- On EdTech Newsletter
- Posts
- Interesting Reads This Week
Interesting Reads This Week
Simple yet important ideas

[ed. This week got away from us due to two consulting projects, meaning that we missed posting both free and premium newsletters. Mea culpa, and we are sharing this post outside of the paywall as partial penance.]
Sign up for the On EdTech newsletter. Interested in additional analysis? Try with our 30-day free trial and Upgrade to the On EdTech+ newsletter.
I will be speaking about the future of online learning at the Digital Universities US conference in Salt Lake City on June 9-11. If you are going to be there, let me know and let’s make a plan to meet up, perhaps for a dirty soda.
It has been one of those weeks. While the holiday in the US on Monday was very welcome, it left me feeling out of sync all week, like I was always a day behind on everything. But what did I read?
The third rail
Much of the good stuff I read this week seemed to center on insights that many of us already know intuitively but that aren’t always stated outright. These are important ideas that need to be expressed and clarified. One such piece, the first part of which I wish I had written, explores how the actual courses themselves are often overlooked in discussions of student success.
In Edcarta, Nicole Poss addresses what she calls “The Curriculum Gap That No One Owns.” She makes the important point that students rarely leave because of a single factor.
When students leave, it’s rarely a decision made in one moment. It’s the result of a slow accumulation of misaligned expectations and frustrating experiences
Recognizing this, colleges and universities have developed a range of efforts to address these problems.
Walk into any campus meeting on student retention, and you’ll find a full table. The advising team is there, sharing updates on early alert systems, call sequences, and plans to hire more staff. Student affairs is talking through expanded mental health support, belonging initiatives, and career services departments, which are trying to meet students where they are. Marketing and enrollment are reporting on outreach numbers and campaign performance. Institutional research is presenting data on stop-out patterns and persistence rates. The provost is often right in the middle of it all, nodding thoughtfully and jotting down action items. Everyone is working hard. Everyone is doing something.
And yet, students are still leaving.
A big part of this problem is that the curriculum itself—the content of courses and programs and how they are structured—is often not addressed. The people who support teaching and learning on campus, such as staff in instructional technology units, teaching centers, and libraries, are present but typically lack control over the content or curriculum, and don’t have the authority to act when they see a problem.
As a result, these issues often go unaddressed, leaving students disappointed and frustrated by the subpar experiences they sometimes encounter in the classroom. In some cases, these problems compound. In the worst cases, they contribute to or even cause students to drop out.
Poff suggests a range of actions that institutions can take to address this issue. This is where she and I part ways. While I agree with her general call to action, I believe that in most cases, the list of actions (below) will be far more difficult than she makes it out to be.
* Advising teams can flag patterns in student conversations about course-related friction. When advisors start hearing the same feedback about a class, that data can be tracked, shared, and addressed.
* Student affairs can surface co-curricular gaps that point to instructional issues. If students are constantly seeking tutoring for the same course, or if student orgs are filling in content knowledge that should be taught in class, that’s a curriculum signal worth flagging.
* Institutional research can disaggregate retention data by course, by section, by delivery method. They can help identify if certain instructors, formats, or course sequences are consistently leading to dropouts.
* Curriculum teams can be empowered to do more than project manage. They can become stewards of the student experience, looking at quality across modalities, tracking alignment, and proactively surfacing issues.
* Faculty and program chairs can be asked to review student performance and withdrawal data alongside their course content. Instead of focusing only on academic freedom, we can focus on academic responsibility.
* And provosts can lead by rebalancing ownership. If retention is everyone's job, then curriculum, arguably the single most influential factor, must be included in the strategy, not just the schedule.
While I increasingly see schools addressing issues like DFW rates, I think that most of these actions are non-starters at most institutions. Especially, but not exclusively, for on-campus courses, because they touch that academic third rail: faculty control over their own courses.
Maybe I’m too cynical. Maybe I spent too long at an institution where even requiring faculty to have a course syllabus was considered a violation of academic freedom while firing someone for the content of their social media posts was perfectly acceptable. Still, I think most of the actions outlined above will too often be non-starters.
Institutions can begin by collecting the data that Poff describes, and they can start by using persuasion and the power of example to improve curriculum structure and content. Most faculty take pride in their teaching and want to do it well, with good outcomes. But for the most part, you can’t mandate it.
Not the Plagiarism Panic Post
Over in Australia at Reflexive Machine, Luis Lozano-Paredes tackles that viral New York Magazine article about student cheating titled “Everyone Is Cheating Their Way Through College.” He makes the valuable point that if everyone is using generative AI to cheat, maybe the problem lies less with the students and the technology and more with the assignments themselves.
He suggests a range of alternatives to help start rethinking assessment.
1. Process‑over‑product Assess version history, peer critique notes, and oral walkthroughs, not just the polished PDF.
2. Synchronous, social learning Denmark’s model reminds us to allocate daylight hours to messy collaboration and push solitary drafting (where AI shines) outside the classroom.
