@tante I think "Intelligence" in AI is fake and dangerous. '
Mostly because of the people who think it's real and it's dangers – especially when failing to recognise an error – can be kept under control.
AI isn't able to acknowledge it can be wrong when the training incorporated contradicting factoids. Logic isn't overruling training, incomplete training makes AI blind to anything that is a simple conclusion. Without inferring there is no intelligence.
https://social.tchncs.de/@Ulan_KA/113603599263313317
No replies.
────
────