Hacker News new | past | comments | ask | show | jobs | submit login

My generative melody models have done it for over a year but that's with human curation, so it's not self-improving. It's typically easier to curate than to generate, and it's especially true for music and images. It's much simpler to recognize a good melody than to compose a new one. The same applies to writing, but to a lesser extent.



Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: