I have to admit, I slept on Claude at first, and that was a mistake. I was already using several other AI tools and just didn’t see the need to further complicate my stack. Perplexity for research tied to a local LLM for summarization, and I didn’t see the need for more chefs in the kitchen. Except, I should have fired them all and moved over to Claude Code, because it’s far more versatile and powerful for the workflow I use for ideation and research.