News for hackers. Stuff that matters.

A comment on Using a Markov chain to generate readable nonsense with 20 lines of Python. Parent comment

Re:Using a Markov chain to generate readable nonsense with 2... by tempodox on Monday January 01, 2024 @09:22AM (#38831961)
And the logic salad problem will not be solved by LLMs because there is nothing in their construction that makes them capable of understanding logic.
[ Parent | on HN ]
Re:Using a Markov chain to generate readable nonsense with 2... by brookst on Monday January 01, 2024 @12:01PM (#38833247)
Can you elaborate? I would argue that logic has even stronger token prediction characteristics than language.

As a total handwave, I would expect an LLM trained on formal logic and a huge corpus of proofs to produce pretty strong logic output.

[ Parent | on HN ]