Articles like the one discussed by @janerosenzweig.bsky.social and linked below (h/t @benpatrickwill.bsky.social) make me wonder how much #GenAI companies test their tools in real classrooms. Ed tech researchers and teachers know a lot about what does/not work - why not work with them?

Jane Rosenzweig@janerosenzweig.bsky.social · 13h ago

In my first-year college writing course about AI, I made a chatbot that could help students "stress test" their arguments by providing potential counterarguments. The class was about AI so it made sense to try it. It wasn't bad at the task but when my students saw that it didn't give answers and /1

“In terms of magnitude,” Lawrence Warren, one of the three Census Bureau economists who authored the paper, said of the employment and earnings declines, “we feel comfortable saying this is kind of in the same ballpark as the earnings effect of a Great Recession.”

The Most AI-Exposed College Graduates Are Entering Another Great Recession

A new Census Bureau report offers concrete evidence of how certain majors are faring in the job market — which may shift what students choose to study.

chronicle.com

This study shows the existing research on human-AI synergy likely underestimates its potential because the studies do not put humans in the best position to succeed. They provide recommendations for changing that, to which I would add the growing area of research on hybrid-human-AI collaboration.

Human learning is an understudied but promising lever for boosting human–AI synergy | PNAS

Humans collaborating with AI hold the promise of achieving superior outcomes compared to either acting alone (i.e., human–AI synergy). However, the...

doi.org

Good summary by @caseynewton.bsky.social that hits at a critical question: what are the compelling consumer use cases for #GenAI agents? I struggle with that. What do I do for myself now that an agent could do as well, and are the benefits worth the risks? 🤷 www.platformer.news/meta-muse-co...

Muse is having a moment

Are consumer agents the future — or a mirage?

platformer.news

New paper! Despite qualitative researchers claiming that generalizability is not a goal, their published work does indeed contain generalizable claims. We document this and provide recommendations for how all researchers can better calibrate their generalizability claims. doi.org/10.1177/2515...

Dear Publishers: Few people read printed journals any longer and vertically-oriented tables are a scourge. Please make all tables horizontal. Thank you for your attention to this matter!

Capisce Owl Tilts Head Curiously

ALT: Capisce Owl Tilts Head Curiously

static.klipy.com

It makes sense that when students try to engage agentically in science classrooms, how the teacher responds can affect students' motivation. But I haven't seen such detailed mixed methods analyses before. Student extrinsic motivation goes up when teachers do not support agentic engagement.

My department and college are searching for FOUR new tenure-track assistant professor colleagues! Please share, and application reviews begin October 15.

Sorry I am no longer an active user but just post this job alert: With Peter Dayan and Michiko Sakaki, we are looking for a postdoctoral researcher who could work on cognitive computational modeling wth digital education data. Come to Tübingen!!: uni-tuebingen.de/fakultaeten/...

Could not have asked for a better host at WashU and around St. Louis than Andy Butler. Great day at the workshop, fun meals with him and his advisees, a fascinating tour of the city, and to cap it off a St. Louis City SC game! Don't tell anybody, but St. Louis is awesome.

This review of developmental aspects of self-regulated learning is much-needed and welcome. I share the authors' concern that any apparent drop in SRL strategy use in adolescence is very likely due to measurement issues. We need better SRL measurement in longitudinal designs. doi.org/10.1016/j.le...

“I don't think this is really sustainable because of the efficiency gain” that AI offers, says Thom. Any mathematicians—especially early-career researchers—risk being “isolated” if they don’t use AI to accelerate their work, he adds. Open source run on prem, perhaps? www.wired.com/story/mathem...

Mathematicians Hate AI. They Can’t Quit It

Powerful AI models have created an existential risk to the field, but researchers can’t stop relying on them because they’re too useful.

wired.com

Hey look, it's me on why teachers still (and will always!) matter in the age of AI. youtu.be/eOlbeMU1j9w?...

Why Teachers Still Matter in the Age of AI

YouTube video by LEAD Graduate School & Research Network

youtu.be

This has some good suggestions about how to break the multitasking habit. It’s hard. Takes multiple strategies, done for long periods of time, with support. But it can be done. wapo.st/4y2grQ7

Analysis | ‘Toxic multitasking’ is breaking your brain. Here’s how to take control.

Doing more than one thing at once, especially on screens, can become addictive, which harms the functioning of our brain and nervous system.

wapo.st

It’s going to be tough to think about anything else today. www.nytimes.com/2026/09/20/m...

