Clay Shirky's piece in the Chronicle aptly addresses the wicked problem of AI in higher education. Instead of a solution, he revitalizes what campuses do best-face disagreement and uncertainty. My main criticism is we have stopped managing these challenges and instead endure them.

Opinion | The University Was Made for the Age of AI

Colleges have never been coherent. Their strength lies in managing disagreement and uncertainty.

chronicle.com

Why isn't the Dept. of Edu calling on AI developers to ensure their tools "incentivize rigor in the age of AI," instead of putting that onus on campuses? The power imbalance is shocking and such a move will surely pass responsibility to faculty in the classroom. www.ed.gov/media/docume...

I made Reading Faulkner in the Age of Thinking Machines as a gambit of sorts—could I make something using AI that students would want to engage with and read and not offload to AI? open.substack.com/pub/marcwatk...

Reading In An Age Of Thinking Machines

I’m releasing a project I created to help my online students read, with the hope that others might be able to use, adapt, or otherwise feel inspired to try something new this coming semester.

open.substack.com

I created an illustrated version of Faulkner's As I Lay Dying. 59 image plates, interactive student reading experience w/ annotations and voice recording students can create/share with their professor, downloadable PDFs and ePub. It's a whole new ball game for digital humanities!

Confronted with a rapidly evolving technology, many educators have little precedent to draw on when creating classroom-level AI policy. For @edutopia.org, I asked 8 educators across disciplines/ grade levels how they establish shared values + expectations about AI use in their classrooms:

8 Classroom AI Policies Developed by Teachers

Setting clear, shared expectations around student AI use is uncharted territory for many teachers. Educators across disciplines and grade levels share their current policies.

edutopia.org

Fable does an excellent job of creating physics for an interactive 3D board with connected string and motion. Very interesting to see how it reacts to implied gravity. For this demo, each note card is a nested site I developed, so it serves as a gallery of experiments. Bonus Always Sunny meme

One thing I'm trying to do with my students this fall is focus on local topics that are meaningful to them, like the cost of housing on campus. I used Claude + real sources to create an interactive city builder, including reflective elements, and a survey about housing for their peers.

It's a trope at this point, one wearing thin on readers. There's little nuance or variation within these articles. I'll also add that it is a shrewd decision that goes beyond fetish to drive clicks and likes across socials. Provocations should do more than decry change.

Timothy Noah@timothynoah.bsky.social · 4w ago

In my latest, I deplore the fetish for “End Of” pieces in the Atlantic that pander to an older readership. newrepublic.com/article/2129...

App smashing will likely become the norm for AI usage. I designed a prototype of a creepy carnival midway about how higher edu is failing to deal with AI. 1) GPT 5.5 for design/site layout images 2) Gemini 3.1 image to video 3) Fable 5 to code the site Lots of possibilities to consider here!

Analog learning is a response to AI that education should consider. But if it becomes just another way to police student AI use, it will fail. The better goal is creating intentional spaces to help students understand when AI supports their learning or thinking, and when it quietly replaces it.

Analog Is Having a Moment. But Is It Sustainable?

The growing backlash against AI-saturated classrooms has created a renewed sense of exploration into analog learning methods.

open.substack.com

Claude Code can make a static image interactive. In this case, I had Gemini sketch a reduced distraction classroom, where AI is limited (I know, the irony!) The mouse snaps to areas, zooms in, and guides the viewer around the room. Really simple use case that makes a design pop.

Is it cheating if AI guesses the answers to a test? Students upload course materials to ChatGPT and create a practice test to study. The concepts and themes of the actual questions/answers are nearly identical to the actual test. Proctored assessment isn't a moat against AI.

It’s launch week! The Norton Guide to AI-Aware Teaching ebook publishes tomorrow! To celebrate, today’s podcast features an interview with my co-authors @anetv.bsky.social, @marcwatkins.bsky.social, and me, with special guest interviewer Stacey Johnson. www.buzzsprout.com/2069949/epis...

Come read along with us! The ebook of the The Norton Guide to AI-Aware Teaching launches 7/1 and you can join 500+ educators for Perusall's 4-week virtual book club. www.perusall.com/events/norto...

