To me it’s as simple as “who knows best how to harness the premier LLM – Anthropic, the lab that created it, or this random person?”

That’s why I’m only interested in first party tools over things like OpenCode right now.