Hacker News new | ask | show | jobs
by Yorrrick 4429 days ago
Whatever consciousness is, it's not obviously required for an AI to start manipulating its surroundings to its advantage, which is one if the fears here.
1 comments

True... the possible dangers of AI can exist even if AI never achieves true consciousness. Many forms of life have evolved to service their own needs at the expense of other forms of life, and most of them either have no brains or brains too primitive to be likely to involve consciousness as we know it. There's no reason AI can't do the same thing.
I recommend a book by Peter Watts called Blindsight. It explores the question whether consciousness is needed for intelligent behaviour.