Too dangerous to release?!

Anthropic has released Claude Mythos Preview, a new AI model that won't be made publicly available due to its advanced security research capabilities. The model scored 77.8% on SWE-bench compared to Opus 4's 53.4%, and demonstrated the ability to autonomously find and exploit zero-day vulnerabilities across major operating systems and browsers, including a 27-year-old OpenBSD bug and a 16-year-old FFmpeg vulnerability. Anthropic plans to release safeguards with an upcoming Claude Opus model instead. The video takes a skeptical, humorous tone toward Anthropic's fear-based messaging around AI safety, drawing parallels to similar 'too dangerous to release' narratives around GPT-2, while also reflecting on how AI is gradually making traditional developer skills less essential.