I'm not sure that would be very helpful if you were prosecuting these 2 people in a court of law.
10.08.2025 20:24 β π 0 π 0 π¬ 1 π 0@patricksturg.bsky.social
social science methodology, sport etc.
I'm not sure that would be very helpful if you were prosecuting these 2 people in a court of law.
10.08.2025 20:24 β π 0 π 0 π¬ 1 π 0I mean that if a bunch of left-wing liberals flee X to Bsky because X is too rightwing then follow a heuristic of "block jerks, don't feed trolls", it's a recipe for very boring ideological conformity. This is why Bluesky is boring and rapidly losing subscribers.
10.08.2025 20:08 β π 0 π 0 π¬ 1 π 0will have a read!
10.08.2025 20:01 β π 0 π 0 π¬ 0 π 0the big problem is when you move beyond banning what is unlawful (not itself unproblematic given many laws), how do you decide what can and can't be said? That is how you end up with Twitter banning women from saying there are only 2 sexes.
10.08.2025 19:46 β π 0 π 0 π¬ 1 π 0do you also think that banning legal speech is problematic, i.e. once the threshold moves beyond what is lawful?
10.08.2025 19:37 β π 0 π 0 π¬ 1 π 0And that pressure can result in lawful speech being banned, such as women expressing gender critical beliefs on pre-Musk Twitter. Making perceived harm the threshold is therefore highly problematic for free speech in a democratic society.
10.08.2025 19:20 β π 1 π 0 π¬ 1 π 0So who gets to decide whether these tweets, that you presumably selected as being egregious examples, are a "coded incitement to violence" or observations?
10.08.2025 19:05 β π 0 π 0 π¬ 1 π 0Haven't been on here for a while but popped over to take a look because Twitter is so awful. Some interesting discussions going on but I'm mostly struck by the rigid ideological conformity and thirst for punishing those who indulge in wrongthink. Should I try Threads?
10.08.2025 18:23 β π 1 π 0 π¬ 5 π 0Not defending them obviously but it kind of does make a difference in understanding the motivation in that mindless morons boo the opposition no matter the extent of the bad taste. I've witnessed it many times.
10.08.2025 17:42 β π 0 π 0 π¬ 0 π 0It was the Palace fans
10.08.2025 17:26 β π 0 π 0 π¬ 1 π 0the obvious defence the people you have screenshotted would mount is that they aren't calling for terrorist violence but predicting it will happen.
10.08.2025 17:21 β π 2 π 0 π¬ 2 π 1It's rooted in complying with the law. But practically, trans-identifying women ('trans men') aren't the problem here, no man is going to care if they use the men's toilets. Alternatively, they can use mixed-sex toilets.
10.08.2025 17:13 β π 1 π 0 π¬ 0 π 0My, admittedly somewhat limited, experience of GPT-5 has not been good. Particularly weird that it didn't even know how to code API calls to itself.
09.08.2025 18:39 β π 2 π 0 π¬ 0 π 0New working paper with Tom Robinson, Laura Fung and Caroline Roberts. We use an LLM to classify occupations in surveys in real time, probing 'intelligently' when more info is needed. Results show big reductions in cost, time and respondent burden.
osf.io/preprints/so...
π£ "Uncovering Digital Trace Data Biases: Tracking Undercoverage in Web Tracking Data" by Bosch et al. @orioljbosch.bsky.social @patricksturg.bsky.social Read: www.tandfonline.com/doi/full/10....
Featured in our latest issue!
New working paper with Chuyao Wang and Dan De Kadt, "AI labeling reduces the perceived accuracy of online content but has limited broader effects". The title is a concise summary but read the full paper for details: arxiv.org/abs/2506.16202
23.06.2025 10:02 β π 1 π 1 π¬ 0 π 0The greatest tennis match of all time.
08.06.2025 19:08 β π 2 π 0 π¬ 0 π 0Interesting, I've never been to Reading.
08.06.2025 16:53 β π 2 π 0 π¬ 1 π 0these are what are called ad hominem arguments. Look it up.
08.06.2025 16:03 β π 1 π 0 π¬ 0 π 0On my way to London for #MASS25. Excited for two days of discussions on Mobile Apps and Sensors in Surveys. Shoutout to my fantastic co-organizers (for the 6th (!) time) @peterlugtig.bsky.social, @jkhoehne.bsky.social, Bella Struminskaya and generous local host @patricksturg.bsky.social
03.06.2025 11:44 β π 5 π 2 π¬ 1 π 1headshot of professor patrick sturgis
π€ Recent advances in AI are beginning to reshape how we design, implement, and analyse surveys.
@patricksturg.bsky.social⬠will be sharing his expertise at UCL's AI for Survey Data Collection Methods and Data Analysis workshop on 9 July.
@clscohorts.bsky.socialβ¬
No @theguardian.com the SC judgment is absolutely clear. What you are struggling with is how you lot are going to keep misrepresenting the law now the judgment is in.
www.theguardian.com/world/2025/m...
Isn't having coppers out looking for you being tracked by the state?
24.05.2025 18:06 β π 0 π 0 π¬ 1 π 0Really can't see what the problem is with AI face recognition as long as it's only used for identification. Just much more efficient and less costly than using humans to do it.
www.theguardian.com/technology/2...
Does local area social mobility affect political alienation? Andrew McNeil & Patrick Sturgis (@patricksturg.bsky.social) find it depends. Read OPEN ACCESS: buff.ly/ptn4JQC
@polstudiesassoc.bsky.social @uoypolitics.bsky.social #ukpolitics @sagepub.com #polsci #polsky
New working paper "Survey Experience and Nonresponse in an Online Probability Panel: A Survival Analysis" with Katya Kostadintcheva & Jouni Kuha. We find some interesting predictors of nonresponse. Read the paper to find out: arxiv.org/abs/2505.03971
08.05.2025 13:01 β π 2 π 0 π¬ 0 π 0Unpopular opinion: Vice Chancellors aren't overpaid and VC pay is just a distraction from the real issues.
www.thetimes.com/uk/education...
West Ham have gone backwards under Potter, thatβs 7 on the bounce without a win. The only reason weβre staying up is the points won under useless Lopetegui. Potter will have zero fan patience if we get off to a poor start as we usually do. Iβd sack him if it were up to me.
26.04.2025 16:31 β π 0 π 0 π¬ 0 π 0yeah, wouldn't work for buses but a reasonable solution for trains.
24.04.2025 08:00 β π 1 π 0 π¬ 0 π 0There could be designated carriages where people who want to scroll tiktok with the sound on can all sit together.
24.04.2025 07:55 β π 0 π 0 π¬ 1 π 0