3. AI as witness, not ghost‑writer Students document how the model was used, what they kept, what they overrode, and why.
4. Micro‑vivas & design juries Short, recorded defences force articulation of reasoning, an old trick from the architecture studio that scales surprisingly well.
5. Civic epistemology Make the entire reasoning chain public by default. If employers can see how and what, the degree’s reputational value rebounds
Buying and not buying into the hype
At Artificial Intelligence Made Simple, the authors argue that while most people think of hype as noise, in low-information fragmented systems without a central authority, hype can actually function as infrastructure.
That description pretty much captures higher education EdTech in a nutshell. The authors go on to argue that:
in fragmented systems without centralized authority, hype functions as a kind of soft power: solving cold-start problems, aligning incentives, and triggering massive coordination cascades.
I’m not sure whether it’s because of or in spite of my time at Gartner that I find the Gartner Hype Cycle to be a compelling explanatory tool. Certainly, part of it is due to Jackie Fenn, one of the people who developed the concept, who is one of the smartest, nicest and most humble human beings you could ever hope to meet.
Hype certainly plays a role in propelling technologies from their first introduction to extensive interest and along the path toward widespread adoption. In that sense, it does function as infrastructure.
But I’m not sure that’s enough.
There also has to be a decent product behind it. Hype alone is not sufficient.
What happens when the hype subsides? Gartner argues that the product tumbles into the Trough of Disillusionment, and only genuine usefulness can pull it out on the other side. So, the infrastructure of hype only goes so far.
In higher education EdTech, I would argue that hype needs to be paired with peer adoption for the adoption cascades described by the authors to occur. Without bandwagoning, widespread uptake won’t happen.
I found the post useful and thought-provoking, though I did wonder how Sangeet Paul Choudary was going to manage to get a whole book out of the concept. I suspect I’ve partly answered my own question, and I should read the actual book when it comes out.
Just Google it
Alex Sarlin and Ben Kornell have the best rundown on the latest EdTech announcements from Google’s annual announceapalooza, Google I/O. These are significant, given Google’s role as the biggest EdTech company in the world. I don’t say that lightly—I’m typing this while wearing a MoodleMoot Global 2024 T-shirt.
The EdTech-relevant announcements include:
LearnLM: Google’s learning-focused model is now directly integrated into Gemini 2.5, Google’s core AI model.
Gemini’s new quiz functionality: Gemini can now create practice quizzes (with tailored feedback and follow-up quizzes) on any topic with minimal prompting.
Free Gemini upgrades: Higher education students in the U.S., Brazil, Indonesia, Japan, and the UK can get free Gemini upgrades, with more countries to be added soon.
Additional new features: Google also announced forthcoming features, including video overviews.
What intrigued me most, though, was Google’s announcement about the application of their chatbot-like AI assistant, Project Astra, to tutoring.
Project Astra is the Google AI team’s public research prototype for a “Universal AI Assistant,” and at I/O, Google unveiled a conversational tutor use case, in which Astra can use photo input to identify problems, provide step by step feedback, annotate problems or even generate diagrams. (note: Google bought Edtech AI startup Photomath in 2023).
While this product is still a research prototype, it is a harbinger of things to come in the AI tutoring space.
The more I research AI tutors, the more torn I become. There’s so much promise, but also so many challenges and some pretty underwhelming implementations. I guess I’ll just have to wait and see how it all pans out.
Not a parting thought, but say what?
According to Business Insider, there’s a growing trend among some parents to throw what are known as “college bed parties” to celebrate their children’s choice of college.
Gone are the days of wearing a college hoodie and snapping a selfie while posing with an acceptance letter.
Today, for young women in particular, college commitments call for an actual event known as a college bed party. This celebration involves decorating a student's bedroom with the colors, logos, and items that represent the college or university they've committed to — and sharing those images on social media.
The trend gained traction during COVID as a logical workaround for in-person celebrations. But instead of fading post-pandemic, it's only gotten bigger and bolder. Families are shelling out hundreds, even thousands of dollars, on the bed party craze, and small business owners riding the wave of this growing niche are here for it.
Among the businesses supporting these parties are bakers and cake decorators, which completely changed my view of the trend. I can totally get behind more cake, even if it’s decorated with a college logo.
My real parting thought
It’s that time of year again, so I’m back at the Ogden Music Festival, a wonderful excuse to sit among the cottonwoods and listen to all kinds of great music.
Last night, I got to hear an old favorite: Pokey LaFarge. I wanted to share my favorite song of his, “Cairo, Illinois,” but I realized just in time how absolutely depressing the lyrics are. So, I’ll leave you with this one instead.
The premium On EdTech+ newsletter is intended for a single recipient with occasional forwarding allowed.