Congrats Dan!
Posts by Timothy Graham
So happy to have my first journal article published today! I argue that LLMs animate a paradox in which similarity only emerges through difference. I show how this paradox is operationalised by LLMs as they generate a natural language response (1/2)
link.springer.com/article/10.1...
I actually set up a bot to spam replies to one of them - harmless stuff and labeled as a bot - and he blocked it 🤷
Back in the halcyon Twitter days there used to be a small group of researchers and internet pundits who tried to “debunk” the field of social boy research, under the proviso that such bots didn’t exist and, if they did, couldn’t do anything meaningfully harmful or impactful
Laure, at this point I’m afraid to ask. Good to see the reality of this - even if it’s professionalised now, unfortunately. There’s still people out there who are “bot denialists”.
Inside a social media bot farm
We've built a searchable & visual database of every X's Community Notes ever proposed. 2 million & counting.
The dashboard displays both visible notes (those shown publicly on X) & not-yet-visible notes (proposed notes that have not yet met the visibility thresholds). notetracker.socialmedialab.ca
Alton Towers
because noone wants to see this
Daily reminder: do not give Tucker Carlson any attention. He is a grifter and a sophist, and he is part of the reason we are here right now facing domestic threats to global peace and wellbeing.
Late capitalism mask off
This is the one where the only surviving people on earth are four stranded astronauts who were on a moon mission when the deranged lunatic in charge of the -
Trump saying a whole civilization will die tonight, never to be brought back again.
FIFA World Peace Prize material right here.
This is a sickening sentence for anyone to utter - let alone the sitting US President.
That’s true, though to be fair this looks less politically motivated/driven and more economic - at least from the evidence available. It’s clickbaity slop that causes political consequences, certainly, but that seems less of a concern than simply finding stuff people click on and like
Latest commentary from me and @snurb.info in this excellent investigative piece by Michael Workman and the ABC NEWS Verify team
Generative AI and its intersection with engagement-driven social media logics continues to be a thorn in the side of democracy...
www.abc.net.au/news/2026-03...
How have I only just discovered this game? Spoilers: it does have some graphic content at the end, but it speaks so much to our media system and the attention economy
ncase.itch.io/wbwwb
exactly!
Latest commentary from me in BBC Verify on the surge of AI-generated war misinformation.
tl;dr the barrier to creating convincing synthetic conflict footage has collapsed, while engagement-driven monetisation is accelerating its spread.
www.bbc.com/news/article...
Congratulations Dan!!! Very proud supervisor moment
Super thrilled to have completed my final seminar today and to be told that I’m almost ready to lodge my thesis for examination :”) huge thanks to my supervisors @timothyjgraham.bsky.social @petamitchell.bsky.social and Benjamin Nicoll for their help over the past three years
Thanks. My point is: focus on the substantive findings of the Nature paper, which is the focus of the article. Whether Moses was racist or not is beyond the point. Infrastructure embeds politics - this we know. If social media is infrastructure, then we ought to understand how it is political.
Thanks Tilman - fair point. But it is contested and not “largely debunked” - I kindly suggest to read beyond the first google result. It’s an illustrative example, at any rate. The Nature paper and my empirical research provide the core evidence. Along with a growing body of research.
Do it! That old traditional roguelike I’ve been working on since before Covid finally has a build almost ready. It’s super CoQ inspired! If you’re interested I’ll send a link - I’ll pop it on itch soon
Really interesting new paper from some of my former @gesis.org colleagues in @bigdatasoc.bsky.social: "From (almost) open to heavily restricted data access – The development of the Twitter/X developer policies"
doi.org/10.1177/2053...
#commsky #computationalsocialscience
🤣 so erudite that I actually replied in a civil, rational, Habermas-idealist, Plato-approved way
Congrats Caroline!
Aaaand here comes the fan mail