你应该害怕人工智能吗?
Eliane Mitchell

19 September 2017

人工智能接管假想事件，即计算机或机器人接管世界，消灭人类——这是科幻小说和世界末日电影中常见的比喻。但超级智能人工智能真的是我们应该害怕的东西吗?

In thisTEDTalk，科学家和哲学家格雷迪·布克不这么认为。While movies likeThe Matrix,Metropolis, andThe Terminatorexacerbate humans' fears of being supplanted by technology—that is, that we might develop technology that is much too advanced for own good—we forget, in Booch's view, an important point. Engineers are not looking to build sentient machines, they are looking to build "simple brains" that can simply carry out tasks. And even if engineers did manage to develop the technology to make systems that have a theory of mind and ethical and moral foundations, he argues, we would teach them our own moral systems, not ones which would try to subvert us. Plus we can be assured that we can always unplug what we have built.

但是，Booch是否对超级智能人工智能的天真过于乐观了呢?是否有一些技术的发展让你担心?Enter your comments below, and check out his TedTalk here:

https://www.ted.com/talks/grady_booch_don_t_fear_superintelligence/

Comments(2)

Harold G. Neuman's picture

Harold G. Neuman

Wednesday, September 20, 2017 -- 11:26 AM

For the first twenty plus

For the first twenty plus years of my life, I loved the world of science fiction. Heinlein; Herbert; Anderson; Bradbury; Clarke and so many others I have forgotten. I recall a short quotation, the origin of which escapes me now: What the mind can conceive, the man can do. I think this answers the question regarding innocence and AI. Let's think, for a moment, that science and technology are virtually limitless (I'm only playing the Devil's advocate now, so don't get your knickers in a bunch). After all, we do have a tendency, as a sentient species, to believe we can do most anything. And, in the long view, up to now,we have done much---some of which we probably should not have done. Its creators anguished over the advent of nuclear energy and the bomb. Paul Tibbetts carried that anguish for many years after flying the Enola Gay over Japan and dropping an atomic bomb. Technology does not worry me. Our will to restrain it does.

elianem's picture

elianem

Tuesday, October 31, 2017 -- 9:59 PM

Hi Harold,

Hi Harold,

I like the way you put this a lot. Your ending line makes me think of Amazon, Google, IBM, Facebook, Apple, DeepMind, Microsoft, and IBM's recently formed coalition, called the Partnership on AI to Benefit People and Society. The coalition aims to refine AI companies' practices to promote things like "fairness... transparency... values and ethics," etc, in the hopes of preventing technological "catastrophe" and better (self)regulating the inventions they create with AI. But Professor Annette Bernhardt of UCSF offers a different proposal (see here:http://bit.ly/2il8dOH) which argues that regular citizens should be included in the coalition's decision-making process. I am curious to hear your thoughts.

Tags