Darren Dahly

A screenshot from Adobe Acrobat, where it is offering to prepare a summary of a PDF. The page from the PDF shown is a recipe for "Mixed up hot dog salad #1". The book is Mettje Roate's 1968 "The New Hot Dog Cookbook".

poisoning Adobe's AI assistant by having it summarize a 1968 cookbook that exclusively contains hot dog recipes

Sorry, I think my f**king eyes just rolled out of the back of my head.

Most social psychologists maybe.

Oops. Ooooooooooooops.

I do hope that nobody has been given or denied a job/promotion based on their SpringerNature citation counts in the past 15 years.

arxiv.org/pdf/2511.01675

h/t @nathlarigaldie.bsky.social

UCC Futures - Collective Social Futures Festival of Social Science 2025 | University College Cork Learn, Study and Research in UCC, Ireland's first 5 star university. Our tradition of independent thinking will prepare you for the world and the workplace in a vibrant, modern, green campus.

Do attend this week’s UCC Futures - Collective Social Futures Festival of Social Science 2025! You can find an overview of the social & research events here: www.ucc.ie/en/collectiv...

Even worse 😉

And to be blunt, a medic and an information systems expert is not a complete and sufficient set of experts needed for this kind of thing to work.

But that step is almost always skipped. So money was spent, timelines can't be met, and the hope is that someone with the needed expertise will rescue the effort after the fact.

This seems to describe at least a few "innovation" projects in the HSE.

And important people want the answers NOW. This has to be done NOW. But it can't be.

The solution of course would have been to consult with an expert at the start who would know how these data will need to be analyzed to meet the aims of the project, and how to format the data to facilitate this.

A hospital has implementation a digital health project. They want to evaluate whether the new digital system is less error prone than the old manual entry based system. They have collected data on the relative accuracy, but in such a borked format it will take ages to make sense of it.

And like most scam victims, the politicians and voters that fell for this one only want to forget it ever happened.

No industry is perfect. But the imbalance is hard to miss. One sector must prove its products are safe & effective. The other profits by avoiding that responsibility. If we care about transparency and accountability, scrutiny shouldn’t be one-sided. (Source: For The Greater Glory of Science) #EpiSky

bsky.app/profile/john...

Not this again

'however, standard culture methods did not yield readily cultivable microbiota.'

😉

lads i found ANOTHER Table of Impossible Summary Statistics in a "green economics" paper 🚩

Can you see the problem?

@mrjamesob.bsky.social

I've never been a fan of group work as a teacher or learner, but now I feel it's even trickier to manage as people have different interest in use of genAI to do assignments. It's a whole other thing that should in principle be negotiated, but how to manage this practically, and at what cost?

AI’s Modest Productivity Boost 3 years into widespread adoption, AI has delivered a ~0.4% productivity growth bump, putting it on pace with PC-scale lift at best (in ~1/2 the time) rather than radically reshaping the macroeconomy.

Just the blurb from this week's blog app piece penciling out the AI impact on productivity generated a robust discussion, so figured I'd make it my freebie of the week:

bobeunlimited.substack.com/publish/post...

The COVID inquiry conclusion that lockdown could have been avoided – and subsequent headlines – seems to have been based on a single opinion…

Here Today, Gone Tomorrow? Toward an Understanding of Fade-out in Early Childhood Education Programs John A. List & Haruka Uchida X LinkedIn Facebook Bluesky Threads Email Link Working Paper 33027 DOI 10.3386/w33027 Issue Date October 2024 An unsettling stylized fact is that decorated early childhood education programs improve cognitive skills in the short-term, but lose their efficacy after a few years. We implement a field experiment with two stages of randomization to explore the underpinnings of the fade-out effect. We first randomly assign preschool access to children, and then partner with the local school district to randomly assign the same children to classmates throughout elementary school. We find that the fade-out effect is critically-linked to the share of classroom peers assigned to preschool access—with enough treated peers the classic fade-out effect is muted. Our results highlight a paradoxical insight: while the fade-out effect has been viewed as a devastating critique of early childhood programs, our results highlight that fade-out is a key rationale for providing early education to all children. This is because human capital accumulation is inherently a social activity, leading early education programs to deliver their largest benefits at scale when everyone receives such programs.

"...the fade-out effect has been viewed as a devastating critique of early childhood programs, our results highlight that fade-out is a key rationale for providing early education to all... largest benefits at scale when everyone receives such programs"
#EconSky #PolicySky
www.nber.org/papers/w33027

On Environmental Democracy and the Collisons’ “Abundance-Verse” Andrew Jackson 21 November 2025 I. Introduction A clamour of praise greeted John Collison’s “how to get Ireland moving” Op-Ed following its publication in the Irish Times on 25th October: the Irish…

I’ve written something on John Collison’s Op-Ed and environmental democracy

Next time you’re out in Dublin, check if you’re being “socially de-risked” by a friendly face from his think-tank

elajucd.com/2025/11/21/o...

Fair point. The effect of removing coffee from coffee drinkers is plausibly different from getting people to start drinking coffee.

So there you have it, twin study estimates were greatly inflated, and molecular data sets the record straight. I walk through possible counter-arguments, but ultimately the uncomfortable truth is that genes contribute to traits much less than we always thought.

👊

And this isn't dramatics. Elon literally used his power to gain access and destroy USAID which has surely resulted in many many preventable deaths.

What do you call social scientists who work in the business school? Widely respected.

Microsoft has open sourced the Zork trilogy of text games Microsoft announced today that the code for Zork, Zork II and Zork III will be made available open source under the MIT License.

Microsoft has open sourced the Zork trilogy of text games - Now future generations can all be eaten by a grue www.engadget.com/gaming/micro... #zork #infocom #retrogamer #retrogaming #retrocomputing