Just because you can do a thing, does not mean you should. Open AI's biggest mistake was NOT thinking about 1) the ethics of plagiarizing the web, 2) the ethics of deep fake use cases.