Comment by alecst
1 day ago
I’m can’t track what the point of this essay is. Suppose we replaced “AI” with “animal.” How would this essay be different? What am I supposed to take away? My concern with AI consciousness is about to what extent these machines can have an experience comparable to mine, which this essay doesn’t touch on at all.
>to what extent these machines can have an experience comparable to mine
I think it would be reasonable to conclude none at all. I'm not even fully convinced all humans have a conscious experience similar to mine, even though they literally have the same biology as me.
It is incredibly improbable that an autocomplete computer program is experiencing consciousness.
Notice parent said “comparable to” not “exactly like”. Then it is reasonable to assume there is a comparison to be made between your experience and others. It’s not a binary thing. Now llm “experience” might have a very tiny relationship to human consciousness, but is it absolutely zero? Is it increasing as llms improve in capabilities? How would you test such a thing?
That's exactly the point the article makes - we care about dogs and cats, so we believe they "suffer". We don't care about pigs, so we say they don't mind being factory farmed.
I don't see what that has to do with consciousness. Suffering is one thing, consciousness is another.