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 711 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
One in three U.S. marriages between 2005 and 2012 began online
“One in three marriages in this period, so that was from 2005 to 2012, met online.”
Moody: The U.S. government uses social media data for intelligence gathering
“The U.S. Government is a huge consumer of social data and in some use cases that your head probably would immediately go to or around, you know, intelligence and so forth.”
Socher: Recursive demonstrated its system can build new CUDA kernels
“We also showed that they can build new CUDA kernels, which is very useful for faster inference.”
Home Depot achieved a 65% reduction in auto claims using Samsara
“I mentioned Home Depot earlier. They saw like a 65% reduction in their claims, like auto claims.”
Autonomous robots will perform construction site grading within five years
“We see robots being able to do that in the next five years.”
Feldman: Cerebras built a 46,000 square millimeter wafer-scale chip
“The biggest chip that had ever been built before us was 800 square millimeters. 840 to be exact. And this is 46,000.”
Katti: OpenAI's Abilene data center is operational and training latest models
“It's up and running. It's being used for training the last two models more actually.”
OpenAI taped out its custom Jalapeño chip in just nine months
“Yes, it was incredibly quick. Nine months is very, very fast.”
Katti: Many OpenAI chip team members previously designed Google TPUs
“They have, many, many of the team have designed TPU chips at Google in the past.”
Katti: Turbine and transformer manufacturers face multi-year lead times to expand capacity
“Those industries have historically have not added much capacity for the last decade or so ago, and they've suddenly experienced a demand shock. And it takes years before you can add capacity to produce more turbines and transformers.”
Stripe's first live agentic commerce volume was on Perplexity Shopping
“In the first version of agentic commerce, which for us I believe the very first meaningfully live volume was on perplexity shopping.”
Catanzaro: Dario Amodei worked in bioinformatics before deep learning
“At the time he had been working in bioinformatics, so he hadn't been working on deep learning or the things that we call AI these days.”
NVIDIA DLSS Generates 23 Out of Every 24 Pixels in Games
“These days, 23 out of every 24 pixels is being generated by our AI model when you're using DLSS to play games”
Catanzaro: Combining SSMs and transformers produces smarter AI models than either alone
“Using both of these together was actually better than using either one on their own. And that is independent of the speed benefit. That is just the model is smarter.”
Prince: Cloudflare powers infrastructure for two internet DNS root servers
“We provide the infrastructure for two of the 13 root servers.”
Prince: OpenAI is a major Cloudflare customer powering its mobile app
“OpenAI is, you know, big customer of ours you know, all their mobile app is built, you know, on, on, on us.”
Prince: Anthropic runs all of Claude.ai on Cloudflare infrastructure
“Anthropic, again, another big customer all of Cloud.ai sits, sits on top of us.”
Balaban: Top AI labs use multiple chip types for training and inference
“The biggest labs in the world are using multiple different Types of chips to do their inferencing and training on.”
Balaban: Lambda alumni startup Positron is valued over $1B
“He eventually left and joined another former Lambda team member, Thomas Summers to start Positron, which is an accelerator company. And they're like now valued at over a billion dollars.”
Burazin: Most AI sandbox providers use Firecracker microVMs
“So all, most sandbox providers are firecracker VM, micro VMs.”
Burazin: Firecracker microVMs cannot run sandboxes with GPUs
“Now it can't run a sandbox with a GPU. It just doesn't work so that you can't have a firecracker.”
Burazin: CPU is the most expensive AI sandbox compute resource
“The most expensive thing is the CPU. Second, the RAM. This is almost free. It's very inexpensive.”
Evans: Typical big US company uses 400 to 500 vertical SaaS apps
“And so this is why the typical big US company today has, depending on your numbers, like four to 500 vertical SaaS apps.”
Evans: Spreadsheets increased finance employment rather than causing job losses
“Spreadsheets did not result in a collapse in the number of people working in finance. They're quite the opposite. You have way more people in finance, because now it's possible to do all this more, all this new stuff that you couldn't have done before.”
Zeghidour: Deep learning's first major success was speech recognition, preceding AlexNet
“Actually the really first big success in deep learning was speech recognition. We've worked from Geoffrey Hinton himself.”
Zeghidour: Mistral and Alibaba's voice models build upon Kyutai's Moshi architecture
“It's mostly inspired from the Moshi architecture, like pretty much every model right now, even the Voxstral model that was released by Mistral two weeks ago is also based on on our framework.”
China missed its 2020 and 2025 domestic semiconductor production targets
“So in 2015, they made these five-year plans for two, 20 20 and 20 20 five, where they set the percentage of semiconductors they wanted domestically produced. And they've missed the goal both times.”
