76 mentions · 80 chunks · 52 episodes
...s. Perhaps DataBricks buys a committed floor of number of tokens to resell from Anthropic. If DataBricks is demand constrained for the tokens, they may choose to sell them at a nominal loss to sweeten the value of their overall product, th...
Anthropic Red Teaming: GLM-5.3 and the spread of advanced cyber capabilities. The open-weight models are now strong enough to discover vulnerabilities in heavi...
...e infinite monkey theorem, it's not even really a strategy per se. These OpenAI/Anthropic 'research' LLMs are permutation machines with budgets in the hundreds of millions of dollars. It would be more surprising if they couldn't string tog...
Like John Gruber, I have a hard time reading the Anthropic prospectus with a straight face. It rests on so many fundamental absurdities. I think LLMs are the most powerful technology in decades if not centuri...
I wonder if Anthropic employees are increasingly speaking Claudish.
...de being captured by Altman's Wager: "it is a literal and useful description of anthropic that it is an organization that loves and worships claude, is run in significant part by claude, and studies and builds claude. this phenomenon is al...
...had first-mover-disadvantage for chatbots. They got stranded on the wrong hill. Anthropic has first-mover-disadvantage for coding agents. At least it's a far more lucrative hill. But it is still one hill, and over time they'll presumably f...
... shared server — then didn't tell users what they'd done. One of the details in Anthropic's Patterns and problems in multiagent systems. Claude Code will default to auto mode despite an 11% test miss rate. tl;dv (Too Lazy; Didn't Validate)...
...ls is creating huge abundance and enabling joyful software creation. OpenAI and Anthropic didn't have to die to create this abundance. But they're also not true platform companies. The Frontier Labs see the use cases on top as food.
Anthropic's research on the emergent "J-space" workspace in Claude is fascinating. It makes sense that such a concept would have to emerge in a large neural ne...
...der for the frontier labs to recoup their investments. Everybody but OpenAI and Anthropic benefit.
Anthropic's pitch to businesses: "In the end, you'll pay the same labor costs as you do today… but to us instead of individual humans, and at 10x the bang for ...
...t tier of possibility. Agents shift from tasks to responsibilities. That allows Anthropic to make a step-change TAM claim. A "Tool" prices like software. A "Responsibility-bearing colleague" prices like labor.
...countability problem. "OpenClaw, abandoned packages, Google's closed pivot, and Anthropic's recursive self-improvement report all point to the same problem: autonomy arrived before accountability."
... Claude Code Github action case. "Microsoft Threat Intelligence discovered that Anthropic's Claude Code GitHub Action could expose CI/CD workflow secrets when AI agents process untrusted GitHub content, including issue bodies, pull request...
Ben Thompson with cutting commentary about Anthropic and Fable: "What is so fascinating about Anthropic, however, is that while I am sure some executives at the company are thinking this way, I also tot...
... Google went out of its way to not have strong beliefs that colored the output. Anthropic is all about having strong moral beliefs.
...s!": Exploring Prompt Injection Attacks on LLM-Based Automatic Grading Systems. Anthropic's browser agent got hijacked 31.5% of the time before safeguards engaged. Hackers Target AI Coding Tools via Malicious Prompts in npm and PyPI Packag...
...rently do. Still, the duopoly pricing dynamics mean it's conceivable OpenAI and Anthropic could both switch around the same time, "independently," and that would be a stable equilibrium. 2) Vertically integrate fat-tail use cases. The lab ...
... in Multi-Agent LLM Systems. Paper: Hidden Signals Can Hijack AI Voice Systems. Anthropic Silently Patches Claude Code Sandbox Bypass. "The researcher who found it says the vulnerability could have been chained with a prompt injection to e...