News for hackers. Stuff that matters.

A comment on Hard Questions: How We Counter Terrorism. Parent comment

Re:Hard Questions: How We Counter Terrorism by im3w1l on Friday June 16, 2017 @06:10AM (#14567737)
I think we should try to stop terrorism, but not "at any cost".

Using AI model's to monitor all 1-1 conversations in the world is too dangerous of a power - it increases the risk for tyrany much more than it decreases the risk of terrorism.

[ Parent | on HN ]
Re:Hard Questions: How We Counter Terrorism by 19eightyfour on Friday June 16, 2017 @07:24AM (#14568024)
I haven't thought much about this....But my first reaction is -- can we at least test it to see what happens?

I have something that might not be that relevant. Consider China's control of conversations on Weibo as a microcosm of "all 1-1 conversations". Assume they are using some sort of AI there ( there's a good argument for any sufficiently large bureaucracy being AI-like ). Is it really so bad? They've tracked huge GDP growth for 70 years and very low civil conflict, and much harmony, I think, from some points of view, definitely.

[ Parent | on HN ]