News for hackers. Stuff that matters.

A comment on OpenAI’s latest research paper demonstrates that falsehoods are inevitable.

Re:OpenAI’s latest research paper demonstrates that falsehoo... by justcallmejm on Saturday September 13, 2025 @02:59PM (#45234450)
This is why a neurosymbolic system is necessary, which Aloe (https://aloe.inc) recently demonstrated exceeds performance of frontier models, using a model agnostic approach.
[ Parent | on HN ]
Re:OpenAI’s latest research paper demonstrates that falsehoo... by HarHarVeryFunny on Saturday September 13, 2025 @05:10PM (#45235336)
No - humans are the counter example.

If you want a model that doesn't hallucinate then train it to predict the truth, and give it a way to test it's predictions. For humans/animals the truth is the real world.

An LLM is trained to predict individual training sample continuations (a billion conflicting mini truths, not a single grounded one), whether those are excerpts from WikiPedia, or bathroom stall musings recalled on 4chan. Based on all this the LLM builds a predictive model which it is then not allowed to test at runtime.

So, yeah, we should stop doing that.

[ Parent | on HN ]