soiax

The title is bad, it is not sandbox escape, it is "RCE inside the sandbox", so only RCE when sandbox is disabled.

Why are people upvoting a unexploitable bug? How is this interesting? There are thounds of these, no one even reports them unless they are exploitable, DoS only.

Right, but we have no clue why, and how the emergent behavior they show works.

If we would know that, there would be no need for interpretability research.

They should also check for hot dog vs not hot dog

Yeah dude .. if you say it multiple times, and louder, it must be true. That would make it a nazi company, right.

There it is... the usual ELON MUSK BAD.

You mentioned prompt injection, now when you talk about larger time horizons, that sounds like a AI alignment issue.

I'm sure there will be actors who don't care at all about "security", saying the positive outcomes outweight the negatives.

Yeah that's false.

from: https://arcprize.org/blog/oai-o3-pub-breakthrough

"Note on "tuned": OpenAI shared they trained the o3 we tested on 75% of the Public Training set. They have not shared more details. We have not yet tested the ARC-untrained model to understand how much of the performance is due to ARC-AGI data."

This sound like you assume that the first thing someone thinks about is security, when building the next big thing.

They will just build something as fast as they can. Last thing you think about is "security".

There were prompt injections in all the big models, and still are. Why would it stop distruption?

You can disable it runtime, with --no-sandbox command line option.

No. There is a reason the author keeps repeating "arbitrary code execution in the Chrome renderer process." Because it's there, not in the browser process.

It's all renderer only RCE-s, no sandbox escape. So it doesn't work on your browser, only if you disable the sandbox.