[gatonegro]

https://archive.ph/tt0pY

I keep seeing talk of artificial consciousness with nothing more than a vague outline of what said "consciousness" would entail, or a hand-wavy "we'll know it when we see it." This is, of course, not that surprising considering we don't have those answers for our own consciousness.

At this stage I'd say we're way more likely to end up developing a system that can trick us into believing it is conscious, à la LaMDA and Blake Lemoine, than an actual conscious entity (whatever that means, exactly). And, if that were to happen, I don't think we'd have a way of knowing that's the case. We could just carry on believing we created artificial consciousness because our not-at-all-conscious robots are too good at saying the right thing with the right prompt.

[matt-attack]

Well interestingly we do the same with every other human on the planet. I basically assume that because I know I’m conscious and you and everyone else are the same chemical design as me, and because you act the same as me, then I just assume you too are conscious. But I have know direct way of proving that. It’s 100% speculation.

It’s entirely possible that you are just acting like a conscious being would act but there is no actual subjective experience inside you.

[gatonegro]

Sure, that's a useful heuristic, but I'd say it works because we deal with other humans. It's safe to assume that, being members of our own species, they are likely to have an inner experience and consciousness too. How do you apply that to an entity that is completely different—chemically, mechanically, etc.—to you and any other conscious being you've ever met?