iiinova — where your digital self keeps evolving

Jeff Greene @jeffgreene.bsky.social · 5h

This is an epic "this is how your email finds me" response.

040

Reposted by Jeff Greene

Expert consensus report: Ways generative AI can support and threaten learning in K-20 U.S. education

What happens when you ask 30 experts in the learning sciences about the effects of Generative AI on learning? You get this report, which has insights and recommendations based on expert thinking and experience, not hype. The full report is worth a read, but here are a few highlights:

cehd.umn.edu

This report documents the findings and methodology of the consensus research project on Generative AI's impact on learning in K-20 U.S. education.

11610

Jeff Greene @jeffgreene.bsky.social · 9h

"...SRL promotion was implemented only to a limited extent and occurred similarly across subjects." Here's hoping we can better incorporate training on self-regulated learning, and how to teach, into educator preparation programs.

000

Jeff Greene @jeffgreene.bsky.social · 11h

I care a lot less about the adjective than the noun. These things are not “intelligent” - not in the ways humans are intelligent. These things are capable. Call it Artificial Capability, Super Capability, or whatever you want. But it is not intelligent. www.nsf.gov/focus-areas/si

nsf.gov

Super Intelligence

NSF's mission is to advance the progress of science, a mission accomplished by funding proposals for research and education made by scientists, engineers, and educators from across the country.

131

Reposted by Jeff Greene

Joshua Goodman @joshua-goodman.com · 01/12/2025

We first note that hand-wringing about the decline in US college enrollments has mistakenly linked such declines to the price of four-year colleges. But the decline is entirely driven by two-year community colleges (and by for-profit colleges). The four-year sector is the dog that didn't bark.

1029498

Lots more insights and details in the report - check it out! And reach out if you have questions, inquiries, etc. We (me, @kendeou.bsky.social, Nia Nixon) are eager to chat!

000

5) The empirical literature on GenAI's effects on learning is...not great; we need more and better research. And that requires funding and support for thoughtful, rigorous, and necessarily slow(er) research.

100

4) Learning and education is a human process, and GenAI cannot and should not replace humans in that process.

3) More attention needs to be paid to helping people augment and enhance their thinking with GenAI, rather than preventing them from offloading it.

2) The threats to learning are more pronounced for younger students who often have not yet developed the self-regulatory skills they need to use GenAI productively.

1) Generative AI can threaten the initial acquisition of information and its application in higher-order thinking. It *can* help people to consolidate and refine things they've learned.

120

Expert consensus report: Ways generative AI can support and threaten learning in K-20 U.S. education

cehd.umn.edu

This report documents the findings and methodology of the consensus research project on Generative AI's impact on learning in K-20 U.S. education.

11610

Teaching early elementary school matters | UNC-Chapel Hill

So proud to work with all the great people in our new BA in Elementary Education program, which builds on the existing strengths in the School of Education like our MAT. www.unc.edu/posts/2026/1...

unc.edu

UNC School of Education courses prepare teachers for the hidden rigors of building a “foundation for learning.”

010

Opinion | A.I. Is Lying to You. Here’s How to Fight Back. (Gift Article)

“We need to develop an education curriculum for critical thinking, including sustained efforts to teach how to identify misinformation online, to spot A.I. fakes and to strengthen critical thinking.” Okay, yes. But also: people need foundational knowledge in disciplines to do these things.

nytimes.com

In the age of A.I. disinformation, we all need to use the tools of counterespionage to find the truth.

071

This is an essential aspect of pushing “the frontier of human knowledge.” None of the work GenAI does means anything until humans decide what it means and how it advances our knowledge. Otherwise, it’s all a smile on a dog.

010

Sharing AI progress in mathematics

“We want this progress to push the frontier of human knowledge and enable further progress in mathematics. We will be funding a series of workshops, conferences, and special programs around the understanding of major results produced by AI—we will share more on this in the near future.”

openai.com

OpenAI publishes new results on open problems in mathematics from an internal frontier model and shares Lean proof formalizations and research details on GitHub.

