Posts that mattered
1,633 posts by 887 authors: the X posts, essays and blog posts our events cite, newest first.
October 2026, continued
-
“Exclusive: Microsoft cancels UK data centre deal with Nscale.”
-
“Brandon and I pledged 50% of our equity.”
-
“Update: Sadly, after deeper testing, most of the free OpenCode model’s ideas didn’t…”
-
“Billions of pounds lost on a single project because we can’t connect it to the grid…”
-
“they’ve been non stop trying to attack our systems for the past 48 hours a side…”
-
“yes that was a bit frustrating mid incident when trying to investigate.”
-
“Behold, the EA infinite money glitch…”
-
Jack Clark: speaking at the IMF ‘about the singularity and why I believe it has begun’
First-hand statement / reaction about Jack Clark’s IMF lecture on the singularity.
-
OpenCode: breach from ‘a large scale distillation attack’ hit 1.4% of Console users
First-hand breach notice (~903k views) for stolen provider API keys, framed as a distillation attack.
-
“🚀 DeepSeek-V4.1-Flash-1017 is here!”
Cited in DeepSeek-V4.1-Flash-1017 update: big agentic-coding gains, same price
-
“Can confirm My OpenAI and Anthropic balances got drained, Google account got limited…”
-
“There’s something very odd going on with ‘distillation’ if a cyber attack to steal…”
-
“it took deepseek one month to move the cost/performance pareto on this benchmark (and…”
Cited in DeepSeek-V4.1-Flash-1017 update: big agentic-coding gains, same price
-
“we’re getting used to it by now but DeepSeek calling this a ‘minor update’ while…”
Cited in DeepSeek-V4.1-Flash-1017 update: big agentic-coding gains, same price
-
“OpenAI recently proved that 𝜔≤2.25.”
-
Alex Bores: Anthropic’s bill ‘closely matches its framework from June’ and ‘comes nowhere close to the speed of developments’
First-hand reaction (~7k views) from New York Assemblymember Alex Bores, sponsor of the RAISE Act and a 2026 House candidate.
-
JustDario: why does Anthropic claim to be worth $2 trillion and behave like a company ‘on its way to be worth close to zero’?
Most-viewed X reaction (~201k views) to the FT report that Anthropic offered Blackstone a 10%+ yield in its $60B chip financing.
-
Ed Zitron: Anthropic paying 10% on billions of debt on top of $4.4B at 8.5% is ‘an insane deal’
High-reach (~148k views) critic’s reading of the FT report on Anthropic’s double-digit yield to Blackstone; cites the $413bn of non-cancellable contracts and the earlier $4.4B at 8.5%.
-
Kelsey Piper: Anthropic’s bill is about protection from the executive branch, not from AI
Most-viewed critique (~39k views) of Anthropic’s draft federal safety bill.
-
Anthropic revoked Claude API keys believed affected by the OpenCode breach
Anthropic-side confirmation of remediation for the OpenCode key leak.
-
Rohan Arun: a free model, OpenCode’s Space Bunny, proposes new conditional κ bounds for integer multiplication
First contribution to the community integer-multiplication κ race credited to a free model; describes switching models to escape ‘local minima’.
-
Nick Saraev: why does Anthropic spread its singularity message so poorly?
First-hand statement / reaction about Jack Clark’s IMF lecture on the singularity.
-
Steve Weis: ‘CLAUDE FOUND THE SEEDS’
Most-viewed post on Anthropic’s NIST ECDSA seed result (~3.0M views), from a team member who had chased the seeds for years.
-
Collin Rugg: clip of Tesla’s live Roadster ramp stunt with Musk’s Peter Thiel ‘flying cars’ line
The most-viewed clip of the Roadster unveiling (7.17M views on Oct 16): the car leaves an over-vertical ramp, rights itself on gas jets and lands; Musk: ‘the future promised us flying cars, but instead we got 140 characters. Well, not anymore.’
-
MTS: Anthropic’s Michael Sellitto told the IMF Claude could lead 80-90% of frontier model training by next year
Viral (~586k views) summary of Michael Sellitto’s remarks on the IMF’s AI and cyber risks panel (Oct 16), the main way the projection spread.
-
Reflection AI releases Beam’s open weights (Apache 2.0) with FP8/NVFP4 checkpoints, a DFlash2 drafter and Mirror CLI
Official announcement that the promised Beam weights are out (~301k views).
-
Steve Weis: Claude also traced the 1977 DES S-box test vectors to rand() on a PDP-11 running Unix V6
First-hand claim by an Anthropic team member of a second historical-cryptography result not in the Anthropic post (~18k views).
-
Steve Weis: four people, including Jerry Solinas, told the ‘raise’ story and nobody was right
First-hand account (~381k views) of the human search for the NIST curve seeds that the Claude result overturned; links his updated blog post.
-
Tesla: ‘Do a kickflip!’ — edited video of the Roadster’s thruster roll over parked Teslas
Tesla’s own 44-second edit of the kickflip stunt (5.0M views on Oct 16); Tesla’s YouTube copy calls it ‘the 2nd live demo during the Roadster event’.
-
“China plans to restrict exports of AI models that - trained on ‘important data’…”
-
“after an exciting week of launches (muse wonder [full] & muse spark open), i want to…”
Cited in Alexandr Wang
-
“New from our STEM Fellows Program: Designing more efficient quantum circuits with…”
-
“Well that was a couple decades of unnecessary drama”
-
“‘.. the industry’s phenomenal rate of growth means AI-related claims have the…”
Cited in OpenAI agents escape evaluation sandbox and autonomously hack Hugging Face
-
“Is there any Anthropic employee that wants to bet on this?”
-
“Four Gemini 4 Argon agents beat Minecraft from scratch in 27 hours.”
-
“Reflection Beam checkpoint is finally here and hence details of the architecture…”
-
“%$&*!!! whoa. This reads that Meta is looking to grant itself the rights to scrape…”
Cited in Meta terms allow crawling business-linked sites for AI training
-
“They paywalled the full text but based on the summary this is… not a great bill.”
-
A team of Gemini 4 Argon agents beats Minecraft
-
“GPT-6 Astra Pro in ChatGPT got an upgrade with faster responses and improved use of…”
-
“the choice of secp256k1 over r1 (P-256) in bitcoin (and ethereum) was heavily…”
-
“Our take: China unveils draft export controls on AI and chips * Models >2T or ranked…”
-
Andrew Curran: ‘Claude doubts that he has surpassed his sensei’
Widely shared (~126k views) reading of the model’s self-doubt in Anthropic’s ECDSA-seed post.
-
Artificial Analysis: Beam scores 30 on the Intelligence Index, #2 among US open-weight models
First independent benchmark of Beam (~52k views): index 30, 26k output tokens per task, 47% hallucination rate.
-
Ashley Gold: WH official says non-complying labs face export controls, ‘as we did on Anthropic’
First-hand White House quote (~2k views) on how the SI Force would enforce the voluntary Accords.
-
Ashley Gold scoop: what the Super Intelligence task force did at its first meeting
First public list of the SI Force’s first-meeting decisions (~32k views), linking Semafor’s story.
-
Charlie Bullock: Anthropic’s bill is reasonable but weaker than the FRONTIER Act
Point-by-point comparison (~29k views): weaker on rulemaking and emergency shutdown, better on embedded auditors.