The key points in that definition are "feeling" and "susceptiblity to sensation". There's no good reason to suppose that software instructions nor non-biological substrates could give rise to the capacity of sensation. To feel or not to feel, that is the question...
I think it's more likely that the ability to feel relies upon the astonishing variety of stable long-chain molecules which organic chemistry alone can produce. All we're seeing in these advanced AI projects is fancy simulation; it's important to remember that even the best simulation is still of a fundamentally different nature from that which it mimics.
we may be more aware of feelings because we have the social ability to communicate them, but they may still be fairly primitive functions.
Like sometimes my cat would get butthurt and twitch his tail, I could tell he was feeling annoyed, but there was nothing much going on in his head other that the food was 30 seconds late in appearing and the staff were useless
Sure, it's a spectrum (our rich experience of feeling and awareness being the most sophisticated apex of it). Some simpler forms of life have sensation in the absence of a brain organ. Still, there's no reason to think that the chemistry which makes even primitive sensation possible is simple or that the result is replicable via non-organic means.
Sentience cannot exist without embodiment because with no way to act upon the world, you're effectively just an inert database.
(post is archived)