Fable vs mythos...does benchmark result even mean anything?
https://www.nytimes.com/2026/06/09/technology/anthropic-ai-claude-fable-mythos.html?smid=nytcore-android-share
Nope. #MLsec reality is more complicated than a benchmark
https://berryvilleiml.com/docs/no-security-meter-ai.pdf
Called Claude Fable 5, it is twice as expensive as the company’s previous flagship system.
Gary McGraw