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 560 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 0 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
Derek Thompson: California's high-speed rail project is dead
“If you're asking for my personal opinion, This project is dead.”
Google cannot replicate Perplexity directly on its cluttered search homepage
“They cannot make this exact product on the Google homepage. It's so hard to know when a query is purely informational or not, and then the Google search page is already, like, so cluttered, that's the answer box, the knowledge panel,”
Google has better consumer product taste than both OpenAI and Anthropic
“Out of the list you mentioned, Google is the only company that actually has the product taste to do this.”
Altman: No Great Businesses Were Built on GPT-3 Except Copywriting
“But, with the possible exception of copywriting, no great businesses were built on GPT-III.”
Altman: The world is astonishingly asleep on the AI platform shift
“I think that's why I'm so excited for startups right now. It is because the world is still sleeping on all of this to such an astonishing degree. And then you have, like, the YC founders being like, no, no, I'm gonna, like, do this amazing thing and do it very…”
Tan: OpenAI's $9/hour real-time voice API threatens call center-reliant national economies
“The Ongoing usage-based pricing is nine dollars per hour, and it sort of points to a sort of powerful thing. Like, if I were a macro trader, I would be very, very bearish on countries that have, that rely very heavily on call centers right now because, you kno…”
Friedman: AI is the fastest-improving technology in human history
“It's the fastest any tech has ever improved. I think. Yeah. Certainly faster than processors, certainly faster than the cloud.”
Tan: Open-source AI access for consumers is the best insurance against tyranny
“We want all consumers to be able to have from the bottom up, ah, the same access to that same technology. And that's, ah, yeah, the best insurance against tyranny.”
Heller: GPT technology is underhyped because GPT-4 operates at postgraduate level
“So this is gonna sound probably insane to most people who hear it, but I think the GPT technology is under hyped because what we see in GPT-IV specifically as a model, and I'm assuming the same will be true as new models come out in the state of the art advanc…”
Habib: Foundation model barriers are capital and talent, not secret sauce
“Like to me, the barriers to entry of training, one of these models are mostly capital and talent. Like the people needed are still very specialized and very smart and you need lots of money to pay for GPUs. But beyond that, I don't see that much secret sauce, …”
Taggar: FAANG Product Launches Rarely Kill Fast-Growing Startups
“Cause like, if you think about like, we're like scratching our heads, trying to find examples of times that like a fan company launched a product that killed a fast growing, successful startup. So like clearly more often than not, like structural advantage isn…”
Best: Consumers now prefer paying for high-quality content
“I think the pendulum's really heavily swinging the other way now. Like people that were sort of like my generation were like the least likely to want to pay for stuff. And we are getting increasingly more likely to want to pay. And then younger generations are…”
Allred: Human capital is the most unoptimized asset class in history
“Human capital is the most unoptimized asset class we've ever had.”
Full-Stack Insurtech Startups Will Likely Fail and Burn Cash
“Folks who are attacking, you know, every aspect of this at once are probably not going to get any of them right and burn a lot of money right out the gate.”
Khosla: Many investors prioritize short-term revenue over long-term strategic assets
“My big beef with a lot of investors is they want revenue, They want to meet plan as opposed to collect assets for this larger ascent to Mount Everest.”
Susskind: String theory cannot describe real particles without modification
“We know that the formal, by formal I mean mathematically rigorous structure that string theory became, it became a mathematical structure of great rigor and consistency that It, in itself, as it is, cannot describe the real world of particles. It has to be mod…”
Susskind: LHC gave no new information, leaving particle physics stalled
“We seem to have run out of new experimental data, even though there was a big experimental project the LHC at CERN. What if that is a great big machine that produces particles and collides them. And I would say I don't want to use the word disappointingly. Wel…”
Susskind: A vast, varied multiverse is the only sensible cosmological explanation
“No, it's not mine, but I think it's the only thing that does at the moment seem to make any sense, is to say the universe is extremely big, much bigger than we can see, and varied. Varied means it has properties which are different from place to place.”
Susskind: Stephen Hawking was wrong about information loss in black holes
“Stephen Hawking's very, very brilliant insight, even though I think he got the final answer wrong. It was very brilliant insight to ask what happens to the information that goes into black holes?”
Mathilde Collin: Slack cannot replace email for external business communication
“Now, in a work environment I feel like email, the protocol is actually perfect because it's the only protocol that you can use to communicate with people outside your company. And so even if you have great tools like Slack that enable you to Slack in, to messa…”
Segment co-founder Ian Storm Taylor predicted a massive business behind Analytics.js
“I think there's a big business behind Analytics.js, which is this routing library.”
Reinhardt thought building a business on Analytics.js was his co-founder's worst idea
“And I was like, that is literally the worst idea I've ever heard. Like, it's 500 lines of code. You've grown a little bit. It's 500 lines of code and it's already open source. I have no idea how you build a business around that.”
Graham: Capable high schoolers should not start companies prematurely
“There are plenty of high school students who could start successful startups but they shouldn't.”
Amazon made a critical error by competing against its third-party vendors
“I think platforms should be neutral. And so the idea that you compete with your vendors, to me, you've got a business needs to keep its, All of its major stakeholders happy. Yeah. And your vendors are one of your key stakeholders. If you don't treat them like …”
Altman: Virtual reality is a fake trend because buyers rarely use headsets
“A fake trend, or at least a fake trend as of, ah, August, 2018, I would say is VR. I do believe VR will be big someday, but today most people that I know, ah, that own a VR headset use it never or very rarely.”
