You must log in or register to comment.
Llms are not built for accuracy. The more people that understand that the better.
Llms are designed to mimic human answers to questions, nothing more.
The fact that they happen to get things right half the time or more is a happy coincidence, not a function of the goal.
Yup, the oligarchs want us to be reliant on AI because they’ll be in control of the answers we seek while we will consider ourselves enlightened.
Same as it ever was.
This is just the new preacher
Ooof.
Oh ah, I mean “hmmm”… (OP’s title:-)
Tried this prompt with deepseek-r1:14b, it seems to get it right about half the time, otherwise giving the wrong quote or claiming it wasn’t really Thomas Jefferson who said it.
