As large language models grow more capable, Chalmers' hard problem of consciousness becomes less abstract and more urgent.
If reality is indistinguishable from a perfect simulation, does the distinction between real and simulated collapse entirely?