Barbara Piotrowska

You Know Who Else Collected Metadata? The Stasi. Files obtained from the archives of the East German secret police show how the Stasi mapped social networks.

You can read more about the context and other information collected by the secret police, as well as get a flavour for their operation here: www.propublica.org/article/how-...

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 💭

Multi-purpose populist policymaking in practice: the Polish academic evaluation reform This paper theorizes populist policymaking as multi-purpose governance that can simultaneously deliver responsiveness, ideological consolidation, and institutionalized patronage. We apply this fram...

Read the full paper: "Multi-purpose populist policymaking in practice: the Polish academic evaluation reform"(Piotrowska, Szkurłat, Szydłowska, 2025).
www.tandfonline.com/doi/full/10....

This shows how populist reforms can be formally legal, yet designed to entrench ideology and institutionalise patronage. Higher education policy, often overlooked, is a crucial site of populist governance: shaping elites, promoting ideology, and redistributing resources.

But most of all, 3️⃣Catholic-affiliated journals gained disproportionately, embedding a conservative-nationalist agenda and rewarding pro-government institutions.

1️⃣Humanities and 2️⃣Polish-published journals were given a boost – framed as responsiveness to scholars worried about internationalisation.

The reform of Poland’s academic evaluation system looked like a technical adjustment. In practice, it reshaped how prestige and funding were distributed across universities and disciplines.

Together with Izabela Szkurłat and Magdalena Szydłowska, we argue that populist policymaking is multi-purpose governance. A single reform can:
➡️appear responsive to “the people”
➡️embed ideological content
➡️channel resources to loyal actors

Multi-purpose populist policymaking in practice: the Polish academic evaluation reform This paper theorizes populist policymaking as multi-purpose governance that can simultaneously deliver responsiveness, ideological consolidation, and institutionalized patronage. We apply this fram...

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....

journals.sagepub.com/doi/abs/10.1...

Next...🧵

Have you ever wondered why it’s so hard to compare public administrations across countries?

This article tackles the confusion around Administrative Traditions

3️⃣AI detection - LLM group displayed significantly lower ability to quote one own's essay - AI use detection could leverage this by randomly sampling students for oral exams discussing the essay. Alternatively, only suspect cases could be subject to such an exam.

In my module, I do that by asking the students to complete independent research first and only then engage with the genAI task.

2️⃣ When integrating LLMs into teaching activities, start with cognitive effort, followed by the use of genAI - "Brain-to-LLM participants could leverage tools more strategically, resulting in stronger performance and more cohesive neural signatures".

1️⃣ Blank page syndrome - LLMs are often seen as an antidote to it. But the paper finds that "the LLM group produced statistically homogeneous essays within each topic". Hence, if you want your piece to stand out, using LLMs to fight the syndrome might not be the best idea.

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:

AEP Lecturer in Politics and Research Methods

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 ⬇️)

I’m not 100% sure that this won’t actually improve his chances among voters of other far right candidates.

Day 2 is over, but there are three more to go: www.kcl.ac.uk/ai/assets/pd...

3️⃣ However, if you lack the motivation to learn, no amount of technology will help - and replacing cognitive effort will lead to skill atrophy.

1️⃣ Learning requires thinking deeply (exerting effort), over extended periods, and at regular intervals.
2️⃣ GenAI can support this by boosting motivation to exert effort (e.g., through gamification) or by adjusting difficulty to an appropriate level (e.g., customised explanations).

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.

en.wikipedia.org/wiki/Detenti...

6/6 I hope that you find it helpful. Please share any feedback on how your interaction could be improved.

5/6 Each deepened scenario is narrated immersively, and then reflection prompts are presented one at a time, for example:
• What values guide you?
• What steps could you take?
• Who could you lean on?
After each: type Done to move on, Develop to dive deeper, or Another scenario to switch paths.

4/6 You can share personal context (e.g., age, occupation, region, dependents) for tailored scenarios, or go anonymous and get a generalised one. I find that adding at least your broad job info makes it feel much more realistic.

Opinion | The Emergency Is Here Trump is disappearing people to a Salvadoran prison for terrorists. And he says he wants to send “homegrown” Americans there next.

3/6 I originally built the GPT as a novelty after my students enjoyed the text‑based games in my Policymaking in Non‑Democracies module. After hearing the latest Ezra Klein podcast (www.nytimes.com/2025/04/17/o...), I realised it could be useful more broadly.