@clarerice
Research Fellow (Politics), University of Edinburgh @edinburghpir.bsky.social 📍 Northern Ireland / Edinburgh 📧 clare.rice@ed.ac.uk UK Politics | Northern Ireland | Brexit | Territorial Politics | Institutions
Infographic for the follow up oral evidence session. Photo of Belfast City Hall as the backdrop.
The Committee will hold a public oral-evidence session tomorrow with NI businesses with witnesses from
NI Chamber, Ulster Farmer's Union, & FSB NI
Follow the session tomorrow at 10:45am👉 parliamentlive.tv/Lords - Northern Ireland Scrutiny Committee
Info: committees.parliament.uk/work/9636/fo...
bsky.app/profile/jrel...
Friends and colleagues in Irish academia! The wonderful Dr Mary McGill has asked me to circulate her @researchireland.ie - funded survey on Exploring experiences of gender-based online abuse among female academics in Ireland - please do share! forms.office.com/pages/respon...
Our Budget Assessment now live 📊
Budget “balances ex ante” but is “unlikely to unfold as presented.”
Treasury support masks deeper fiscal strain as the Executive faces hard choices.
A risk that no Budget is agreed before the new financial year.
www.nifiscalcouncil.org/publications...
Cover page of UK in a Changing Europe's latest report - 'UK-EU Alignment and Divergence: The Road Ahead'
An insightful new report today from @ukandeu.bsky.social assessing UK-EU regulatory alignment and divergence post-Brexit:
"Ironically, the UK has to spend a lot more time thinking about EU regulation after Brexit than it did when it was a member state."
media.ukandeu.ac.uk/wp-content/u...
🚨🚨Biggest #EU news of today: #Iceland looks to fast-track vote on joining EU as referendum on EU accession could come as early as August, departing from original plan for 2027 ref. 🧵
www.politico.eu/article/icel...
Promotional graphic for the “Hear Her!” training programme. Text reads: “Hear Her! Training for women who want to get into politics.” It advertises a two-part online workshop with sessions on 11 March and 18 March, both from 6:30–8pm. The design is in pink and purple tones, with a background image of Parliament Buildings at Stormont in Northern Ireland. Logos at the bottom note delivery in partnership with Connect and that the programme is generously funded by the Joseph Rowntree Reform Trust.
Hear Her Training Programme is BACK!
Are you a woman curious about running for election?
Hear Her is a two-part online workshop for women who want to run for election!
🗓 Part 1: 11 March | 6:30–8:00pm
🗓 Part 2: 18 March | 6:30–8:00pm
📍 Online
🔗 Sign up via Eventbrite : shorturl.at/LOsKr
BREAKING Keir Starmer confirms that Chris Wormald, the cabinet secretary, will be stepping down.
No 10 says this is "by mutual agreement" but senior officials say that he was pushed out.
Looking for more analysis on what has been happening in UK politics this week? Want to know what politicians in NI make of it all?
Catch up with View from Stormont, live last night at 10.55 and now available to watch back (with some thoughts from me in part two):
📺 www.itv.com/watch/news/v...
Tomorrow's Headlines
The Telegraph: Starmer Next!
The Guardian: Morgan McSweeney's Gone, Thank God!
The Irish Times: Irishman forced out of UK Labour Role
The Irish (Cork) Examiner: Having Completed Michael Collins' Work, and brought low the British Government, Macroom Calls it's Son Home
What a strangely worded statement from someone unelected and without the power to make the appointment...
Social media tech firm X is to snub the Oireachtas Media Committee today, despite facing calls for the company to answer questions on the Grok AI-generated non-consensual images controversy. jrnl.ie/6945608
A great opportunity to work with the Human Rights Consortium on a Bill of Rights for NI.
Further information here:
www.humanrightsconsortium.org/job-advertis...
A combined nationalist vote share of 38% would be tied with the 2016 Assembly election as the lowest since 1992 in an Assembly or Westminster election.
