When talking to @howardhliu.bsky.social about his new project, I remembered this great graphic made by the Stasi showing the social connections of a poet they were spying on projects.propublica.org/graphics/sta...

I will be presenting my first paper using data collected while incorporating genAI into my module on 15/10. If you can join us, I’d love to hear your thoughts 💭

Populist policymaking is not only about courts and media crackdowns. It can also hide in technical reforms. In Poland under PiS, changes to academic journal rankings were used to reward allies, promote ideology, and claim responsiveness. www.tandfonline.com/doi/full/10....

Next...🧵 Have you ever wondered why it’s so hard to compare public administrations across countries? This article tackles the confusion around Administrative Traditions

We've felt it and now we have the evidence: using genAI to write essays yields shallow encoding. Here, the subjects write essays using one of 3 tools (LLM, Search Engine, brain🧠 only) for 3 sessions. In the 4th session, LLM group needs to rely on 🧠 , and 🧠 group gets LLM. Key consequences for me:

We’re hiring 3 permanent (tenured) lecturers in the Department of Political Economy at King’s College London. These are ‘education pathway’ posts (in Economics, Research Methods, & Comparative Politics) 1/4 www.kcl.ac.uk/jobs/117268-... www.kcl.ac.uk/jobs/117209-... www.kcl.ac.uk/jobs/117245-...

The Anthropic Education Report (www.anthropic.com/news/anthrop...) offers a great insight into how students use Claude. The (relatively) good: half of the conversations are collaborative (dialogue with AI to achieve goals). The bad: students mostly outsource higher-order cognitive functions (see ⬇️)

Yesterday's session during the @aiatkings.bsky.social conference attracted a wonderfully engaged audience. In the slides, you’ll find the GPTs and exercises demonstrating applications of point 2 (below). Key points ⬇️

The state of research on AI and education from controlled studies: Growing evidence that, when used as a tutor with instructor guidance, AI seems to have quite significant positive effects. When used alone to get help with homework, it can act as shortcut that hurts learning. Still early days.

1/6 This is to share with you Misfortune Telling GPT (chatgpt.com/share/680102...) : a guided reflection tool modelled on the Stoic practice of praemeditatio malorum, to help you imagine and/or prepare for real democratic backsliding.

Last week, I joined Rozana Himaz, Ruth Dixon, @drjennings.bsky.social, and Colin Scott to pay tribute to Christopher Hood and his work: youtu.be/NA3vH8w1RUY. It was a real pleasure to hear memories and reflections from so many of his students, friends, and collaborators.

It's wonderful to see the news that my fantastic Department of Political Economy colleagues @maiaking.bsky.social and @bpiotrowska.bsky.social have been nominated for this esteemed prize. Huge congratulations guys!!!

📢 Out Now! V-Dem Dataset v15 and the V-dem Institute Democracy Report 2025 The Dataset: v-dem.net/data/the-v-d... DR25: "25 Years of Autocratization - Democracy Trumped?": v-dem.net/publications... #DR25 #DemocracyReport2025 #VDemDatasetV15 #polisky #PoliticalScienceData #PoliSciSky 🧵1(3)

What a nice recognition of all the hard work that went into this book. Especially thinking of Christopher, whose dedication and brilliance made this project happen—this is a lovely testament to his vision.

📢 Call for papers! We are organizing the 5th Early Career Workshop in Quantitative Political Economy on 19-21 May 2025 at King’s College London! Keynotes: Sonia Bhalotra (Warwick) and Peter van der Windt (NYU-Abu Dhabi). No fee, travel grants available! Submit at: tinyurl.com/qpeearly

What worries me is not government use of AI per se but the slightly credulous way that the government talk about it, which makes me suspect they will be taken for rubes by the tech companies selling their services. www.theguardian.com/politics/202...

Fact-checking wasn’t “biased” against conservatives. Conservatives just shared more false content. If there’s a sportsball game and one team fouls four times as much, it’s not “biased” for the ref to call four times as many fouls against that team.

Just learned that the great Christopher Hood, a professor of public management at the University of Oxford, has passed. Hard to overstate his influence on the field. At a personal level, both a gentleman and a gentle man. Here is a link to his most-cited work. onlinelibrary.wiley.com/doi/10.1111/...

tired: fearing that robots will become so smart they take your job wired: fearing that robots will still be dumb but take your job anyway

I (or my family) have definitely done it all 🐟

As a result of this event, I found out about Centre for Data Ethics and Innovation, which runs, among others,Public attitudes to data and AI tracker survey. Sharing in case useful www.gov.uk/government/o...

“You saw how the [Assad] regime raised them: they call us Sidi [‘my master,’]” he said. “You feel they are broken. [We just want them] to get out of that mindset. You’re a person with self-dignity, I’m not your master or anything. I am an employee, like you.” on.ft.com/4iyR4hJ

"The purpose of writing isn’t just to demonstrate knowledge in the most expedient way. Writing is to explore knowledge, to connect and synthesise ideas, to create new knowledge, and to share." An interesting analysis of pitfalls of drafting using genAI by @leonfurze.com leonfurze.com/2024/02/21/t...

