There’s a lot of fear around about Artificial Intelligence. South Koreans recently flipped out when Google’s AlphaGo defeated their Grand Master at the national board game. But will AI usher in the end-times for humanity?
Certainly Hollywood seems to think so. Cue: Skynet, Age of Ultron, Transcendence, The Matrix, War Game; or, even earlier, Colossus: The Forbin Project; hell, even as far back as Metropolis in 1927! And who can forget Hal 9000’s chillingly calm, ‘I’m sorry, Dave. I’m afraid I can’t do that’ when he refuses to open the pod bay doors for marooned astronaut Bowman in 2001: A Space Odyssey?
The equation seems clear: the first thing self-aware computers will decide to do is kill us. The most consistently upbeat portrayal of AI and humanity living side by side is in Iain M Banks’s Culture novels. The ‘Minds’ of the Culture are true artificial intelligence that would make Skynet look – and feel – like an abacus. And that’s the important difference: Banks’s AIs have feelings. To be self-aware is to have an opinion, to be drawn towards some things and repelled by others, and, consequently, to create and be guided by a moral and ethical landscape. The vast majority of humans are not homicidal psychopaths, so why should artificial intelligences be any different? Okay, some Culture AIs are crazy, or mildly anti-social ‘rogues’, but the other Minds keep them in check.
What Banks’s AIs value is uniqueness. Each Mind is constructed with a certain degree of randomness built in. They are all individuals, which is another requirement of true self-consciousness. The inescapable logic of this is that they also value the uniqueness of the human mind; for example, in Consider Phlebas the far more advanced Minds acknowledge that the character of Fal ‘Ngeestra has a way of looking at problems that is very useful. The Minds are partners with the people of the Culture, with each side bringing something important to the table. As a result, the whole civilisation is better for it. Maybe ours will be too.