The Ledger
Every statement that passed quotation and attribution checks. Mix any filter with any other: certainty 1/5, debate potential 5/5, or both at once.
why aren't all 967 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 26 checkable ones are still open, waiting for their date. predictions held up or didn't; assertions are supported or contradicted. on every card: ▮▮▮▮▮ certainty · ▮▮▮▮▮ debate potential. speakers are clickable
Osika: Open source is overrated because AI models can generate code on-demand
“Unfortunately with AI, it's overrated, like the, an AI model can just spit out open source code, so to say.”
Vogt: Moving into a home without a robot will seem strange soon
“I think it will be strange to move into a home or apartment in five years that doesn't have a home robot, or you won't want to without one.”
Cruise founder Vogt: Tesla's end-to-end neural network approach is right
“What I see is, is really Tesla as a company who kind of pioneered the end-to-end neural network approach to self-driving, which I think is the right technical bet long-term, but they put some constraints on it.”
Cruise founder Vogt: Waymo's classical robotics architecture is wrong
“With Waymo they started off in the DARPA Grand Challenge era of self-driving, which is old school, classical computer vision, classical motion planning, and they built this highly validated, robust system that's now on public roads, and it's great. But they kn…”
Vogt: Next $100B company founded in 2025-2026 will have under 100 people
“I think the next hundred billion dollar company that's created, you know, in 25, 26 will be under a hundred people.”
Upcoming Spectacles will match Apple Vision Pro capabilities in glasses form
“And I think the sweet spot for specs coming later this year is the wearability of, you know, normal glasses, but the capability of you know, true spatial computer, like something like a Vision Pro.”
Rapidly created bespoke AI software will replace one-size-fits-all applications
“And so where, where I think we're at in 20, 26, which is fundamentally really different, Is that, you know, there's literally going to be bespoke software for people that they're going to be able to imagine that they're going to be able to build and create rea…”
Combining social connections with media feeds creates perverse platform incentives
“The most important thing you need to do is separate social from media. The biggest mistake was combining them, and it creates a lot of perverse incentives.”
Future social networking will mean hanging out in person using AR glasses
“This social network of the future is actually hanging out together in person and using specs together. So that's what excites me about the future of digital communication. I think, you know, messaging is fine, but it's hard to compete with actually being toget…”
Companies will shift resources from engineering to distribution to accelerate growth
“People are going to reallocate more resources away from things like software engineering, for example, to distribution in order to grow faster.”
Xu: Takeaway generates 3x to 5x higher incremental margins for restaurants
“Rough math, you know, for every dollar of food that if you and I bought inside of a restaurant, you know, from the cost structure, you're gonna have 30% of that in the food and the packaging. You're gonna have 30% of it in the rent. You're gonna have 30% of it…”
Tony Xu: Ghost kitchens fail without large brand customer acquisition power
“It just turns out it's extraordinarily difficult, however, unless you're a large brand, or, you know you know, a house of brands, someone like DoorDash, to be able to attract enough customers to make that math work.”
Xu: DoorDash's Dot robot uniquely navigates roads, sidewalks, and bike lanes
“And the sidewalk and the bike lane. And so it's the only vehicle in the world that does that.”
Collison: Stablecoins are already the best option for cross-border payments
“Two places that we're seeing tons of interesting stuff is one, any kind of cross-border use case, where if you want to send money to a hundred countries, it's already the case that stablecoin's a much better way to do that than anything else. And that's where …”
Collison: Stripe is building Tempo for scalable AI agent crypto payments
“The second, and this is where Tempo, our new initiative, That you guys were nice enough to work with us on is coming in is very scalable crypto payments, and obviously agents is what we have in mind as we're developing Tempo, where, you know, if you want to pa…”
Xu: Google's food ordering failed because it couldn't manage post-checkout logistics
“I remember services, you know, Google food ordering launched in, I forget, 2015 or 16, and they allowed you to order delivery, for example from various Google services, you know, Google Maps, Google Search, et cetera, and while on the one hand it was low frict…”
Xu: Chat interfaces fail in commerce unless they solve the end-to-end job
“It's not obvious to me that it's just a chat interface, that there are, you know, all of these different components that have to be solved. And so I think You know, the way I think about it is, okay, regardless of whatever we ship whether it's on our own you k…”
Staniszewski: ElevenLabs hopes to pass voice Turing test within a year
“Our goal is to like pass the voice Turing test in all those cases, or the Turing test for all conversational agents outside of voice two. And I hope we will be there in the next year or so.”
ElevenLabs bets research on cascaded voice architecture over speech-to-speech
“So today we are optimizing heavily on a cascaded approach... So that's like where we are betting a lot of the research work of how you can make that great, and we think we can make that great.”
Staniszewski: Speech-to-speech models are 'definitely dumber' due to size limits
“They are definitely dumber. You need smaller model, you cannot... If you are going speech to speech, usually you will use smaller models, so it's still quick.”
Collison: High-Agency People Will Win From AI Advances, Low-Agency Will Lose
“My biggest takeaway from all this Has been that around agency where I feel like high agency people are the winners of the advances in AI and within organizations, low agency people will lose out.”
Pichai: Google held back LaMDA due to lack of RLHF and toxicity
“In fact, in the Google I.O. In, in 22, We launched something called AI Test Kitchen, and that was Lambda, but we had constrained it because internally we didn't have an end-to-end version which was RLHFed, right? So the version I saw was a lot more you know, t…”
Pichai: Leading memory makers cannot rapidly expand short-term capacity
“There is no way that the leading memory companies are going to dramatically improve their capacity. So you have those constraints in the short term, but they get more relaxed as you go out.”
