News for hackers. Stuff that matters.
Re:Google scrambles to manually remove weird AI answers in s... by vidarh on Sunday May 26, 2024 @04:25AM (#40480527)
And for most cases that human explanation is likely with a disturbing frequency a complete fabrication after the fact. See experiments on split brain patients.

With respect to repeatability, yes, LLMs are currently frozen in time. That is not an inherent limitation, but it is one that is practical for a lot uses and a problem for some.

There are no comments on this comment yet.