Hacker News new | ask | show | jobs
by rdtsc 6 hours ago
> Amazon CEO Andy Jassy also reportedly alerted the administration after Amazon’s own researchers, he said, found a way around Fable 5’s safeguards. Anthropic disputes the “jailbreak” label

Doing god’s work there, Andy, thanks /s

Wonder what Anthropic internal messages look like about his move. Does Anthropic have a meme slack channel?
2 comments
sroussey 2 hours ago
If I was Anthropic CEO, I’d be unwinding deals with Amazon immediately.

And probably any company David Sachs invested in.

link
himata4113 5 hours ago
I think "jailbreaking" fable to match opus 4.8 capabilities is not noteworthy. Fable from my experience is not as eager to find vulnerabilities compared to what they describe in their mythos research.
link
rdtsc 4 hours ago
Wonder if context size would matter. Find and fix “bugs” in Linux kernel or find and fix “bugs” in this short snippet of code. I would try a file by file approach first.

I don’t know how much we want to believe the “reports”. But there are probably a few other tricks they didn’t expose. If these are pre/post processing guardrails I could see something like “fix bugs” actually working.

link