Aaronson: Human stupidity is a bigger near-term threat than superintelligent AI
“When I think about, like, the future of civilization, you know, let's say the next 20 years, the next 50 years, I tend to worry less about super intelligence than I do about super stupidity. You know, I tend to worry about, you know, killing ourselves off or y…”
Benjamin: Modular architecture is the only viable route to rapid quantum scaling
“And in fact, this network approach is, I think the only one that if you wanted to Manhattan project it, right, or moonshot it, you're one of these big projects that have happened in the past where you have a goal that's extremely challenging and you just decid…”
Xie: Very few crypto projects actually need to conduct token sales
“So a very, very few times is it reasonable, in my opinion, to do a token sale. There's a lot of really great projects that can do decentralized applications or create their own blockchain or whatever, but they don't actually need to do a token sale. They can j…”
Brundage: Technical transparency in AGI could enable international non-aggression pacts
“And I think if you actually had the full development of the FAT methods, and you had accountability and transparency for even general AI systems or superintelligent systems, I think that would open up the door for a lot more collaboration. If you could sort of…”
Schacht: Teaching science primarily to fill STEM job shortages is mistaken
“I think it is an absolute mistake to say that we should teach children more science because there's a lot of STEM careers out there that, that, that need to be filled. Like, that's absolutely the mistaken approach.”
Watts: Antidepressant research is often short-term to serve drug manufacturers' interests
“The research into the effect of antidepressant drugs often is very short-term. It looks at them for a very short period of time without much follow-up. So because that serves the purposes of the people making these drugs.”
Altman in 2018: Virtual reality isn't a trend because headsets sit unused
“If you contrast that to virtual reality, everyone talks about it as the next trend. It may be in the future. But today, if you talk to people who have VR headsets they don't use them every hour. They don't use them every day. Most of them don't use them every …”
Altman: VR in 2018 Is in the Palm Pilot Era, Not iPhone
“Doesn't mean that VR will not be big someday, but this is, you know, I don't think you can make a high conviction bet on starting a VR company today, at least that the, or at least that the TAM will be huge. It still feels like we're at the Palm Pilot, not the…”
Sidor: Rigorous baseline optimization will advance AI more than complex architectures
“And you know, it's not the kind of sexy research that people want to see, where you have like some hierarchy of big RNN, but it actually, this kind of research, I think at this point will advance field the most.”
Sidor: Strong Engineering Is Much More Valuable at OpenAI than Writing Exotic Models
“So, so essentially becoming a good engineer for our team is much more valuable than, for example, people spending you know, months upon months implementing exotic models in TensorFlow.”
Brockman: Solid Engineers with Zero AI Experience Are Productive Day One at OpenAI
“So someone like that can actually become productive from day one.”
Hoover: Living in virtual reality worlds will be a good thing
“I think the world where we live in VR is actually a good thing. And I know a lot of people get really scared and nervous about like a future where we all live in like a ready player one or at least like some sort of virtual world like that. I actually think it…”
Mr. Money Mustache: US car usage should be cut by 90%
“You know, in my personal views, for example, I've become more hardcore over the last year that cars are just like the biggest stupid inefficiency that we have in the United States, and we need to just cut that whole shit down by at least 90%.”
Iorns: Biotech Startups Do Not Need Business Co-Founders
“People have this Kind of illusion that they're, like, antisocial or, like, they don't understand business, and I just, like, completely disagree with that. I think scientists have to be very articulate. They frequently present in front of large audiences. They…”
Altman: Investors rarely do much despite big promises
“Investors sort of like write the check. And then despite a lot of big promises don't usually do that much, sometimes they do. But your employees are really the ones that build the company over years and years.”
Chaubard: AI models are recursively distilling into one another, Claude into Kimi
“What I think is kind of happening right now, but like clawed kind of mother birds into, Kimmy too, and then now Kimmy too is post-training thinking machines, and so it just keeps going and going.”
Wang: VCs who passed on Scale AI now write data think pieces
“Every time we would go out to fundraise, even though our numbers were great and we had great revenue, you know, VCs and investors would always be very skeptical. They'd be like, Oh, I don't know if this is a good business. Does it have longevity? Is this durab…”
Tan: Meta's Muse Spark matches Claude Opus at one-eighth the cost
“When I was using you spark with my open claw, like it became clear that it was as good as Opus, especially for that sort of agentic flow with skill files, but it was like eight X cheaper actually.”
Huang: Controllability is the single biggest breakthrough needed for AI agents
“And so I think controllability is probably the single biggest breakthrough that we need for agents at every single level.”
Tan: All competitive alpha lies in building custom, self-improving AI models
“I feel like all the alpha is in building your own AI. I mean, if someone else is using whatever is off the shelf, but you're, you have a thing that can recursively self-improve, and it is, you know, I mean, the mechanic, people are very flippant about markdown…”
Tan: AI models before Opus 4.5 were toys; newer models are peers
“In the Opus 4.5 and later moment, like I would, you know, I would trace it to like that model was the thing that changed everything for me. The things before that were usable, but like toys that release, there was something magical that happened. And now I can…”