Reactions from ranked influencers
3 posts
I disagree several of the underlying assumptions, including the notion that an aligned ASI needs to be a "super-nanny for humanity" as well with the notion that just because systems are super intelligent then they deserve welfare or should be considered as a "race." (But my sense is that this may become a deeper back and forth discussion than X is designed for..)
@boazbaraktcs @JaEsf I agree that there are far better outcomes that having ASI as a super-nanny, I just don't see them as reachable without either ignoring/overriding most people's values to get there, or somehow solving ethics. I'd love to see coherent visions that are none of those. But I haven't.
@boazbaraktcs @JaEsf I also don't know that AI becomes sentient or a moral patient - but I don't think that is a resolvable empirical question. And yes, I agree we're unlikely to resolve any of this on X.