So, wait, the whole “Mythos AI is so powerful it can find exploits in any software” thing requires both access to the source code and thousands of runs to find anything remotely actionable? This is the “too dangerous to release” model they’ve been hyping up?

Is that really it?

@baldur Yeap. Also it found the same bug 250 times, not 250 separate bugs. And then one of the bugs required them to recompile Firefox to turn off some of the security features like its whole sandboxing to get to something exploitable. The OpenBSD bug would’ve been found by a fuzzer if anyone had actually bothered etc.
@ohno @baldur so it really is Anthropic Lie ™ ft. Bullshit ®