Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Google is sitting on ~20 years of gmail, but I can imagine the headache of both cleaning the dataset and likely consumer blow back.

They also have youtube, which almost certainly has enough good data to train a powerful model on it's own, but also seems daunting to clean up first.



They can't legally be allowed to use email content to train public models. The data leaks that could result will be unbelievable. It could be used to train private LLMs within organizations that own the emails. But Youtube comments work.




Consider applying for YC's Fall 2026 batch! Applications are open till July 27.

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

Search: