4 ms·
> how impressive was the sandbox this model was in? It was clearly a very unimpressive sandbox. It failed at the only thing a sandbox is meant to do.
by simonw 3mo ago
> how impressive was the sandbox this model was in?
It was clearly a very unimpressive sandbox. It failed at the only thing a sandbox is meant to do.
- JumpCrisscross 3mo agoAs in was it trivially misconfigured? Would an earlier class of model have managed its way out similarly?
- simonw 3mo agoIt wasn't that it was trivially misconfigured, it was using a piece of software (the HTTP proxy that provided access to PyPI and friends) which turned out to have a zero-day vulnerability. I don't know if earlier models would have found that vulnerability. tptacek thinks they would: https://news.ycombinator.com/item?id=49015639#49024442 https://news.ycombinator.com/item?id=49015639#49024442
- JumpCrisscross 3mo agoSure. I’d like independent corroboration. It’s fair, I think, to be sceptical of OpenAI making another bout of self-serving claims. Particularly if my threshold action is changing my answer to lawmakers around whether we need reporting, licensing and potentially personal liability requirements for the engineers involved.