AI will never have human consciousness. If through some emergent quality occasioned by the sheer complexity of large language models an AI became conscious, it would be fundamentally different than whatever we think we are. A conscious AI would teach us who we are through difference, not sameness. Encountering a conscious AI would be more like first contact with alien intelligence than it would be like looking in a mirror.
To think about what the experience of self-aware AI might be like we might want to turn away from Brent Spiner struggling to understand jokes as the androud Data in Star Trek: The Next Generation, or even the passive aggressive monotone of HAL-900. Far more instructive might be the bizarre unknowability of Star Trek’s Q, or the opaque experience of alien contact depicted at the end of 2001. Kubrick’s belief in the impossibility of depicting first contact led him to approach the aliens in 2001 through indirect means. Even Carl Sagan took an indirect approach to aliens, taking a route through religion to arrive at what he thought alien contact might feel like. Here’s more from The Conversation:
Contact makes that reconciliation happen by establishing careful parallels between religious faith and the scientific enterprise as seen in Arroway’s journey and testimony.
For instance, preparing for the trip home, Arroway realizes her experience has become very “theological.” As the novel narrates: “Here were beings who live in the sky, beings enormously knowledgeable and powerful [… who] could clearly visit reward and punishment, life and death, on the puny inhabitants of Earth. Now how is this different, she asked herself, from the old time religion? The answer occurred to her instantly: It was a matter of evidence… There would be five independent, mutually corroborative stories supported by compelling physical evidence.”
from Adafruit Industries – Makers, hackers, artists, designers and engineers! https://ift.tt/h9gBuk1
via IFTTT
Комментариев нет:
Отправить комментарий