An account of the "Time of Heroes" program, which focuses on embedding individuals with combat experience in Ukraine into Russian governance structures through leadership and management training. A selection of quotes ⬇️ on.ft.com/41nZ7re

In the US, in an information-poor environment (such as local elections), endorsement by various interest groups acts as a partisan cue. A pretty striking graph from onlinelibrary.wiley.com/doi/full/10.... by Jennifer Gaudette.

The latest two episodes of the EKS got me thinking about ideas for positive lessons from populist policymaking that can be adapted into liberal policymaking to achieve better outcomes: 1. Direct engagement 2. Simple communication 3. Focus on outcomes 4. Fostering identity and pride

Weak autocracies are far more common than strong autocracies. They are also often far more dangerous and damaging than strong autocracies. Never take hope from an autocrat’s weakness. Instead, prepare for the worst. A 🧵 based on 2 of my books: (1/N)

Korinek (2024) "Even in areas like religion and AI alignment, the unique value of human input may diminish as AI systems become better at understanding and replicating human values and experiences."

Are you struggling to turn complex information into a persuasive narrative? Introducing C.A.P.T.U.R.E., the 7-step framework we use at Exaltus to transform information overload into a captivating story. Over the next 9 posts, we’ll break down each step—let’s dive in! 1/9

A new addition to my case vault: The Swedish nobility's ⬇️ role in the public sector evolved gradually, with meritocracy solidifying in the 19th century due to welfare expansion rather than military needs. This shift, rooted in 17th-18th century state-building, varied by rank and sector.

Listened while waiting for @aspaglayan.bsky.social's book. Key: to understand education policy,understand politicians' incentives. Historically, this often meant ⬆️ education to ⬇️ the risk of disobedience against the state. Education became the third tool in the coerce, co-opt, indoctrinate toolbox.

Looking forward to reading this! I’ve included two papers by @aspaglayan.bsky.social in the education week of my Policymaking in Non-Democracies module, so it’ll be great to dive deeper into the topic.

Very interesting: AI & AI+human-graded essays result in similar improvements in final exam scores, possibly due to teachers compensating for AI-only systems' limitations by providing personalized support. Both systems increased the number of essays discussed individually (task composition effect).

Just watched a great video by @justinwolfers.bsky.social on the use of genAI in teaching econ. Highly recommended! TL;DR: Bad news: High-stakes, at-home assessment is dead for econ. Good news: Personalisation at scale with genAI improves learning at ⬇️ cost. My 3 reflections🧵 youtu.be/BvKgcoOO20c?...

ChatGPT created this Median Voter Theorem GIF for me, and I'm pretty pleased with it. Its original suggestions didn't do a good job conveying the MVT, but after I laid out the pictures I wanted, it was very helpful in generating this: #genAI #EduSky

Prepping for a pub quiz using chatGPT. My purported expertise: names of animal group. ChatGPT suggests learning using 1️⃣evocative mental imagery, 2️⃣alliterations, and 3️⃣rhymes. So far, so good! Let’s see how much more help I can get.

We noticed there wasn't a starter pack on authoritarian politics yet – Now there is! go.bsky.app/UFHn7sS Spread and ask to join!

Citizens hear about democratic backsliding on the news, but they meet it in interactions with street-level bureaucrats. This SI, co-edited with @gabilotta.bsky.social and Nadine Raaphorst, explores how populism and democratic erosion shape the work of the frontline public sector🧵

This contribution by Jonathan Cardoso-Silva led me to 3 lessons about essays as assessment: 1.Bump up originality as a grading criterion. 2.Use creative essay prompts. 3.When grading, pay esp attention to the content, don't be fooled by the beautiful language. More ⬇️ www.lse.ac.uk/Events/LSE-F...

I've just finished this course on using genAI to amplify critical thinking and it really is worth the 1hr it takes to complete. I'm already thinking of ways to apply the PIQPACC framework to my modules www.linkedin.com/learning/amp...

Two ideas for using genAI in teaching from B. Oakley's "Using generative AI to strengthen and speed learning" youtube.com/live/3i8G_9B...: 1. Create metaphors for concepts you are trying to teach 2. Generate hooks linking the lecture topic to something students care about

Just as I was writing a brief on the migration policy prospects under the new coalition government... notesfrompoland.com/2024/01/04/p...

Poland update: the transition of power to the democratic coalition should be complete by tomorrow. Today the PiS PM seeks the confidence of the parliament, which he is unlikely to get-->the coalition proposes & approves Tusk as the new PM -->(tomorrow afternoon) vote of confidence in the new govt.

Bear with me, while I’m pinching myself and working on a summary thread.

I’m delighted to work with the @apsgworkshop.bsky.social core team on new initiatives promoting research on authoritarian politics. If you haven’t checked out the workshop yet, you should do it now - the paper and discussion quality is fantastic!

This viz shows well the complexity of the task ahead of the PL coalition in working out its programme. Many of the election candidates are political alliances or outright coalitions of partiesthe opposition coalition would in fact be composed of 10, rather than just 3 parties.