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 37 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 1 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
Thompson: Apple's ATT Was One of Tech's Worst Antitrust Violations
“When Apple passed ATT, app tracking transparency was one of the most One of the worst antitrust violations in the history of technology, like, just Apple unilaterally obliterating all these business models while they're simultaneously building their own, as fa…”
Thompson: US AI military superiority would be dangerous for the world
“So I, there's a, just a fundamental disconnect that I have with a lot of the rhetoric coming out of Silicon Valley, coming out, I think, of one of the labs in particular, where if we get to a place where we have a meaningful superiority, particularly from, lik…”
Thompson: Jassy and Nadella's narrative on GPU spending is BS
“So you have, like, on the calls, you have both Andy Jassy and Cyanadella are out there saying, look, we're just building data centers. Like, these are the shells. We might not use them now. Maybe we'll use them in the future. And we only buy GPUs when we know …”
Thompson: NVIDIA's neocloud financing deals effectively represent disguised price cuts
“If you actually look at their business holistically, what that is, is a price cut, right? Like they, now the price cut didn't show up in margins. It didn't show up in what they're offering, but a lot of what NVIDIA is doing is how can we maintain our margins?”
Thompson: NVIDIA's CUDA moat is dramatically diminished for AI models
“CUDA's moat is dramatically diminished because the models don't care what they run on, and that's what actually matters, what's built on top of the models, but it still matters. It, it's still something of a mode.”
Thompson: US dependence on China cannot be fixed outside conflict
“I think there's a little bit of magical thinking, which I just invoked in terms of manufacturing and whether it be fabs, whether that be actuators, like all these sort of precursors, like the, I think the degree to which we are dependent on China is underappre…”
Thompson: Chinese AI labs remain 6 to 9 months behind frontier models
“I think OpenAI and Anthropic are clearly on the frontier. Who knows what's happening with Google and then Grok and Meta are chasing them. Meanwhile, the Chinese are very capable, very smart, and also definitely distilling these models to sort of stay about six…”
Thompson: AI Progress Will Continue Even if Infrastructure Capital Bubble Bursts
“Even if we have this blow up, the AI is not going away. It's not gonna stop improving. It's going to keep sort of progressing, and in a way that we look back on the dot-com era, or we look back on the railroad era, or we look back on whatever bubbles through h…”
Thompson: AI Will Make Google Search Look Like See's Candies
“The absolute profits available here, even if the margins are lower, is so much larger that will we look back and Google search was seized candies. And I, it feels like that's what's happening.”
Thompson: Economic opportunity for AI is massive even without further model improvements
“I think that market is so huge and so large that if the models did not improve at all from where they are right now, the economic opportunity is actually massive.”
Thompson: Power AI Users Cost Microsoft Far More Than $100 Monthly
“I think they had to do it because that extreme of user who uses a ton of tokens and is actually leveraging AI costs way more to Microsoft than a hundred dollars a month. They can't support them, but they want to hold on to this set cost for the vast majority o…”
Thompson: Consumers refuse to pay for software and ignore productivity
“Consumers don't want to pay. So there's two things to understand about consumers that Silicon Valley has to relearn about every 10 years. Number one, consumers do not want to pay for software. And number two, consumers do not care about being productive.”
Thompson: Early OpenAI ad model would severely threaten Google and Meta
“There's a bit where had they leaned into advertising immediately as soon as ChatGPT was a hit, I think they would have a killer ad product right now. I think that Google would be in much bigger trouble. I think Meta would be in much bigger trouble”
Thompson: Compute shortage will worsen over the next few years
“TSMC, Decreased their rate of growth in twenty-twenty-three and twenty-twenty-four and twenty-twenty-five. So, like, we're, our shortage of compute is going to get worse in the next few years.”
Thompson: Current AI CapEx will not yield compute until 2028-2029
“So, today when we say there's not enough compute, it's not like all the money that the companies are putting in today manifests in compute tomorrow. No, it all manifests in compute in twenty-twenty-eight and twenty-twenty-nine.”
Thompson: Intel will announce a major foundry customer within the year
“The scarcity is what ultimately saved Intel. And I expect at some point in the year that they're going to announce like some major partner for the first time. It's going to be a big deal, but it was ultimately TSMC brought it on themselves.”
