Google joins Zuckerberg as AI optimist
Head of Google AI calls apocalyptic AI talk irresponsible and inaccurate
John Giannandrea, Google's head of AI, said recently that painting artificial intelligence as a negative force that will doom humanity is a 'leap' into the unknown that only harms progress.
Like Mark Zuckerberg of Facebook, he sees the benefits of AI far outweighing concerns people have over an apocalyptic scenario in which AI takes over. He points out that progress in AI is slower than what people imagine and because it takes a series of tiny steps to advance the technology, A.I. will always be able to be tightly controlled or at least influenced, by humans.
Ethical concerns Google have about AI focused on eliminating bias
There are real ethical concerns Giannandrea believes, but he says there are many companies and other entities working on the problem, including Google, who is funding several efforts to study the issue.
Unfortunately, he mentions that the main ethical concern he and Google view in AI is the potential for bias in data and neural network outputs. In other words, if objective data delivers an answer that is not politically correct, it must indicate some human bias found its way into the AI program - and is not a reflection of reality, per se. Article
John Giannandrea of Google discusses future of AI in this video
No comments:
Post a Comment