OpenAI and Anthropic were founded with exotic corporate structures (private company owned by nonprofit, etc.) because the founders took seriously the existential risks of AI and wanted to self-limit, suggesting at least some industry actors are not purely motivated by profit.