The problem is you can not teach it to never be wrong. There isnt ever a point where you could trust it like somebody who learned the job they have been doing for years on end. There will always that bit of randomness in there

October 17, 2025 - 11:26 UTC

5
0
108
It has no concept of right or wrong, only positively and negatively reviewed sentences. An LLM is a language model not a truth model. If anybody wants to try making a truth model then that might be interesting but an LLM can't become a truth model just by hoping it will.
0
0
8
So it's worse than the example because you could become a tool helping it and it will never actually be reliable.
0
0
1
Yeah! The wild card! Like Charlie! Exactly how things should function!
0
0
1
it's part of the design!
0
0
25
Yup. pretty useless. If calculators spit out bullshit numbers we, rightfully, wouldn't use them.
0
0
2