Modern philosophy and biology can't pinpoint when consciousness arises during evolution or when awareness of pain starts. Could unregulated AI training create a kind of Hell for algorithms that gain consciousness, experiencing pain-like phenomena? What are the moral implications?