@short_straw@ZachAbramowitz it basically allows Claude to run the AI Deep Research on its own and then keep running additional AI Deep Research as needed to get to the bottom of an issue. It has been really helpful.
100%! the moment I can do what Opus 5.5 can do on a local LLM it will be done for these companies. Agreed too re watermarking, it is really making me want to try ChatGPT again. I just bought a $5k Mac Studio and it can barely run the frontier open source models. I think the move is for a company to allow people to rent compute to run their local models. Are you going to stop using Claude when the watermark starts?
@short_straw@AnikaCodes have you tried to run a local llm? You need at least a $15-20k computer to even come close to a mid Claude model. Even then, its at best meh.
Nothing funnier than a judge denying your motion to continue, giving you a whole speech about "no more delays," and then putting you 67th on the trial docket. See you two hurricane seasons from now, your honor.