> might be even more reliable simply impossible. when a input parser breaks, it breaks. with a LLM, it breaks... sometimes, for different reasons. Let alone th

> might be even more reliable simply impossible. when a input parser breaks, it breaks. with a LLM, it breaks... sometimes, for different reasons. Let alone the enourmous computing power needed compared to simple input parsers.