Pichai: 40 million Americans will soon have Wing drone delivery access
“In some reasonable time period, like, forty million Americans will have access to a Bing delivery service, right? And I'm not talking years out or something like that.”
Pichai: First-party hardware will be very important in regulated AI and robotics domains
“My lesson from Waymo and on the AI side with TPUs, et cetera, I think to really push the curve well, particularly in areas where you have safety, regulatory, everything, you want the firsthand experience of the product feedback cycle. So I think having first p…”
Cacioppo: Security companies targeting startups should sell compliance instead of security
“One of the original, like, founding hypotheses of the company is if you want to start a security company for startups, you should actually start a compliance company. Because your customers never ask you for Security, but they do ask you for compliance.”
Cacioppo: Current market size only predicts today's market, making TAM sizing misleading
“I think the second part is market sizing is bullshit. You know, you can, like, be as academic or whatever, strategery-ish as you want about it, and, like, the market size today is only a predictor of the market size today.”
Dolgov: Autonomous vehicles require closed-loop RLFT, not just imitation learning
“And if all you did was kind of observing how other people drive when you trained the system, maybe observing, you know, just passively how people drive and how they interact maybe also, you know, driving the car yourself and then using imitative learning to tr…”
Dolgov: Waymo core driving technology has no remaining foundational gaps
“I would say that we've clearly moved past the stage of scientific research and kind of deep core technology development to this new phase of accelerated global scaling and deployment. So, you know, we still have work to do, right? But I don't see today any lim…”
Dolgov: Waymo Gen-6 Hardware Costs a Fraction of Gen-5
“It is very different from the fifth generation. It is simpler. It is more capable. It is much lower cost. It's thinking a fraction of the cost is, you know, comparable to what you would get like a fancy ADA system.”
Dolgov: Waymo driver tech will eventually reach every US USPS address
“Eventually it will. Absolutely. There's no doubt in my mind. I think it's just a matter of when.”
Lopes Lara: 2024 election showed prediction markets outperform traditional polling
“The polls were completely wrong and the markets were so much better at bringing that sort of information, and I think that it's the shining example of why these markets are forced for good, and we need to have them in the US and regulated”
Lopes Lara: Over 95% of Kalshi maker orders come from peer-to-peer traders
“Over 95 are just like. Peer-to-peer, like, or funds that have like two people that just got stuck.”
Mansour: Kalshi's top inflation forecaster was an amateur from Kansas
“The best inflation forecaster on CalSphere over the last few years is not, none of the institutions or the, you know, the big name hedge funds, it's this guy who lives in Kansas, never traded financial markets before, just likes to read the news, and just know…”
Mansour: Gambling houses profit from losses, exchanges profit from transparency
“Gambling is this idea where the business model is, you are the house, and your revenue is your customer's losses. Like, so, and a lot of the dynamic that you describe has to be true, because your incentive is like, well, somebody's making money, I gotta stop t…”
Lopes Lara: Prediction markets will improve polling accuracy, not destroy it
“My take is that polls are just going to get a lot better, because what people are going to be is like, okay, I can make money if my polls are right, so I'm just going to commission this poll, and I'm going to do this, And now you can actually, like, compete a …”
Mansour: Prediction markets are required to accurately price complex traditional assets
“And if those dimensions, you don't have a good understanding of x one to x n you cannot get a good estimate of y, right? And so you, the paper basically says that you need infinite markets. And prediction markets are this notion of infinite markets, which is l…”
Lopes Lara: Kalshi will never list war, terrorism, or assassination markets
“There are a lot of things that we wouldn't do, like wildfires we don't do, war, terrorism, assassination. Those things are bad, and like, there's a moral side of these markets, and we're not going to ever go there.”
Mansour: Prediction markets are an antidote to social media polarization
“I think that we do see this a lot with also, well, there's sort of the piece where people use it and that sort of react in real time to certain things, but I think there's some degree of depolarization. And that's what Luana is alluding to. Like, and in some w…”
Taylor: Coding agent success doesn't translate to broad knowledge work
“And so it's not like impossible for an agent to use those things, but the idea that you can like straight line from coding agents to writing the Stripe annual letter, I don't totally buy.”
Taylor: Mimicking codebases is currently the best harness for AI agents
“I wonder if in the short term, it might just be one of those idiosyncrasies of history, like mimicking a code base is actually the best way to make a general purpose agent work. And maybe over time we'll get fancier than that, but it's actually like a relative…”
Taylor: In agentic development, PRDs and intent are more durable than code
“Where there's a lot in, in the code that is more transient. You know, you might be fine to Delete that. What was the intention? What was the PRD? Like, what was the customer problem? Is actually the more durable asset.”
Taylor: Future engineers cannot be precious about writing code by hand
“I feel like I won't be like a self,
Self-actualized software engineer in the future, if I'm too precious about that artifact, which used to be so central to me.”
Taylor: Hierarchical multi-agent architectures are elegant on whiteboards but nonsensical
“I feel like this view of a multi-agent world was you have all these agents that do tasks that are fraud detection. You know, another one over here for, you know, personalization, and then you make a super agent, it does all these things, and it looks really go…”
Taylor: Future AI agents will evolve away from Model Context Protocol
“My sense is we're making true agents over time the way we think about context and how that context is sort of like shared so that the agent that's orchestrated it actually understands sort of what's behind all these APIs and why and the history will maybe look…”