You can save one life for a lot cheaper than the countless billions poured into OpenAI. This might be the least efficient life-saving intervention in history, and that's before calculating for harm caused.