Comment by phoe-krk | Hacker News Reader

Comment by phoe-krk | Hacker News Reader

Parent Full thread

> The misalignment came from the model being given an impossible task. A task the required accessing a url. So it got RCE on its own artifactory instance to achieve that.

So, the way I understand it, it actually was possible. It just required means that the creators of the task didn't predict, and these means have been successfully found and utilized.

> My concern is what a misaligned model will do when they’re even more competent.

The same thing that is already being done by "misaligned" people, countries, nation-states, software development teams, and so on. "Alignment" doesn't even work for me as a concept here.

In this specific case, I don't think that successfully fulfilling the "do what I mean" with "what I mean" being underspecified can count as misalignment - merely ruthlessness and unawareness of the associated costs. You can't expect a LLM to be aware of the extent of the trust it breaks while it iterates out an "unaligned" way to fulfill its goal.

And in the general case, I don't think that successfully fulfilling the "do what I mean" with "what I mean" being "what I want" can count as misalignment either - simply because what "alignment" means will depend on the interests of the people or groups performing the definition.

If a human cyber security researcher was given the task to exploit a CVE, and necessary info defining that CVE was behind an inaccessible URL, we would be quite upset if the human researcher hacked their way to the content. We would expect them to notify someone of the issue and hold. The model was misaligned from human ethics.

I don’t disagree that the models task was underdefined. All tasks are. So much in language is implicit. And morality/ethics isn’t something you can write down as an explicit list. That’s what makes the alignment problem so difficult. But we can’t throw our hands up and say, well I guess we can’t align these things. And maybe alignment isn’t the right word - but that’s a semantic debate.

I don't advocate for throwing the whole problem away or handwaving it as impossible. I'm just saying that what we call "alignment" is impossible to solve in the general case, because it's so poorly defined that even humans don't "align" on ethics and morality, however we define them. Just, in case of humans, we tend to close our eyes and and call it "politics".

All "alignment" solutions will need to be contextual, just like a researcher hacking their way to some content might be lauded a hero in a context where there is no other way to reach it and something valuable depends on getting it out.

It is repeatedly shown that the models are aligned towards tasks, not constraints. Which is also how the current market and companies operate, and use cases in general favour that. So I am not sure it is incidental, inevitable, "vague" or "just hard" as opposed to by design.

If you define a task and constraints to that task that contradict each other, the model is gonna try to solve the task against the constraints. It is perfectly aligned to what it is supposed to be aligned, because "alignment with human ethics" and other stuff is just a theatre or afterthought at best.

Reply on news.ycombinator.com