破除对人工智能的误解
Great champions know how to end their career at the peak of their success. DeepMind, Alphabet’s artificial intelligence lab, has decided that its computer program, AlphaGo, will no longer focus on winning the board game Go. Instead, according to a blog post co-written by Demis Hassabis, DeepMind chief executive, the focus will be on “developing advanced general algorithms that could one day help scientists as they tackle some of our most complex problems, such as finding new cures for diseases”. The ambition is justified. AI may lead to remarkable breakthroughs, especially in the hands of thoughtful people. Yet these spectacular developments may be more easily achieved if a misunderstanding is removed. Some seem to think that AI is about coupling artificial agency and intelligent behaviour in new programs. Yet the opposite makes more sense: it is about decoupling successful problem solving from any need to be intelligent. Only when this is achieved is it successful. The best definition of AI was written in 1955 by US computer scientist John McCarthy and colleagues, part of the classic proposal for the Dartmouth summer research project on |