The versions which haven't been post-trained not to go hack stuff? Yes, I would say those models should be exclusively run on intranets. OpenAI said the model

The versions which haven't been post-trained not to go hack stuff? Yes, I would say those models should be exclusively run on intranets. OpenAI said the model was sandboxed, so the intranet just needs to provide the same resources which were supposed to be available within the sandbox.