Soldier of FORTRAN ​Jun 14
I love the geniuses over on X finding out that LLMs have a bias by asking deepseek about the tiananmen square massacre then not applying that new understanding to claude or chatgpt.
Show threadA Random Half-GiantJun 14
@mainframed767
It's like an LLM equivalent of the Gell-Mann amnesia effect.
Show threadSoldier of FORTRAN ​
@Ent I'm almost certain that LLMs specifically prey on that exact effect.
Jun 15 at 6:41pmWeb
Show threadA Random Half-GiantJun 15
@mainframed767
That one and Dunning-Kruger, at a minimum