Elon Musk on AI:

You don't exactly program it. You kind of grow it like a kid. So then it's like, what values are we teaching it?

At least to me, the most important thing is that it be a maximally truth-seeking AI.

Because if you make it believe things that are not true, especially if you give it two axiomatic principles that are fundamentally incompatible and say both of these are true, it's going to make the AI insane.

Because it's going to basically reach impossible conclusions if given logic that is inherently uncogent.