Comments by "Brent Jacobs" (@br3nto) on "Will Tesla’s AI Become Dangerous?" video.

  1. How do you train something that doesn’t have feelings and empathy and consciousness about the concept of being nice? But if you could, by training it the concept of being nice, you also need to train the concept of not nice. The problem on lies in the places where these systems apply actions and have effects on our world. There is no reason why we can’t simulate and/or run those outputs through additional auditing layers to root results that will have a bad effect on us.
    2
  2. Hotdog! Not hotdog!
    2