By figuring out how bad actors might use DALL-E, ChatGPT, and other generative AI programs for evil, OpenAI can hopefully prepare for the worst.
Source link