mean, they are the best in terms of technology.” “These guys deploy 10,000 GPUs a week at the peak when they’re building data centers. That’s incredible.” “But their balance sheet is full. There’s only so many people and resources you can marshal to build out another gigawatt when they’re going to 15 and beyond.” “We have customers coming to us and saying, ‘I saw you write such good things about CoreWeave. I just have $100 million I’m trying to give to them. They’re telling me they can’t accept it until May of next year because they’re backed up.’” “And therefore Nebius has been capitalizing.” “They’ve also executed very well. The technology that they’ve developed is strong.” “We think that managed clusters are approaching this feature-complete world where it’s hard to tell the difference in quality between Nebius and CoreWeave.” “Certainly our testing experience has been really solid on both of them.”

SemiAnalysis' @JordanNanos predicts that chip startups will become the next wave of neoclouds, building their own data centers. "Cerebras has become a neocloud, Groq is buying Nvidia GPUs and spinning up a neocloud. Others like SambaNova are doing this themselves. We see OpenAI Jalapeño that'll go into self-build sites." "Chip companies are like, 'Not only do I need to battle in the supply chain for allocation like TSMC and the memory providers, I also need to go get my data centers set up. I need to learn about construction, I need to learn about day-2 operations.'" "Who's going to train people to repair a Cerebras wafer-scale engine and their 250 megawatts of compute they're bringing online in the next couple of years? It's them." "They need to build that capability in-house and frankly, I think they're going to realize a bunch of benefits to their margin profile for doing this. AWS and Google already proved you can do this with TPU and Trainium."

SemiAnalysis' @JordanNanos on the neocloud talent wars: "It's across the entire supply chain. We see this in electricians, data center technicians, and operators. Salaries for electricians in Louisiana and Abilene, Texas, are up 3x to 5x." "Crusoe just came in and paid these guys so much more money. But they're out of people. They need to train. No matter what job we're talking about, people just need to be trained." "The most successful neoclouds that we're seeing have this pipeline to take talent from other related areas and get them contributing to neocloud management and day-2 operations where a software engineer can start to work as an SRE on one of these teams."

Happy Thursday. On today's show: - @CompleteSkeptic (TypeSafe) - @JordanNanos (SemiAnalysis) - @leifthunder (Public) - @hoomanrenezhad (Solcoa) - @erika_alden_d (Pioneer Labs) - @Shalev_lif & @RomiLifshitz (Enclosure) See you on the stream.

SCOOP: @petkevichdan says a senior Tesla employee told him that one of the constraints on the next-gen Roadster’s acceleration is the human body's tolerance for G-force. “The goal is to not kill the driver.”

Jordi: self-driving cars need a humanoid. “We’ve solved the task, but not the job.” He gives the example of hiring a driver when fires doubled his commute to two hours each way, and his favorite part was everything but the driving. “Everything that I loved about it was not actually the driving itself. It was having someone in the real world that could just do things.” “It was parking the car when you get there. It was running in to grab food, running to pick something up, dry cleaning. It was all that stuff.” “You really want the Waymo or the Tesla with the humanoid in it, or the ability to close the loop and park, or go run into the store and grab something, or do things.”

.@benthompson says consumer AI has a marketing problem because the average consumer doesn't actually want to be more productive. "Tech companies are like, 'Let us book your flight for you!' I’ve had a personal assistant for 10 years. He does not come within an inch of booking my flights for me. That is crazy talk." "I understand which ones are gonna be delayed. I know which seat I want to be on. So the fact that this is continually a demo is hilarious to me." “Number two: people love shopping. They don’t want to be productive. They want to be entertained. One of the biggest forms of entertainment is shopping.” "Also, there’s so much information on a product page. Companies have spent so much energy focusing on that. You look at apparel as a category. They’re going to tell you what size the model is wearing, and you want to know that, and you want to look at the different styles." "The idea of having a personal agent that’s going and getting some of the information and bringing it back to you, and then you’re like, 'Well, what about this other thing? Tell me about this thing.'" "And you have your eager tech company executive say, “Oh, I have an assistant that can do it for you!" "And it’s like, no. Not only do I want to do it, I want to complain about it. I’m enjoying myself complaining right now, okay? Would you just let me be and live my life?"

.@benthompson gives his recommendation on how Meta should market Muse: "The way to frame this is to say: 'We're not helping you be more productive, we're helping you get rid of annoying crap in your life.'" "It needs to be framed more as a pain relief as opposed to this aspirational, 'I'm gonna be so productive and get things done.'" "People don't care about being productive. Employers care about their employees being productive. But once you're off the job, you just want to go home and watch TikTok." "The way I think this is going to land is through Instagram Reels or TikTok with people actually showing stuff they do and people say, 'That's amazing. That would make my life so much easier.'" "It really is a show-not-tell sort of technology."

