I recently read an explainer on how LLMs work. It was great. It described how the probabilities of the next work were calculated. I'm pretty sure it was by a comm scholar because I remember thinking that this was impressively technical. But I can't remember who wrote it or where I read it. Help?

Geopolitics are complicated

105 minutes without checking the news.

Today I shall mostly be migrating

I'll be giving a talk titled "You've got a friend, until they block you: The management of personal ties in the digital age" hosted by the Centre for Digital Humanities, Cultures and Media, The University of Manchester, on 25/9/25 at 3pm. If you're in the area, please come!

New special issue, "Comparative Approaches to Studying Privacy," edited by #CPRN is now published in Social Media + Society! journals.sagepub.com/topic/collec... w/ @lutzid.bsky.social, Lemi Baruh, Kelly Quinn, @masurphil.bsky.social, Carsten Wilhelm (comparativeprivacy.org)

How does the professionalization of Trust & Safety shape global content moderation practices? We compare Community Guidelines using the typology of harms from TSPA to identify policy priorities and exclusions, available #openaccess in New Media & Society: doi.org/10.1177/1461...

You can rationalize it however you want but staying on twitter is a terrible choice. It is not the “town square.” It is the world’s disease-ridden dump. You don’t need to be there to think critically or write relevantly or stay abreast of current events.

The best time to quit X was two years ago, the second best time is now...