$100 Million. The Most Advanced AI Model Ever Built. Just Marketing Hype.

Anthropic spent $100 million on an AI model too dangerous for public release. Then they pointed it at curl. The maintainer published his verdict.

$100 Million. The Most Advanced AI Model Ever Built. Just Marketing Hype.
Photo by Oleg Laptev on Unsplash

Five claimed vulnerabilities. One confirmed. Severity: low.

Anthropic built an AI model that they said was too dangerous for public access. They called it Mythos. Twelve launch partners put up $100 million in model credits. Amazon, Apple, Google, Microsoft, NVIDIA, and the Linux Foundation.

Bloomberg ran features. Fortune ran follow-ups. The operation had a name that sounded like a DARPA program: Project Glasswing.

Then someone pointed Mythos at curl. 178,000 lines of C code deployed on billions of devices. One maintainer.

Daniel Stenberg, the sole maintainer of curl, got the security report. Five claimed vulnerabilities from the most expensive AI security operation ever assembled.

And his verdict arrived...

This post is for paying subscribers only

Already have an account? Sign in.

Subscribe to Can Artuc

Don’t miss out on the latest issues. Sign up now to get access to the library of members-only issues.
jamie@example.com
Subscribe