Super, I will watch that tomorrow; thanks! By coincidence I saw a trailer for that video in this one, which is very good.
For the sake of thread conversation: one of my thought processes is that I am not sure what ethical minefield one is jumping into by funding the tech-bro AI companies. I think we’re going to get a reckoning with forthcoming price-increases anyway; engineers who’ve been making hay with running sub-agents overnight may have to rein in their wild abandon when the subsidies run out.
That said, I will probably grab one of the least-worst options (Claude) and learn how to use it for a few months. I can see me pushing the button on local LLMs, and it would be a fun project to experiment with, but I wonder if there is an early-adopter time/price tax at present.