There are useful things you can create with AI code that function without AI and allow students to think about their values before turning to a chatbot. One example: I adapted the Pause Before Your Prompt activity from BU for my online classes, to include reflection and a visual PDF export.

We’ve talked extensively in education about the harms and impact of AI through critical AI literacy and how to use AI responsibly through AI fluency, but it feels like few people have had the opportunity to explore the creative possibilities of using these frontier models to code

Context Is All You Need

Why creative AI use takes judgment, taste, and a willingness to be inefficient.

open.substack.com

Granta ending their partnership with the Commonwealth Prize doesn't solve the issue of AI in publishing. Editors need to become familiar with how AI models generate prose, classifiers (AKA AI detectors), and examine proof of authorship techniques and tools. AI isn't going away in publishing.

Granta stops publishing short story award winners over AI controversy

Literary magazine will no longer engage in ‘external publishing partnerships’ after Commonwealth prize furore

theguardian.com

What does AI-aware teaching mean? Here's our full conversation with the authors of a new book on the topic, @derekbruff.bsky.social @marcwatkins.bsky.social and Annette Vee: youtu.be/XNNWURc24RY

Honestly, a six month pause in AI deployments would be welcome for nearly everyone in edu. Most campuses haven't processed GPT-4, let alone Fable. Having time to get our bearings and figure out what path we'd like chart is far more valuable than a new model or feature set dropping every few months.

The Fable of AI in Education

What I built with Claude 5 before it disappeared and why almost no campus can afford to build the same.

open.substack.com

Something to consider in light of Fable and other resource intensive frontier models— will any firm even have the capacity to run advanced models for hundreds of millions of users? www.theregister.com/on-prem/2026...

Only half of US datacenter capacity planned for 2026 is actually under construction

Another fun example of AI hype and reality colliding

theregister.com

I built some impressive applications using Claude’s Fable, but most campuses won’t have access to these features. If we want faculty and students to have creative opportunities with AI beyond text, we need to consider who will have access to advanced AI models. open.substack.com/pub/marcwatk...

The Fable of AI in Education

What I built with Claude 5 before it disappeared and why almost no campus can afford to build the same.

open.substack.com

Trump really doesn't like immigrants using AI. Jesus this is bad.

The divide between those who don't use AI and those that use the mostly free models deepen when affluent students or well-resourced faculty gain far greater access to AI than their peers. Case in point, I couldn't even make a simple deck using Claude 5, editing w/sonnet, opus @ $20. I'd need $100!

Anthropic is retaining user data for 30 days with Mythos, supposedly for safety review, including on enterprise plans with zero data retention. This may be a nonstarter for many campuses and industries.

We shouldn't destroy ourselves in arguments about AI. campuses need to focus on conversations that allow faculty to form shared values and shape what role, if any, AI should have in education. open.substack.com/pub/marcwatk...

We Shouldn't Destroy Ourselves Fighting About AI

How we talk to one another about AI needs to improve.

open.substack.com

In today's Tea for Teaching podcast, @anetv.bsky.social , @marcwatkins.bsky.social , and @derekbruff.bsky.social join us to discuss their forthcoming book, The Norton Guide to AI-Aware Teaching. teaforteaching.podbean.com/e/ai-aware-t... @cyberthread.bsky.social

It was wonderful spending time with @anetv.bsky.social & @derekbruff.bsky.social talking about AI-Aware teaching on The Tea for Teaching Podcast. We need to start having serious and sustained conversations across campuses about what role AI should have in classrooms and teaching

Having critical and thoughtful discussions with students about their AI use and AI’s impact on their community creates the type of creative and curious responses for students to learn about this technology without asking them to use a chatbot. open.substack.com/pub/marcwatk...

The Norton Guide To AI-Aware Teaching & Using AI As Provocative Pedagogy

Most writing about AI in education treats it as a problem to manage.

open.substack.com

A foundation model like Claude or ChatGPT IS NOT trained as a text classifier (aka an AI detector). PLEASE STOP DOING THIS! These are two entirely different AI systems (generative vs. classifier). This is why AI literacy matters. People need to understand how these tools work.

The Granta AI scandal grows. Several other stories may have been AI generated. A single paragraph at the end of the award announcement says Commonwealth is investigating. No response from Granta. Even the writing within the award announcement reads like it is AI generated. What are we doing here?