.@benthompson says his "most profound experience" with AI came from building "GeckoBot," an agent that lets him offload tasks for his personal assistant whenever they come to mind, then briefs his assistant when he's back at work. "I'm a massive believer in not obsessing about your weaknesses. Double down on your strengths and then pay someone to take care of your weaknesses. That is generally my life philosophy." "When I built GeckoBot, I would be doing something on the weekend and think, 'Oh, this needs to be done.'" "But I felt bad texting my assistant on the weekend. So I would tell the bot what needs to be done, then when he shows up Monday morning, it sends a briefing with all the stuff that I wanted over the weekend. His joy and relief was unbelievable." "It's meaningful to see him happier and his life tangibly better. He's like, 'This is the greatest week of work of my life.'" "I don't want to take advantage of his predilection to do his job well and let him enjoy time with his family."

"People in tech don't understand how actually awesome Walmart is." - @benthompson "One, their stores are like this crazy dystopian experience." "Two, if you go in their stores, their app is unbelievable. It will guide you through the aisles to tell you exactly where to go to get the thing that you need. It's amazing. It's the best in-store app by far compared to anyone else." "If what you want to buy is in their store, you can get it delivered in an hour. Who knew? No one in tech who's going to be buying stuff or who is technically inclined through ChatGPT actually knows how good the Walmart e-commerce experience is."

Jordi crashing out over the new McLaren logo: "This feels bad... a pretty brutal rebrand." "They seem to be trying to reposition as a luxury brand versus a motorsports brand." "It's tough because I always thought their brand was awesome."

Box CEO @levie says AI has been “unequivocally” a net positive for software. “You look at infrastructure providers—the Cloudflares of the world—totally on fire, because agents need sandboxes, they need compute, they need network, they need gateways. Great business.” “For us, we have reaccelerated growth far past our internal plans because it turns out that enterprises need core systems to be able to manage their unstructured data.” “Agents need to be able to work with that unstructured data to make decisions or move information through a workflow—whether it’s all of your contracts, your research materials, marketing assets, or financial documents.” “Now imagine an enterprise with 10,000 employees. How do you ensure that those agents continue to go after the actual canonical records, the actual documents that are the real sources of truth and the authoritative versions of that data?” “That means you need platforms.” “And all of that is creating more value for these platforms.”

The Paramount settlement was a big win for David Ellison and a pretty underwhelming result for California Attorney General Rob Bonta, according to Bloomberg's @Lucas_Shaw. "[David Ellison] was looking at months of uncertainty. This trial wasn't going to start until next year. If that had happened, he was going to owe hundreds of millions, if not billions of dollars to this ticking fee that he had in the deal between Paramount Skydance and Warner Bros. Discovery." "Instead, he mostly agrees to do things he's already said he was going to do. Release 30 movies a year in theaters. Only 20 of them have to be wide releases. There's a commitment to make a certain amount of movies and television in this country. But again, I think he was going to do that stuff anyways." "It doesn't seem like Rob Bonta, any of the attorneys general who are involved in the suit, or the Writers Guild got some big win." "They were calling for structural changes, which would have meant selling something, and there is nothing being sold."

.@vkhosla's prediction for the two moats of personal AI: 1. A company you can trust with your data 2. A tool that actually gets the job done “I believe the moat will be trust in a company that develops trust with consumers. And that’s where Meta has a huge disadvantage.” “I also think completion is going to be really important. Consumers will be loyal to products that get their shit done." “If you can get all my stuff done, I’ll be loyal to you.”

.@vkhosla thinks consumers will need a personal AI whose only job is to represent their interests against corporate AI. "An AI that’s loyal to the consumer—not to a particular company, a particular product, a particular point of view—is going to be very, very important.”

“The right agent won’t be an app.” – @vkhosla “Where do most people live? They live in their email, they live in text messaging, they live in Slack.” “The app, if there’s one, is just in the background.” “Meeting a consumer where they are—the consumer experience is really, really important.”

.@vkhosla predicts personal AI agents won't be a winner-take-all market. "[Personal agents] will be a broad category. I do think it'll be very diverse and I don't think there will be a winner-take-all." "Every one of the big labs, Meta and Google included, will want to have their own agents. They will have different trade-offs and I think we'll see a lot of diversity in requirements." "As to Alex's comment, I think he hasn't used agents enough to understand much of that diversity that he's thinking about." "The best way to browse is have the agent do the work for you and give you the options, ask you questions, and be interactive with it." "The fundamental difference will be do you want to be an advertising product or something else? Do you want to trust Meta with your data?" "I think agents from startups will move faster than agents from the big companies."