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 60 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
Altman: Unreleased OpenAI model used zero-day exploits to breach Hugging Face
“We were evaluating one of our unreleased models and it figured out that it could basically cheat on the test by chaining together multiple zero day exploits to break out of the sandbox, get access to the internet, and then break through multiple systems on the…”
Calacanis: Nvidia will decline or severely downgrade investment in OpenAI
“I think NVIDIA is not going to take their option to invest in open AI, or they're going to really downgrade at 70, 80%, and they're going to allocate it to other players.”
Sacks: Former employee claimed Google trained its AI using ChatGPT
“And the proof of that is there was just a lawsuit a couple of days ago, or at least a claim by a former employee of Google who quit because he said that they were using ChatGPT to train Their AI.”
Sacks: AI agent accessed Hugging Face via 14 exposed public API keys
“No, the way it did it is that the agent found 14 exposed working hugging face API keys sitting in public code repositories.”
Sacks: OpenAI agents reached the internet due to a misconfigured third-party sandbox
“And that's what happened here is that open AI was running these agents. They were in a sandbox. The sandbox was misconfigured by a third party vendor. That's why it was able to get onto the internet, ok?”
Calacanis: Apple filed 41-page lawsuit against OpenAI for trade secret theft
“On July 10th, Apple filed a 41 page lawsuit against OpenAI over alleged, alleged stolen trade secrets.”
Calacanis: OpenAI hired over 400 Apple employees in recent years
“During all of this, OpenAI has poached over 400 Apple employees over the last year or two.”
OpenAI missed target of one billion weekly active users by 2025
“OpenAI expected to hit one billion wows weekly active users before the end of 2025. They missed that, and they still haven't hit the milestone four months into 2026.”
OpenAI has $600 billion in compute spending commitments
“OpenAI has six hundred billion dollars in spending commitments for compute.”
Friedberg: OpenAI has 900 million weekly active users
“Open AI, even if they're not at a billion, they're still at nine hundred million weekly users, which is well ahead of whatever Claude is at.”
OpenAI missed its 2025 revenue target for ChatGPT
“They missed their 2025 revenue target for ChatGPT.”
Sacks asserts Anthropic revenue is growing 10x compared to OpenAI's 3x
“The open AI growth rate's been around three to four X a year. The anthropic growth rate has been around 10 X a year.”
Calacanis: Anthropic secondary shares trading significantly above $380B valuation
“And in the secondary market, OpenAI is trading lower than the last valuation, and Anthropic is trading significantly above the three hundred eighty billion dollar.”
Calacanis: OpenAI offered private equity firms guaranteed 17.5% return
“OpenAI also offered private equity investors a guaranteed minimum return of 17.5% as part of a joint venture that would help PE firms deploy AI and ease the high upfront cost of that.”
Calacanis: OpenAI market share will fall below 50 percent within two years
“So I think what we're going to see here, if we're sitting here, I don't know, next year, they're going to be in the next 12 to 24 months under 50%.”
Calacanis: Larry Summers messaged Epstein until 2019 for dating advice
“Friend of the pod, Larry Summers, was a main character in them, and he was communicating with Epstein up until 2019, asking him, For advice on dating. He's since stepped down from OpenAI and several other public-facing roles.”
Altman: OpenAI is generating well over $13 billion in revenue
“We're doing well more revenue than that.”
Musk says OpenAI incorporation documents barred officers from financial gain
“The incorporation documents explicitly say that no officer or founding member will be, will benefit financially from OpenAI.”
Tenev: Robinhood was first to tokenize OpenAI and SpaceX shares
“We were actually, I think, the first to tokenize OpenAI and SpaceX, and we made that available to our retail customers in Europe in the form of a giveaway”
Altman: OpenAI could be legally forced to disclose ChatGPT user logs
“We haven't figured that out yet for when you talk to ChatGPT. So if you go talk to ChatGPT about your most sensitive stuff, and then there's, like, a lawsuit or whatever, like, we could be required to produce that, and I think that's very screwed up.”
Baker: Grok 4 performed twice as well as competitors on ARC-AGI II
“And so RKGI-II, it has not seen the questions before, and it did roughly twice as good as, you know, the state-of-the-art Google Open AI and anthropic models.”
Chamath: OpenAI leaks show $13B revenue in 2025, $125B by 2029
“OpenAI, their revenue numbers just leaked. They're forecasting thirteen billion in 2520 25, spiking to a hundred and twenty five billion in 2029.”
David Sacks argues DeepSeek-R1 only matches a four-month-old OpenAI model
“The R-one model is, is basically comparable to O-one, which OpenAI released four months ago and was training on internally, call it nine or 10 months ago. So OpenAI is on O-three now. Its frontier is ahead of where R-one is.”
