The LLM Critics Are Right. I Use LLMs Anyway

I agree with almost every valid critique of LLMs, from environmental damage to the flood of low-quality code, yet I still rely on them daily. This dissonance is shared by many engineers who applaud critical talks while using tools like Claude Code. Despite the risks to open-source trust and geopolitical instability, I believe we must embrace local, open-weight models to maintain independence and shape the future of AI rather than fighting an unstoppable tide.

"In a world hurtling towards AI, we believe humans are the best agents."