My top goal for AI is to help me see better in every way possible. An interesting post by Charles_Rosenbauer on the HTM Forum answering the question, "Is there a Basal Ganglia theory equivalent to HTM?" raises the issue of blind spots to learning created by limbic disinhibition,
which got me to thinking of how to fix our constitutional limitations. The bad news is it may require having our neural nets make changes to our DNA that we don't understand in advance. The good news is that it is looking more and more possible. Here's how, for people who want to start training that part now:
Submitted July 27, 2017 at 05:01AM by phobrain http://ift.tt/2h5fRix
No comments:
Post a Comment