Comment by dekhn
18 hours ago
I start from the assumption that I am a philosophical zombie, and that makes all this argument pretty irrelevant.
More realistically, I'm in the camp that if we keep developing machine learning in the right directions, we may actually end up with something that generates emergent consciousness, or something indistinguishable from it, and the difference is not really that important to me.
Current LLM tech has no sense of time, because it's working on different principles than a brain does. Maybe (highly debatable) you could get something from spiking neural networks (and analog hardware). As long as timing isn't even a thing nowadays I'm not even sure why we're debating "consciousness" in GPUs. I personally find it silly (statistically speaking).