TED : We can't control what our intelligent machines are learning

Go down

TED : We can't control what our intelligent machines are learning

Post by Peyrat Tom on Sun Oct 30, 2016 1:03 am


In the video, the author speak about the ethic in AI developing. The progress in this sector allow machines to review us, thanks to what we do on internet for exemple, but also on our phisical characteristics, and that's a real part of the problem. The author think than we can't control the way machines make decisions for us in some areas, cause we don't know exactly how it works (she compare it to "puppy machine creatures"). She also fear the way machines can make predictions about us, like predict the depressions, or the chances for a woman to become pregnant, she think thoose algorithme can make unfair decisions about employement, imprisonement and so on ... based on inference, and even on ethnical characteristics. she conclued by saying than AI can fail in ways we're not prepared to.
In my opinion she's right on many points, allow machines to make decisions based on things who "probably" happened would killed humanity (the "human" side of decisions, not the human beiings), it looks like a kind of pessimism, fatalism way of thininking, based on Murphy's law, guessing the worst way will happened, whatever, and so can lead to make this happen. for exemple a man not hired cause the machine think he well become depressiv, would probably finally become depressiv cause he can't get a job, so he can't pay his rent, and so on ...
But on an other side, well used, and associate with human to take the decisions, and knowledge about how the machine take decisions, it can be really usefull.

Peyrat Tom

Posts : 11
Join date : 2016-10-05

View user profile

Back to top Go down

Back to top

- Similar topics

Permissions in this forum:
You cannot reply to topics in this forum