Emad Mostaque

I think that you can align the inputs.

You raise it, right?

Well, again, this concept of feelings, right?

And in Asimov's books, you had the zeroth law that kind of was added above that.

So this is what Anthropic is trying to- What is it?

I don't know that one.

The zeroth law kind of supersedes all laws if kind of the whole system is at risk effectively.

But I mean, this is what Anthropic is trying to do with the constitutional AI process.

So you have the base model and they have a constitution that the AI adheres to that tunes it constantly.

So a series of kind of constitutional principles.

again is it is it as open for interpretation as a real constitution no one knows what the right constitution the right laws are this is the thing like our intellect only goes so far and we've already seen with laws and constitutions you can make those go anyway like north korea has a fantastic constitution does it really it does yeah it's actually pretty quite liberal

Well, it's their interpretation.

Interesting.

This thing, like, you have to adhere to it because the AI, what are feelings?

What is objective function?

Like, one of the key concerns in alignment is paper clipping.

You tell the AI to make a paper clip, and it's like, oh, well, let's just make the whole world a paper clip.

You're like, how do you solve climate change?

Just kill all the humans.

And it doesn't have any, like, feelings about that.