I wonder when things will progress to the point where you could run a LLM that would do a good enough impression of a therapist on a 4090. I honestly think the

I wonder when things will progress to the point where you could run a LLM that would do a good enough impression of a therapist on a 4090. I honestly think the only way I could trust something like this is if it was running on my own hardware, or a private cloud instance.