@macsparky it takes a lot of effort and guardrails to safely use any non-deterministic system. I’m very pro-AI-where-it-fits, but those use cases are much smaller than the current crop of AI companies are letting people know. For people that want a system like this, I think you should build something up from scratch using pieces you (as the user, individually) understand. I think it’s still a very fertile landscape for labs videos. Try to fill in those knowledge gaps so the typical user can learn what they aren’t being told by LLM publishers…how to not shoot themselves in the foot.