Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror

Comment Re:And ? (Score -1) 25

LLMs are not really designed to learn on the fly. They essentially trust all their sources equally since they just absorb all information they're fed.

That's something newer AIs are trying to solve by letting them experience the real world and develop their own concept of trust and decide which sources are trustworthy based on lived experiences. Just like human children do.

Slashdot Top Deals

You can't cheat the phone company.

Working...