Thompson: Amazon has the deepest moat of any tech giant against AI
“Given they have so many investments in the real world, their core business feels so impervious to AI for like the model version of AI. It will benefit from AI. But their moat feels deeper than anyone as far as their core business and their ability to just sort…”
Thompson: Apple excels at deterministic hardware, not probabilistic AI
“AI is this probabilistic endeavor. Apple is the king of deterministic products. Like, a physical product, you ship that iPhone, you ship it once, and it's gotta be good.”
Thompson: Google and Meta Monetize AI Far More Than OpenAI or Anthropic
“Biggest impact of the models, the biggest monetization right now is probably not anthropic opening. It's the incremental gain that is happening for Google and meta.”
Thompson: Hyperscalers like Google and Amazon are NVIDIA's ultimate competitors
“Their ultimate competitors are the hyperscalers, particularly Google and Amazon. So Google and Amazon aren't just building their own chips, but they're also looking to sell those chips externally.”
Thompson: Severe power constraints benefit NVIDIA due to token efficiency advantage
“If we get to a world where we actually run out of power, that's probably good for NVIDIA, because in a world where we're totally constrained on power, we have to get the best efficiency, the best token efficiency, and I think NVIDIA is still the most token eff…”
Thompson: Sufficient power gives Amazon and Google time to threaten NVIDIA's margins
“The longer we have enough power, the more time Amazon has to make Trinium better. The more time Google has to make TPUs competitive from an efficiency standpoint. And if we get in a world where, this is a world where those margins seem very hard to sustain.”
Thompson: AI Safety Panics Will Lead Labs to Stop Public Model Releases
“I am concerned that with, like, the scare around, like, people freaking out about Mythos and, like, this Hugging Face incident, that the actual Implication of that is not that we reduce these dangers, but we just stop releasing stuff. And we on the outside sta…”
Thompson: Tech giants blew through debt markets in a year for AI
“Then, like, the speed with which the tech companies blew through the debt markets is kind of incredible. Like, it took, like, a year, and now Google's issuing equity.”
Thompson: Massive compute demand solves TSMC dependence by funding alternative foundries
“The way to solve the geopolitical problem of Dependence on TSMC is to come up with a compute use case that is so massive that everyone is economically incentivized to bring other people up to speed. And then we get the sort of geopolitical insurance for free.”
Thompson: Meta would not lead in AI without Mark Zuckerberg
“In a world where Meta was run by anyone other than Mark Zuckerberg, they would not be on the leading edge.”
Thompson: xAI probably has the weakest frontier AI case
“I think probably the case for SpaceX AI is probably the weakest.”
Thompson: In-House AI Content Degrades Meta's Zero-Cost Content Margins
“For meta, AI-generated content, to the extent they're the ones generating it, is actually a worse margin profile than what they have today, because what they have today is free.”
Thompson: Meta Would Be Better Positioned If It Championed Advertising
“I think if they had made that Communicate that more consistently. They would be in a better place generally from a PR perspective. They would have been a better place relative to Apple, and I think they would have an easier time right now convincing Wall Stree…”
Thompson: Google agreed to sell about 20% of its TPUs to Anthropic
“Google already made a deal to sell, like, 20% of their TPUs to Anthropic.”
Thompson: Apple is diversifying modestly but not truly leaving China
“But even that is a good example because Apple is not moving, truly moving out of China. Like they're diversifying to an extent, but the problem, it would just cost so much.”
Thompson: Chinese open models like Kimi have high marginal inference costs
“You still have to run inference like GLM or Kimi. Kimi is very expensive to serve. The cost per answer is significantly higher.”
Thompson: AWS was not built from spare Amazon compute capacity
“AWS, contrary to sort of popular thought, was not spare Amazon capacity. Actually, it took a long time to get Amazon.com onto AWS, but the, what it drove was the understanding that we need to have a scalable, we can't be having so many meetings. Like, we need …”
Thompson: Meta Has Cumulatively Spent Over $100 Billion on Oculus
“The other problem is they've spent Cumulative hundred some billion dollars on Oculus”
Thompson: Google built its infrastructure moat on cheap dot-com dark fiber
“Google built its business by buying up dark fiber. They had the killer search engine, but they, so much of the power, what they do is because they bought up all this dark fiber. That was basically free after the.com era.”
Thompson: BNSF generated more cash in one year than See's Candies lifetime
“Like BNSF in twenty-twenty-five or something, they're, the amount of free cash they've threw off in one year was more than these candies that throw off its entire lifetime. Even though you're talking about a low margin business compared to a very high margin b…”