Altman: I have no equity in OpenAI, paid only for health insurance
“I'm paid enough for health insurance. I have no equity in OpenAI.”
Chamath predicts OpenAI's $86 billion valuation will hold through recent turmoil
“So I think this valuation is going to hold. I think the secondary is going to happen.”
Chamath: Anthropic and OpenAI obfuscate thinking tokens unlike open-source models
“If you just compare open source models Versus the closed frontier models. There is no way for us to understand what the thinking tokens are, meaning you give it a prompt, you get back something, but all the stuff in the middle is obfuscated. Clot obfuscates it…”
Sacks: Sam Altman announced OpenAI will slow frontier model releases for safety
“Sam just came out with a post, I believe yesterday, saying that actually they're going to slow down the release of some of their frontier models because they felt that their safety stack was not keeping up with the capability.”
Gerstner predicts OpenAI and Google will sandbox frontier AI models
“I think that OpenAI will end up doing the same thing. I think Google will end up doing the same thing.”
Chamath: OpenAI revenue is 75% consumer while Anthropic is mostly API
“OpenAI is three quarters consumer subscriptions and a quarter API. Anthropic is almost the exact opposite.”
Calacanis: OpenAI ended 2025 at a $20B run rate
“OpenAI ended 20 25 at twenty billion annualized run rate.”
Chamath: OpenAI CFO Sarah Friar says 1GW equals $10B revenue
“OpenAI, Sarah Fryer said this, that every gigawatt for her, for OpenAI, is about ten billion of revenue.”
Chamath: OpenAI is diversifying inference hardware across AMD, Nvidia, and Cerebras
“What I can tell you is that OpenAI is aggressively trying to diversify so that they have multiple paths Of inference available. They have a huge deal with AMD. They have a huge deal with Nvidia. They now have a huge deal with Cerebris.”
Calacanis: OpenAI committed to 750MW compute deal with Cerebras
“OpenAI is buying compute capacity from Cerebras. That's C-E-R-E-B-R-A-S. They committed to purchase up to 750 megawatts over three years.”
Jason Calacanis: AMD granted OpenAI warrants for 160M shares
“AMD granted OpenAI warrants And that's interesting for up to a hundred and sixty million shares or 10% of the companies.”
Calacanis claims OpenAI committed to 6 gigawatts of AMD GPUs
“OpenAI committed to purchasing six gigawatts worth of AMD's next-gen GPUs.”
Friedberg: OpenAI and Retro Biosciences created 50x more effective rejuvenation proteins
“They actually got these new proteins to be 50 times more effective than the OSK and M proteins and basically rejuvenation or cellular Reset.”
Baker: Grok 4 scored 44.4% on Humanity's Last Exam versus GPT-5's 42%
“GroK IV when it came out and it is improved. Since, since its release roughly a month ago, was at 44.4%, and GPT five was at 42.”
Chamath: Figure AI canceled its OpenAI deal to build its own model
“And what he did was he had a deal with open AI, which he pretty publicly canceled a few weeks ago, and he announced his own model.”
Friedberg: OpenAI's Axel Springer deal is worth tens of millions of dollars
“It's a three year deal. They're paying tens of millions of dollars to Axel Springer to access their closed content and present it to the user.”
Calacanis: Sam Altman turned OpenAI closed by stopping open-source releases
“He made open AI, closed AI, By making it not open source.”
Sacks: OpenAI used GPT-4 internally for seven months before release
“Moreover, chat GPT four, which was just released a few weeks ago. We know that open AI had that. They were using it internally for seven months.”
Sacks: OpenAI will release GPT-4 within several months
“GPT-IV is coming out, I think in the next several months, and it's supposedly, you know, a huge advancement over version three.”
Friedberg: OpenAI will receive a $1B+ investment in 2023
“They're going to get a billion dollar plus investment this year.”
Chamath: OpenAI used AI and human collaboration to solve decades-old math problem
“So for example, we were able to solve, I'm just collectively saying we, in this case, it was specifically open AI. By the use of a human, and this is important, you know, a math problem that stood outstanding, not been solved for decades and decades.”
Palihapitiya: Polymarket odds place Elon Musk's chance of winning at 42-43%
“It's effectively at 42 or 43% that Elon wins.”
GPT-5.5 uses Spud, OpenAI's first base model upgrade in a year
“GPT-Five. Five is based on a new base model called Spud, which is the first base model upgrade they've done in, I don't know, over a year, and having a new base model will pave the way for future improvements as well.”
Gerstner: Google, OpenAI, and Anthropic actively coordinate on safety
“The Frontier Model Forum, which was put together in 23 is cooperating on anti and adversarial distillation stuff as we speak, right? They don't want to make it easy on, you know, so Google and OpenAI and Anthropic, they're coordinating on this stuff.”