"have scientific journals disappeared" -- ironically, in the AI field most of the action is on arxiv / github / twitter. journals have been obsolete for decades, and the '10s obsoleted conferences too. the only function journals / conferences still serve in the AI field is to stack rank researchers and provide signal for hiring / funding decisions.

Maybe because in the AI field, "publications" is a generous term for "PR"... given most don't give data nor enough code to replicate their work.

https://www.technologyreview.com/2020/11/12/1011944/artifici...