3 ms·
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
by im3w1l 9y ago
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.
- 19eightyfour 9y agoI 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.