Przemek Grabowicz 🇵🇱🇪🇸🇩🇪🇺🇸🇮🇪
przemyslslaw.bsky.social
Przemek Grabowicz 🇵🇱🇪🇸🇩🇪🇺🇸🇮🇪
@przemyslslaw.bsky.social
CS faculty at University College Dublin and the University of Massachusetts Amherst: responsible AI, social media, news media, computational social science, applied NLP, open-world AI
RT≠endorsement, RT=interesting read
More: https://przemyslslaw.github.io
A small correction: the percentage from outside the network went up from 50% to 80%.
November 5, 2025 at 9:02 AM
We believe that the only way to achieve neutrality ⚖️ and free speech 💬 in social media is through transparency. Right-wing and left-wing policymakers need to talk with each other and online platform representatives to figure out how independent researchers can safely study social media. 🧵7/7
We want neutrality ⚖️ and free speech 💬 in social media. But can we achieve that without transparency? I don't think so. So right-wing and left-wing policymakers need to talk and figure out how independent researchers can study platforms. An excellent piece for @science.org, with my inputs. 🧵1/2
"Access to data by researchers is a crucial element of ensuring transparency and democratic accountability,” @lewan.bsky.social says. “The public is entitled to know what platforms and their algorithms are doing to the information landscape.”
I wrote for @science.org about Meta, TikTok and the DSA 🧪
November 3, 2025 at 12:58 PM
This research subsumes our previous study of the 2025 German federal election (attached figure and paper), before which Elon Musk supported and campaigned for the far-right AfD. We conclude that social media platforms can develop in different ways. 🧵6/7
zenodo.org/records/1489...
November 3, 2025 at 12:58 PM
We find that the fraction of posts from outside the network went from 20% to 50% following the 2023 change of the algorithm announced by X owner Elon Musk (attached tweet). It's possible that the algorithm also started to amplify political extremes around that time. 🧵5/7
x.com/elonmusk/sta...
Elon Musk on X: "A major update to the recommendation algorithm is rolling out over the next few days. This will help surface smaller accounts and posts outside of your friend-follows network. As always, this will be made open source and undergo continuous improvement." / X
A major update to the recommendation algorithm is rolling out over the next few days. This will help surface smaller accounts and posts outside of your friend-follows network. As always, this will be made open source and undergo continuous improvement.
x.com
November 3, 2025 at 12:58 PM
A prior study by @inference.vc of the 2016 Twitter's feed algorithm did not find evidence of amplification of political extremes in seven countries, including Germany (attached figure), so the feed algorithm of X must have changed considerably since then. In fact, it did, as we show next. 🧵4/7
November 3, 2025 at 12:58 PM
We test whether the feed algorithm favors political extremes, by regressing the number of feed appearances against user engagements received by each post and political party affiliation of its author. We find that each post from political extremes appeared about 20 minutes more in the feed! 🧵3/7
November 3, 2025 at 12:58 PM
The study (attached) was conducted during the culminating month of Polish presidential election. We tracked news feeds of four sock-puppet accounts and analyzed public posts of 485 Polish politicians and related users. 🧵2/7
doi.org/10.5281/zeno...
November 3, 2025 at 12:58 PM
This may sound like an impossible task, given how this year Trump and Republicans started to oppose online platform transparency [1]. However, in 2019 Trump and Republicans drafted a platform neutrality legislation [2]. 🧵2/2
[1] www.politico.eu/article/us-c...
[2] edition.cnn.com/2019/08/09/t...
White House proposal would have FCC and FTC police alleged social media censorship | CNN Business
A draft executive order from the White House could put the Federal Communications Commission in charge of shaping how Facebook, Twitter and other large tech companies curate what appears on their webs...
edition.cnn.com
November 2, 2025 at 5:23 PM
What if it's interpretable AI and users are in charge?
October 27, 2025 at 5:16 PM
I'm feeling inspired and thankful for having so many insightful conversations with extremely bright people! 😊🙏
September 27, 2025 at 1:08 PM
I've been pigeonholed by both sides. Whenever it happens it's a problem. That said, this topic is often twisted and exaggerated for partisan gains. What we need is balance, neutrality, facts, rules, and procedures.
September 19, 2025 at 12:59 AM