Post #1874316

@reiddragon What did this moron expect? LLMs are actually working exactly that way: they parse the user's input and collate code by using tokenised code snippets in their data base. You never get the same result twice, because even minor subtleties in the prompting can trigger an entirely different decisioning. And yes, when coding you often can address a requirement in different ways. A human would stick to one way, an AI ... not so much. And now he's complaining about inconsistency? 🙄