[seanhunter]

Because there’s a world of difference between a reinforcement learning trained special purpose model and asking a general purpose large language model to have a go at something.