@LefterisJP: local tools and local AI matter centralized models are a liability

Also when you want to hack your own system, it feels safer to invest in your own hardware, as I imagine getting banned from Anthropic is a pretty bad outcome for a company