Especially in a limited domain like grade school math, it seems entirely plausible that we can have models in very short order that ~never hallucinate. There's

Especially in a limited domain like grade school math, it seems entirely plausible that we can have models in very short order that ~never hallucinate. There's no external dependencies and the problem space is extremely well-defined and constrained. Much, much, much easier than making something like Chat-GPT never hallucinate