In Fighting for Every Black Child, Did I Betray My Own? (Gift Article)

A decade ago, I made a decision about my daughter’s education that I hoped would challenge New York City’s separate and unequal school system. Now 16, she thinks I made the wrong choice.

nytimes.com

I agree with most of this, and my disagreements are not substantive enough to deny endorsing it. I don't think a lot of the jump scares around #GenAI are warranted. But these systems can cause harm and need to be regulated by the people. "The nuclear analogy is useful because..."

Opinion | Humanity Has Avoided Apocalypse Before. Let’s Do It Again. (Gift Article)

It is up to all of us, through our government, to shape what A.I. becomes.

nytimes.com

💯 "We know what to do: incubate open-access A.I. models, regulate technologies before they go rogue, hold the people who allow them to go rogue accountable if they do and protect the most vulnerable when we, the people, sometimes get it wrong." www.nytimes.com/2026/09/19/o...

Opinion | A.I. Is a Threat, but Not in the Way You Think (Gift Article)

The threats from A.I. are real — and they’re ones we know how to deal with.

nytimes.com

Funny enough, the Four Advantages Mollick discusses here align very well with what I've been discussing in recent presentations on #GenAI. In short, GenAI is a tool that augments, but does not replace, human knowledge, expertise, and agency. Self-regulation is needed to leverage these advantages.

The Overhang

Using your deep knowledge, wide knowledge, taste, and agency

oneusefulthing.org

Such a joy and an honor to deliver a keynote for the Recentering Student Learning in the Age of AI symposium at Washington University in St. Louis Arts & Sciences! I was so impressed with how they are thinking carefully about how to integrate GenAI thoughtfully into their pedagogy...

So excited to be speaking at this event! My talk will focus on student agency as the key to any decisions we make about #GenAI in higher education. Whether we use it, ban it, or temper it, those decisions must start with: "What will help our students succeed, thrive, and realize their agency?"

Recentering Student Learning in the Age of AI

A half-day symposium for faculty, post-doctoral fellows, and graduate students on how to prioritize, support, and enhance student learning in a world increasingly shaped by AI.

artsci.washu.edu

So much this.

Stephen Wild@stephenjwild.bsky.social · 6d ago

Perhaps psychologists should spend more time developing and validating scales

Some scholars are going to use #GenAI in their workflow. This threatens to create a "scientific monoculture." These authors argue preventing scientific homogenization and exclusion requires people to view GenAI as a tool for epistemic reflexivity and as an object to be examined.

Excited to see the latest issue of Educational Psychologist published - the first by the new co-editors Krista Muis and Jason Chen! As always, a great set of articles covering procrastination, interleaving, mindsets, and a career award winner! www.tandfonline.com/toc/hedp20/c...

This is an *excellent* post on cluster analysis. If you ever thought about applying cluster, latent class or latent profile analysis to psychological scales, you absolutely should read this!>

Jan Vanhove@janhove.bsky.social · 6d ago

New blog post: "Cluster analysis: A skeptic’s guide". In which I ask a few questions that social scientists wishing to identify hidden classes in their data ought to address. janhove.github.io/posts/2026-0...

P.S. There is, strangely enough, the notion that cluster analysis is somehow more idiographic, not treating all people as one homogeneous blob. Folks!! Cluster analysis is about sorting people into groups, with group membership having vast implications. This ain't it.

Is it just me, or are other people noticing a striking rise in bad copyediting in published articles? Just read an article referring to the authors' "hypnotized" effects. As someone who often typos, I sincerely appreciate good copyediting. Bad copyediting makes the journal look bad.

Kermit The Frog: 'Mistakes Were Made'

ALT: Kermit The Frog: 'Mistakes Were Made'

static.klipy.com

How do students learn to take charge of their own learning? Drawing on the latest #cogsci research, the Science of Learning, 2nd Ed. breaks down 3 principles for how teachers can support students to learn more effectively and independently over time. 📄 www.deansforimpact.org/tools-and-re...

Okay wait. I thought #GenAI was going to become conscious, smarter than humans, and take over the world. Now the mice are, too?!?!? join1440.com?user_id=66c4...

So pleased to see this article led by Casper Juul published! He did fantastic work interviewing upper-secondary students about how they thought about knowledge across three different subjects, identifying five different onto-epistemic categories these students used to direct their learning.

Student conceptions of knowing: Onto-epistemic categories of epistemic performance in school subjects

Apt epistemic performance across formalized subject settings requires students to adapt their performance to the norms and practices of each domain. Such navigation plausibly entails the classifica...

doi.org

So many people think that the students who can't pay back their loans are going to like Vassar or Oberlin or even Harvard. In reality they're going to schools like Legends Barber College or Construction Training Center. www.npr.org/2026/09/16/n...