218
votes
Basically a way to tell the AI “this made sense” or “this didn’t make sense,” which the ai would then adjust itself accordingly. I dont actually know if this is feasible, the network might not work with an adjustment algorithm like this, but I thought it would be an interesting idea.
Comments: 3
-
03 Feb, '20
DevonThey could just track what gets /reverted and punish that.
-
31 Mar, '20
World ConquerorThis is already a feature in the AI app Replika.
-
15 Nov, '20
Devon Admin"An option to confirm correct answers" (suggested by John on 2020-11-14), including upvotes (9) and comments (0), was merged into this suggestion.