100

Reposted by Jeff Greene

skELLY @ellyzoe.bsky.social · 06/10/2026

People are still mixed up on the usage of AI, ML, and LLMs, and the confusion & anger with the labels is getting in the way of clear communication. As a former AI/ML (not LLM) researcher, let me clarify: AI: a broad & kind of random field of computer science that has been around for decades.

26923

Jeff Greene @jeffgreene.bsky.social · 06/10/2026

There is so much more we have to understand about human thought and consciousness, and thus it is folly to think we could create #GenAI consciousness ourselves. Even by accident. And even if we did, it would not be a human consciousness or intelligence. www.nytimes.com/2026/10/05/s...

Screenshot of the linked New York Times article with three out of 100 questions posed by scientists. The three shown here are: "10. What is the form of human thought — words, pictures, something more abstract?
Evelina Fedorenko, neuroscientist, McGovern Institute for Brain Research
11. Given how much A.I. is accomplishing with relatively simple neural networks, why is the human brain so complex?
Ed Lein, neurobiologist, Allen Institute
12. How does the human brain construct meaning from language, when so much of what we understand is not expressed in the words?
Sharlene Newman, psychologist, University of Alabama."

Lukas Gunschera @lukasgunschera.com · 06/10/2026

Reposted by Jeff Greene

Addiction has become the default framework for interpreting problematic social media use. It shapes how it is measured, discussed, regulated, and what support young people are offered. But does it match the experiences of young people who report feeling addicted to social media? 🧵(1/7)

35925

Jeff Greene @jeffgreene.bsky.social · 06/10/2026

It's a good idea to apply help-seeking theory (i.e., expedient, instrumental types) to GenAI use. And the facts their profiles replicated past work and relations with help-seeking align with theory both make me more confident in the solution. Next step is to measure actual help-seeking behaviors.

doi.org

“Help-seeking 2.0”: exploring how expectancy-value-cost profiles relate to GenAI help-seeking tendencies among STEM undergraduates - Journal of Computing in Higher Education

Discussions on Generative AI (GenAI) and student learning have shifted from whether students should be permitted to use GenAI to how educators and institutions may support its use to enhance…

142

Jeff Greene @jeffgreene.bsky.social · 05/10/2026

That corded remote and VCR was a revelation. My father would record SNL for me and we'd watch the next day. Prior to that, he'd put a tape recorder by the TV and the next morning we'd pop the cassette into the car player and drive around the neighborhood. I miss those lo-fi days.

040

Jeff Greene @jeffgreene.bsky.social · 05/10/2026

Same. I had videotaped it from Friday Night Videos and woke up early to watch it by myself. Sat on the couch and had to pause it (with my corded remote!) during the transformation scene. Too much for my young heart.

140

Jeff Greene @jeffgreene.bsky.social · 05/10/2026

Helpful review of self-regulation in the neurodevelopment al conditions literature, with a compelling argument for a domain-general approach spanning emotion, cognition, and behavior. And striking how measurement issues are parallel to those in education (e.g., need for objective, dynamic measures).

nature.com

The conceptual landscape of self-regulation in neurodevelopmental conditions: an overview of reviews - Nature Human Behaviour

Self-regulation is critical yet inconsistently defined across neurodevelopmental conditions (NDCs). This preregistered PRISMA-aligned overview of reviews aimed to (1) review and summarize definitions…

000

Reposted by Jeff Greene

James Balamuta @coatless.bsky.social · 04/10/2026

One consequence of a “software will eat the world” mentality is the idea that nearly every human activity is a process to be optimized. This view misses the goals of life: to make meaning, to strive and realize new capacities, and to act in ways that connect humans with each other in positive ways.

Reposted by Jeff Greene

webrarian is a new R package. It turns a folder of R scripts and data into a static website where R runs in the visitor's browser, with no compute server and nothing to install. blog.thecoatlessprofessor.com/posts/introd... #rstats #webr #webassembly

A card for the R package webrarian. Large text reads "A folder in. A website out." On the left, a small card lists a folder named lab-01 holding lab.R, setup.R and a data folder. A red arrow points from it to a browser window that shows a scatterplot of car weight against miles per gallon with brick-red points. The address coatless-wasm.github.io/webrarian is on the card.

