It’s the only explanation that I can come up with for these CPU spikes. Cursor’s quality went way up when I also started getting “can’t connect to OpenAI” messages while talking to Claude, but my system performance absolutely tanked. These sorts of CPU spikes would seem to indicate that a local LLM is orchestrating a multi-modal approach to answering questions.
While that’s great news, it’s also something I’d rather have a bit more control over in terms of how much CPU it’s allowed to consume in one shot. But first Cursor would need to admit that it’s what they’re doing 