4 ms·
The article goes into deep discussions of AI maximizing reward at the cost of humanity. Is the answer to develop an AI/ML that is rewarded when it finds and sh
by OrangeMonkey 4y ago
The article goes into deep discussions of AI maximizing reward at the cost of humanity.
Is the answer to develop an AI/ML that is rewarded when it finds and shuts down an AI/ML that is acting suspicious?
- IceMetalPunk 4y agoNot particularly, because who's to say that "cop AI" doesn't screw over humanity in its attempts to shut down an AI it's deemed "suspicious"? The current usual approach here is to train AIs without specific goals or optimization functions. Instead, they're rewarded by human reactions to their decisions. As such, they effectively don't learn "do this thing", they instead learn "do what the humans want, avoid what they don't" -- and through that, they indirectly learn to "do the thing" (because the humans want it) while avoiding dangerous outcomes (because the humans don't want behaviors that lead to those).
- gs17 4y ago> Not particularly, because who's to say that "cop AI" doesn't screw over humanity in its attempts to shut down an AI it's deemed "suspicious"? In the cliche sci-fi story version of this, the cop AI simply realizes eliminating humanity is the only way to cut "bad" AIs off at the source.
- IceMetalPunk 4y agoYeah, I, Robot was a good movie.