I’ve been studying and tinkering with open weight LLMs since the original llama weights leaked. I’ve very recently become convinced that the true data and compu

I’ve been studying and tinkering with open weight LLMs since the original llama weights leaked. I’ve very recently become convinced that the true data and compute requirements needed to fine tune and produce an “unsafe” model are orders of magnitude less than what’s needed today. We are no more than a year away from anyone with a 4090 being able to fine tune their own mistral. The cat is out of the bag on this one.