Dan Luu

The internet created a lot of new mediums. Blogs, vlogs, tl;dr culture, explainer videos, Twitter, podcasts, clickbait, ELI5, etc. It seems like we never saw new mediums emerge for academia. Probably because the same institutions are consuming and generating the content?

I don’t think of it like that. Journals build brands, and academics who contribute to the journals convert those brands into reputation for themselves, and thus into jobs. So the journals are competing on brand (mostly impact factor and similar), not the quality of the medium.

To pick a slightly self-interested example: most ML researchers would (currently) rather have a NIPS paper than a distill paper, even though the latter is a much more interesting medium. Why? The former has a better chance (at present) of getting you a job

This can and does change, but usually very slowly. One of the cleverest things done in establishing the preprint arXiv was to recruit leading physicists like Andy Strominger and Ed Witten to contribute in the first few days. Wham! Brand established.

Great points. I was talking from a fuzzier perspective. Distill represents one shift where more time and effort goes into a paper, but going in the reverse direction and embracing more casual mediums also seems interesting

For a lot of "learning in public" as @nayafia put it, there aren't expectations about the medium you use in the first place. People can pick between blog posts, tweets, emphasizing visuals, switching up tone, playing w/ how data-oriented vs argument oriented their writing is, etc