This is a translation of one of my mini-essays over at Fragmente.
Our problem is not that AI will develop "consciousness" and then destroy us all, Skynet-style.
Because, first, an AI without consciousness could do that too - a sophisticated intelligence does not necessarily have to be conscious.
Second, the evolution of a digital consciousness as we usually think of it under physicalist assumptions is not possible: evolution does not create consciousness; if anything, it evolves it, refines it, organizes it. But this happens in the context of biology and communication, not technology. At most, it seems to me, a sufficiently complex networked system with appropriate AI software could attract, connect to, some form of consciousness: a kind of demonic possession. Argh!
But this too, if possible, could just as well be interpreted as a symptom of the real problem instead of the cause: the self-enslavement of mankind.
Translations of my German mini-essays are for my paid subscribers. To continue reading and support my work, become one!