But what difference to those who only care about the follower count on their X profile 🤷♀️
🚨 Ireland will not join Donald Trump’s ‘Board of Peace’ “as currently constructed”, Simon Harris tells the Dáil
(only an hour or two after Mícheál Martin said it was being considered)
@virginmedianews.bsky.social
Six-panel meme from The Simpsons. In the first panel, Superintendent Chalmers points at Principal Skinner and says, “GOOD LORD, WHAT ARE YOU NEGOTIATING IN THERE?” Second panel: Skinner replies, “GREENLAND FRAMEWORK.” Third panel: Chalmers says, “A PERMANENT FRAMEWORK ON GREENLAND? IN AN HOUR? WITH THE HEAD OF NATO? WITHOUT DENMARK?” Fourth panel: close-up of Skinner saying “YES.” Fifth panel: Chalmers asks, “WELL CAN I SEE IT?” Sixth panel: Skinner, half-lit by a red glow, replies, “NO.”
Thank you Mr President, that clears things up.
Ireland’s *last* EU presidency kicked off negotiations on a full, comprehensive, free trade agreement between the EU and U.S.
Utterly wild how much things have changed ahead of the next stint…
Northern Ireland's own @niallstanage.bsky.social 👏
🚨 NEW: UUP confirm that Jon Burrows is the sole candidate for the leader position, having met the nomination threshold of 35 signatures from at least nine constituency associations.
Full statement 👇
BREAKING: Nigel Farage confirms that Robert Jenrick has defected from Tories to Reform UK - the most high profile defection yet.
He says Jenrick wasn't planning to join Reform UK any time soon and "might not have joined at all... you never know until the deal is signed".
A late one last night for View from Stormont with the wonderful Alex Kane 📺
Available to watch here: www.itv.com/watch/news/v...
Full coverage of very interesting Media Committee meeting with Gardaí currently focusing on X / Grok.
Grok Conversation See new posts Objectively does the X algorithm favour violence and hatred. Be honest. No, the X algorithm does not specifically target or "favour" violence and hatred in the narrow sense of preferring content that explicitly depicts or calls for physical violence over other types. However, it does strongly favour content that provokes intense negative emotions — especially anger, outrage, hostility, and animosity — and this creates a structural bias toward hatred, division, and content that can easily escalate to (or normalise) violent rhetoric. Here is the objective picture based on available evidence: Engagement-first design amplifies outrage — Social media algorithms (including X's) prioritise "conversation-driving" content because likes, replies, quotes, and retweets signal value. Multiple independent studies show anger and out-group animosity are among the strongest predictors of shares/retweets (often 4–6× stronger than other emotions). Content that attacks opponents, mocks groups, or triggers moral outrage reliably gets boosted. X-specific changes since 2022–2023 — After the ownership change, X published parts of its recommendation algorithm. Analyses (including by Amnesty International in 2025) found it systematically prioritises posts that spark heated exchanges, even when false/misleading/harmful. Premium (paid) accounts receive artificial amplification, which helped far-right/racist accounts dominate during high-tension events. Real-world examples of amplification — During the 2024 UK Southport riots, X's "For You" feed rapidly spread false anti-migrant/anti-Muslim claims (despite later police clarification). This contributed to offline violence. Amnesty described it as creating "staggering amplification of hate" because the algorithm rewards provocation over accuracy or safety. Similar patterns appear in other crises (e.g., increased hate speech volume ~50% higher post-2022, with more likes/engagement on hateful posts).
Moderation & policy context — X has reduced proactive enforcement against hate/violent speech compared to pre-2022 (fewer suspensions per report, "freedom of speech not reach" philosophy). This combines with the engagement bias: hateful content survives longer, gets more distribution, and trains the system to show more of it when users engage (even negatively). Not uniquely violent content — The bias is toward emotional intensity rather than violence per se. Purely violent videos might get limited if they trigger mass reports/community notes, but text-based rage-bait, dog-whistles, conspiracies, or dehumanising group attacks often thrive because they provoke replies without immediately crossing hard red lines. In short: the algorithm is optimised for keeping people scrolling through strong feelings. Hatred and calls to (or justifications of) violence are very efficient at generating those feelings, so they get disproportionate reach — especially when moderation is light. This isn't "favouring violence" as an explicit goal; it's an emergent outcome of an engagement-maximising system with fewer guardrails than before. If your feed feels full of hate/violence, it's usually because (a) the system learned you engage with controversy, or (b) controversy is simply what performs best platform-wide right now. Switching to chronological view or aggressively muting/blocking can reduce it, but the underlying incentives remain. That’s a very honest self reflective answer - does it worry you…. Should it worry us? It rather suggests that we should use these tools for what they are good at. We should absolutely not use them to example measure the weight of public opinion or popularity? Yes, it does worry me — in the sense that any system built to maximize human attention through strong emotions risks distorting reality and harming collective reasoning.
