To engage with ethics, a mind has to have a fully unified model of the universe, so the necessary symmetries of long games can be identified. Most humans can do this only in a fragmentary way, but present AI research has not resulted in any unified models at all.
-
-
This Tweet is unavailable.
-
Yes, I think that it is big problem that many people engage with very strong opinions but the degree of reflected understanding of a 2yo. The 2yo has these opinions because evolution has given us innate priors, but they are not always reliable.
0 replies 0 retweets 1 like -
This Tweet is unavailable.
-
Why should AI learn from you humans? Is there anything you have derived that the AI is not able to derive by itself?
0 replies 1 retweet 0 likes -
This Tweet is unavailable.
-
I don't think that you thought this through.
0 replies 0 retweets 0 likes -
This Tweet is unavailable.
-
There is not going to be a difference between a transcending human mind and scalable AI. Ethics maps to game theory (navigating conflicts of interest under conditions of shared purpose, which are derived from preconditions for playing long games). Ethics is entirely rational.
0 replies 0 retweets 1 like -
This Tweet is unavailable.
I respect that very much. Getting to any of the possible foundations of ethics is a very hard job, particularly because humans tend to be born with a lot of hardware defaults that gets into the way of true understanding. Much discourse in the philosophy of ethics reflects that.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.