BANGARANGA 💪

I should visit Moldova

The UK song is like an LLM’s interpretation of “90s blur meets ok computer” gone horribly wrong

BANGARANGA

Eurovision thoughts: 1. Australia should simply be forbidden (also in sports) 2. V good lecture on measurement 3. Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto Ferto

If writing a Guardian oped hasn’t saved his Premiership, maybe Starmer should start a Substack (free at first, ofc)

I think there is great value in loudly, clearly, and publicly modelling for grad students (and colleagues) what good practice looks like. And modelling that bad practice is bad, and should be shunned and criticised.

I think you’ll find Keir Starmer’s massive mandate totals at 18,884 votes. Parliamentary systems…

Even the economists are getting radicalised

MIT has lost 500 international grad students relative to last year. Caused directly by the White House. Closing America's doors to the very best, most brilliant people on the planet is like dropping a bomb directly on America's capacity to innovate. On the prosperity of our children's future.

Has anyone come across some writing - not necessarily anything formal - that they find particularly persuasive/informatice/thoughtful around human subjects ethics and AI? NOT professional ethics.

Gonna to say that I think the disease here is worse than the symptom. Hallucinated citations are bad and real and should have *real damaging consequences* for the reputations of authors who include them. BUT they aren't really a problem in and of themselves. Easy to see & correct, not damaging.

I have been using agentic AI for about 8 months now, quite intently for the last 4.* Today I taught an introductory workshop for MSc and PhD students at LSE (and beyond). * Check the last post for the footnote

I think Labour doesn't understand that while people like me - pretty moderate - really like some of what they've done, the extremist pandering on trans rights and immigration is not just a minus point, it's entirely alienating and unacceptable. And for every me they lose, they've won 0 reformers.

Pro tip: Don't inhale chilli crisp

What’s the bigger indictment of current Labour? - that Starmer is still PM - that Wes Streeting is his biggest rival

I guess winter is back

This is a great use case for agentic AI improving your workflow -- put your PAP into the project folder from the start, and have your agent keep a running tally of deviations, which you package as a neat table in the appendix.

AJPS Editors Blog Post: Pre-Analysis Plans (PAP) Policy Change. We recently clarified our policy regarding the execution of pre-analysis plans and the reporting of supplemental materials. These changes are in line with increasing transparency in reporting ajps.org/2026/05/13/p...

Good thread, but I think the explanation has a deeper cultural root. There is broad cultural disparagement of academia in the UK. “Those who can, do, those who can’t, teach” etc. That we work hard, contribute to GDP, and are the UK’s most successful export is simply an ontological impossibility.

Btw, people outside academia think we are insane when you tell them about the journal system and the importance of prestige journals.

Excited to share an updated paper from a longstanding agenda with Andy Hall, @dthompson.bsky.social, and Jesse Yoder: "Fortunate Sons: Elite Political Selection in American History" We link Members of Congress (+ their families) to the complete count census, covering congresses from 1850-the 2000s

Join us Wednesday at 1:30 PM (ET)! Navigating the non-academic #jobmarket as a political science graduate student or PhD can require a different approach than #academia. "Pursuing Non-Academic Careers with a Political Science PhD" Register here to join us!

I feel a lot of David Attenborough's hard work with nature preservation is to cover for how much of a mess his brother made of that dinosaur theme park.

Gen AI image from the prompt “crickets” LLM comedy or critical failure mode?

the only top political science pub that actually matters is Miller’s, just outside the Palmer House

You actually just don't ever have to feed the top-journal-academic-prestige psychosis machine. You can be your own scholar, kids.

Reminds me of the time I wrote an oped in the guardian after a stinging desk rejection from AJPS

A reminder that applications for this 3-year fully funded PhD position close at the end of next week (May 17). This is a great opportunity to work on climate politics comparative politics in a vibrant environment, with loads of freedom to develop your own project. www.sciencespo.fr/centre-etude...

Didn’t an entire discipline once do this

Don't forget to sign up for our final Seminar Series of the year, with Professor @adamberinsky.bsky.social 💡 Rethinking Misinformation Interventions: Beyond the Search for a Magic Bullet 🗓️ Friday 8 May ⏰️ 4.30pm - 5.30pm Join here👉 www.eventbrite.co.uk/e/department...

We’re hiring two LSE Fellows in Public Policy A two year career development role for early career scholars to teach on the MPP and contribute to LSE’s research community 📅 Deadline: 21 May Apply: https://www.jobs.lse.ac.uk #LSEPublicPolicy

FWIW, I don’t think the major problem in my field, or related ones, is that people need to move faster & produce more. I read a lot of papers as part of the job, and I’m increasingly convinced we’d benefit from fewer papers and better ones, not more mediocre work produced ever more “efficiently.”

Prime with ads. Just how Kubrick intended.

Love Clarence Seedorf as a talking head. Every time he says "visualize it" I visualize an AJPS. Hasn't worked yet but I like the idea.

Anyone else’s kids just entirely and relentlessly anti movie?

Maybe it was telling me to go to bed

Opus 4.7 just used the term “de-significates”

My Spirit Airlines memory: First year PhD. Friend and I book Spirit to Chicago for MPSA. Arrive at gate with backpacks and poster tubes. They want to charge us for the poster tubes as extra carry-ons. My friend gets out some tape and tapes the tubes together. We only pay for one extra item. Legend.

NorCal, Central Valley, SoCal

Pretty thrilled to host @adamberinsky.bsky.social at @lsemethodology.bsky.social next Friday. Please feel free to join us (get your tickets at the link) to learn about misinformation.

Yay, another paper by Berna and Erkan to dismantle wishful thinking about how clear-cut replicability is as a signal of good science.

Visiting Germany from London and I wonder how a society can be so prosperous with such high water pressure, such vigorously flushing toilets, and such effectively shutting windows.

Does anyone know of smart writing/thinking about why LLMs are so bad at writing (and why they seem to have gotten worse, if they have)? As in, is it a training/reinforcement issue, or is it something to do with the act of writing (which is maybe best when it deviates from next token prediction)?

If Labour MPs want to construct an electorate that likes them they’ll simply need to disenfranchise everyone

math department that stops offering its probability theory course because pdfs are notoriously problematic for meeting title II accessibility standards

Some of you are out there getting real R&Rs while the rest of us are like

bank holiday sorry