OpenAI, the company behind the GPT-3 neural network that is capable of generating plausible-sounding text passages on any given subject, have now extended it https://www.theregister.com/2021/01/07/openai_dalle_impact/ to generating images based on any given caption.

How will this impact content creators? Given that the net is trained on data copied off the Internet, what happens if/when it spews out something a little too close, copyright-wise, to some existing work of a human artist/writer?