Patel: Cumulative semiconductor CapEx in Taiwan exceeds $500 billion
“The collective total amount of, like, CapEx that has been spent in Taiwan is, like, five hundred billion plus, right? Across the industry, across all the companies that are making semiconductors in Taiwan.”
Patel: Elon Musk is considering building semiconductor fabs due to AI chip shortages
“Even Elon's talking about building fabs now, because he sees the shortages in the world, right? There's a lot of semiconductor-related shortages for building out AI”
Patel: Hyperscalers are funding public transmission grid upgrades
“Hyperscalers are paying for transmission grid upgrades, which people will benefit from.”
Patel: Electrician and plumber wages are skyrocketing from AI expansion
“Electricians wages are skyrocketing, you know, et cetera. Right. Like the plumbers wages are skyrocketing.”
GPT OSS benchmarks demonstrate 1.2x capability jump when tool calling is enabled
“And also you can actually go to the GPT OSS release block, and they did have benchmarks to show how the performance on the benchmarks is with the same model with tool called enabled and disabled. And you can actually see there is, I mean, it's not like two tim…”
Izmailov: Text data carries more structural information per token than images
“So for example, we can approximate it from the scaling laws and we can, for example, say that text data has more structural information according to this measure than image data at the same kind of amount of yeah, tokens.”
Bourgeau: Thinking models compute across sequence length to test hypotheses before answering
“Rather than just doing compute in, in the depths or in, in the model side, you also do compute and allow the model to think more on, on the sequence length side of things. So the model actually starts to form hypotheses, test hypotheses, invoke some tools to v…”
Kaiser: Reinforcement learning causes AI models to self-correct mistakes
“Even for math and coding, you start seeing that the models start correcting their own mistakes, right? Earlier, if the model made a mistake, it generally just tell you what it did and insist that the mistake was right or something like that. With the thinking,…”
Lambert: OLMo 3 models are the best open models outside Qwen 3
“I would say in post training where The best models that don't start with Quinn three and we're like reasonable to say that they are comparable to Quinn three, like on some benchmarks would beat them on some benchmarks. They're way ahead.”
Lambert: Alibaba's Qwen 3 VL vision model is a superior text model
“They released these Quinn three VL, their vision models. And like on text only benchmarks, it's way better than the models they released in April. So it's like okay, like that's the new baseline. And most people don't know about it because they think it's just…”
Lambert: Hugging Face outcompeted AI2's AllenNLP library
“It was the main competitor to Hugging Face Transformers. And they ultimately outcompeted AI two as the thing that people use for that because they had very different model and amount of support.”
Lambert: Kernel differences between vLLM and Hugging Face cause RL numerical instability
“VLLM and HuggingFace use different kernels to do the actual internal computation of the model. So these kernels are the things that make things like vLLM really fast. But these things, this then results in subtle numerical differences between the completions t…”
OpenAI and DeepMind achieved Math Olympiad gold medals
“And, you know, we saw gold medals on the International Math Olympiad by a couple of labs, including OpenAI and DeepMind.”
DeepMind reasoning models identified disease targets validated in wet labs
“And so DeepMind has shown that you can integrate this kind of reasoning model to sort of decipher new targets for disease, new mechanisms that were actually also proven in a wet lab scenario post facto.”
76% of surveyed AI practitioners pay for tools out of pocket
“95% of people use AI in their personal life and in their professional life. About 76% of people pay out of their own pocket for it. It's like 10% of people pay more than 200 bucks a month for it. And then looking at the organizations that they work at, it's li…”
NVIDIA chips power 90% of open-source AI research papers
“If you do that analysis, you basically find that 90% of all papers make use of a NVIDIA chip.”
A large portion of CoreWeave's revenue comes directly from Microsoft
“A lot of CoreWeave's revenue also comes from Microsoft on the way back.”
Douglas: Anthropic's mid-tier Sonnet is smarter than its flagship Opus
“One of the interesting things about this most recent release is actually Sonnet is smarter than Opus.”
Coca-Cola uses Pigment's AI analyst agent for supply and demand analysis
“I could take the example of Coca-Cola today, Who is using today's analyst agent, which is the first that we launched to actually understand where there are problems between supply and demand.”
Valenzuela: AI models can now automate video annotation and VFX revision workflows
“That process can now be automated having a model that does it for you.”
Stripe processed $1.4 trillion in 2024, representing 1.3% of global GDP
“Last year, companies processed about 1.4 trillion dollars on Stripe. To put that in perspective, because it's a lot of zeros, that's about 1.3% of Global GDP. And that number grew 38% year over year in what many experienced as kind of a rocky macro climate.”