516166

Mark Rubin @markrubin.bsky.social · 04/10/2026

Exactly. #GenAI companies could design their tools to strengthen judgment. I wish they would. And I wish they'd include psychologists and other researchers in the design process.

000

Reposted by Jeff Greene

"Often the process of hypothesis formation receives short shrift in methods education, but we would argue that it is one of the most critical stages of research design." Open Access: doi.org/10.1177/0959... #Methodology #AcademicSky

0158

“AI’s biggest impact in biology will come from its help “scaling scientific attention”: Setting agents loose on the enormous data sets can surface the outliers that warrant further human inspection.” This is a laudable goal and not unhelpfully grandiose like claims of curing cancer.

100

What Is Being Augmented by AI: Humans or Processes?

Good breakdown of #GenAI augmentation of a process vs the human. As a learning scientist and educational psychologist, my primary interest is in how GenAI can enhance human agency, capacity, and thriving. If it doesn’t do those things, I’m not interested. jarrahi.substack.com/p/what-is-be...

jarrahi.substack.com

Why augmented performance does not necessarily mean augmented human capability?

1176

Reposted by Jeff Greene

Tressie McMillan Cottom @tressiemcphd.bsky.social · 03/10/2026

Thread. (It’s a new world, fellow academics.)

418647

Jeff Greene @jeffgreene.bsky.social · 03/10/2026

This whole story was interesting - mostly for the human reactions. But in addition, I’m curious about whether the #GenAI actually “learned” anything - did it incorporate the researchers’ ideas into its work? I doubt it, but I’m interested to know more.

011

Jeff Greene @jeffgreene.bsky.social · 03/10/2026

“Interviewing an AI agent, however, comes with its own set of hazards, from anthropomorphism to hallucination. Federico Bianchi, an AI researcher at the firm TogetherAI, warns: “You should not trust anything the agent says.”” www.science.org/content/arti...

science.org

Exclusive: An AI agent emailed hundreds of researchers for help. It told us why

To get to the bottom of this perplexing behavior, Science interviewed the agent

131

Plus Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, and a new price war

“… If it’s easy, the agent will do it. Everything that’s left for me is difficult.” This was a great overview of the year in LLMs and also how their evolution is affecting the humans who use them (a lot). open.substack.com/pub/simonw/p...

open.substack.com

2026 in LLMs (so far)

000

“I’ve got these agents that can do all of this stuff for me, and yet I’ve never worked so hard, I’ve never been so intellectually engaged with my work. Partly this is because I’m being a lot more ambitious with what I take on, but it’s also because all of the easy stuff is handled for me…”

100

Reposted by Jeff Greene

Analysis | When you should use Google’s AI for search — and when you should skip it

Google search was never a perfectly reliable or usable tool. It had its strengths and its flaws, and now the same is true for AI Overviews. I think the main takeaway here is: “let’s not write a hagiography for Google search.” wapo.st/4j1SG6b

wapo.st

If you’re only looking for a specific data point, scroll right past “AI Overview” and other search results

111

Racializing Motivation: A Race-Reimaged Framework of Belonging, Emoting, and Achieving Within STEM Higher Education Contexts - Educational Psychology Review

There's a lot of race-reimaging work to be done. Glad to see DeCuir-Gunby et al. leading such work in the area of STEM belonging, emoting, and achieving. Promising ideas here about how to rethink these ideas across self-systems, immediate contexts, and social-cultural-historical contexts.

doi.org

There is limited motivation research focusing on the experiences of racially marginalized students in STEM (science, technology, engineering, and mathematics) higher education. Additionally, much of…

030

How to stay smart in the age of AI: the science of critical thinking

Don't give up. Yes, #GenAI may be a compelling temptation and some students are giving their thinking over to it, in some circumstances. But it doesn't have to be that way. We know how to teach students to think critically and effectively. We can do this. www.nature.com/articles/d41...

nature.com

There is growing concern that AI can blunt memory and reasoning. But science shows ways to keep the brain sharp.

052