Have you noticed that OpenAI is focused on intelligence that creates new content, be it pictures or explanations (gpt3)<p>While most of machine training efforts are focused on changing existing truths?<p>If a person is walking and you tell the computer to make that person be a different thing (a dog, a mailman, or a president) you are changing the truth<p>But if you say to the computer "Do a scene where a person walks" you are creating the truth