[thegabriele]

If this the case? If the LLM must be re-trained each time a new problem arises, it means that it doesn't "reason"...so what's the point?