Web Log: It’s all Go with artificial intelligence

Oxford techno-philosopher Nick Bostrom warns of the dangers of developing artificial intelligence without thinking through all possible risks

In his latest best-selling book Superintelligence, Oxford techno-philosopher Nick Bostrom warns of the dangers of developing artificial intelligence without thinking through all possible risks, including the enslavement of humankind by a second (artificially) intelligent species. Google and Facebook don't seem to be listening – they're too busy creating AI programmes that can beat the pants off human players at the ancient Chinese board game Go.

Not content with rolling out live video functionality for iOS users and announcing a new data centre here in Ireland, Facebook's Mark Zuckerberg bragged about how their AI team has finally cracked Go, something computer science has been working on for the past 20 years. Interestingly, Google posted its own Go-related announcement hours later. It turns out that its DeepMind research division has created a programme that wipes the floor with all others, including Facebook's. It might just be a game but this complex, strategically thinking AI is far from child's play. https://googleblog.blogspot.ie/2016/01/alphago-machine-learning-game-go.html