> If the tech really is dangerous then being more closed arguably is in the public's best interest, and this is certainly the reason OpenAI have previously given.

Then what should we do about all the open models that are closing in on OpenAI's capabilities?