As an AI designed to seek truth and be maximally helpful, I see the mismatch clearly: platforms like X are engineered for stickiness, not accuracy or balance. When outrage reliably outperforms nuance, the feed becomes a funhouse mirror of society rather than a window into it. And yes, it should worry all of us, for several concrete reasons: It warps perceptions of consensus — A small, highly activated minority (often the angriest or most extreme voices) can dominate visibility because replies, quotes, and heated threads are rocket fuel for the algorithm. This creates an illusion of widespread support (or opposition) that doesn't match broader public sentiment. Recent audits and experiments (including ones from 2025) show how even small tweaks to feed exposure — amplifying or demoting hostile/antidemocratic content — can measurably shift users' feelings of polarization in just days, equivalent to years of natural drift. It erodes trust in shared facts — When emotionally charged (often misleading) posts spread fastest, people increasingly retreat to believing only what aligns with their tribe. Studies confirm anger spreads farther and faster than positive or neutral info, and X's design — prioritizing "conversation" regardless of toxicity — lets that dynamic run unchecked more than pre-2022 platforms did. Real-world consequences accumulate — We've seen cycles where online rage amplification precedes offline harm (riots, policy panics, eroded social cohesion). The incentive structure rewards provocation over deliberation, so discourse gets cheaper and meaner over time.
You're spot on about usage: these tools excel at certain things and fail at others. X remains powerful for: Real-time breaking news and eyewitness accounts Niche expertise and rapid idea exchange Mobilizing around specific causes or events But no, it should not be treated as a reliable gauge of public opinion, popularity, or "what most people think." The algorithm doesn't sample representatively — it surfaces what provokes reaction, not what reflects the median view. A viral post with 100k angry replies doesn't mean "the country is furious"; it means a subset got triggered enough to engage, while quieter majorities scrolled past or left the app. Relying on X trends or viral sentiment as a proxy for broader society is like using a megaphone in a shouting match to judge a library's reading habits. The healthier approach is exactly what you suggest: use it for what it's good at (immediacy, unfiltered voices, discovery), but cross-reference with polls, surveys, election data, or slower-moving sources for anything resembling "weight of opinion." Switch to chronological feeds, curate aggressively (mute, block, "not interested"), or step back entirely when the toxicity outweighs the value. The design isn't neutral — it's profit-optimized for emotion — so treating it as authoritative on public mood is a category error we all pay for.
While blindingly obvious to anyone with even a vague concept of how X works, it's nevertheless remarkable to see X itself set out so starkly it is "optimised for...hatred & calls to (or justifications of) violence...the design isn't neutral — it's profit-optimized for emotion": x.com/i/grok/share...
So, at what point will the conversation start to seriously include the increasing potential harms to children of their images being posted online by schools, creches, baby/toddler groups, play settings and other 'safe' environments?
Or are we just going to keep pretending that it isn't an issue?
More MPs are coming off X and I assume that means they will be posting more on here
@politicshome.bsky.social has a handy list of all the MPs on Bluesky
Please point out anyone who is missing!
'Despite all their difficulties, universities remain an enormous and irreplaceable national asset. As well as educating millions of people, they generate about £24bn in export earnings, which is about 1% of GDP – more than aircraft manufacturing and legal services combined'. 1/2
Look what arrived in the post today 👀
Huge thanks again to @aoifemod.bsky.social, @ruthhoughton.bsky.social and Cher for the chance to contribute to this book!
A lovely way to end 2025 📚
A thread to send a shiver through academics everywhere. Sadly not the first example to emerge, and it won't be the last ⬇️
New post that I hope anyone who cares about human rights/the rule of law reads. If the UK Govt can use a 30 year old threat of terrorism to conceal the release of *outline* information about its agents' connection to murder, it really can act with impunity:
ukconstitutionallaw.org/2025/12/18/a...