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.

ulysse.gks & FARI / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/

Claude 5 is here and . . . gone! For a few days, the general public had access to what Anthropic billed as the most capable and dangerous AI model to date. This past Friday, the American government shut it down, fearing foreign nationals would use the powerful AI tool to wreak havoc. Or was it all part of Anthropic’s stylized hype? Who knows? AI is weird. AI developers are, too. With each new release comes an increasingly odd series of rules, guidelines, policies, and internet mysticism that has propelled Anthropic from a tiny sliver of users compared to OpenAI’s and Google’s nearly billion users each, to a household name. It’s sad, really, because we don’t need to artificially create a sense of FOMO around an AI product to understand that using natural language to code practical, functional web applications can be really useful.

But that in itself is the fable, at least for higher education. None of the enterprise plans I’ve seen offers the sustained access needed to help students or faculty explore creative use cases of AI beyond simple text and image generation. The future is arriving in a freemium ecosystem that most users and institutions will never be able to meaningfully access. That alone should make even the most staunch AI critic pause and ask who, aside from the already affluent and extremely resourced institutions, will have such access and what this means for opportunities for all students.

A free or low-cost AI education plan might allow you to test advanced AI coding features, generative video, audio, etc., but that level of access isn’t enough to iterate effectively outside of maybe a single class. If you pay $20 a month, you get more access, but soon find their capacity is throttled and rate-locked for four to six hours at a time. Those who can afford elite access to premium features and capacity, often paying between $100 and a staggering $2000 per month, are the ones who will be able to fully use the evolving features teased by new model releases, like Fable.

Practical Applications

I know little about writing code, yet I’ve now used natural language to generate dozens of functional web apps, games, simulations, and even galleries to host them. I’ve used both Claude Cowork and OpenAI Codex to build functional things that I think might help me do my job better as a teacher.

There’s never been a time when I’ve felt more pressure to speed up and also feel the pull to slow down and consider preserving what matters most. Decades ago, Neal Postman coined the phrase Technopoly to illustrate how culture often surrenders agency to new technologies and many believe that’s true of our AI moment now. But Technopoly isn’t a judgment about good or bad technologies, or the companies that develop them. No, Technopoly is all about how we uncritically adopt new tools without thinking. The answer to ubiquitous AI is intentionality, but how do we become intentional when access is so jagged?

I’ve previously written about how Ohio State University’s Committee on Academic Misconduct (COAM) developed a system of standardized “Academic Integrity Icons to promote transparent communication with students about the use of various resources on assignments, including AI, to transparently communicate with students about what an instructor will allow on an assignment. What I said then was imagine logging into your LMS and being able to easily communicate with your students what is acceptable for an assignment using a common set of icons they are all aware of! We should all investigate creating a common template that everyone can use to ensure faculty AI preferences are communicated transparently to students. With Claude, I did just that.

I used the AI Assessment Scale and generated a series of academic integrity icons in a simple HTML app that can be exported and embedded within assignments. I started with Fable, then had to finish it with Opus 4.8. With a flexible badging system, I no longer need to write out lengthy instructions to communicate what’s allowable within assignments for students.

When you can create functional web applications and host them for free on platforms like GitHub, a world of possibilities is suddenly open. You aren’t locked behind the LMS or a vendor; you are now capable of creating tools needed for how you teach. Yet, that is just another layer of this fable. The more I worked with advanced AI features on Anthropic’s and OpenAI’s platforms, the more it became clear to me how limited even the $20 monthly plans I had purchased on each were for me and the implications for equitable access this will have for faculty and students.

Each time, I’d burn through tokens in a matter of a few sessions and be rate-locked for 4-6 hours until I could access them again to finish an app or webpage. That’s not feasible for a student learning how to use a tool or feature, and it’s really not workable for many faculty who already possess those skills. The $20 month plan provided limited access to advanced models, like Anthropic’s Fable, but to truly explore creative applications, you’d likely need access to the $100 a month plan or be extremely good at project design and know what you want.

A Reading Tracker

For years, I wanted to help my online students read more deeply. I’ve loved using social annotation tools like Perusall and the ability to create connections across texts and readers. For this, though, I wanted to create something students could use independently. Imagine being able to copy and paste a text, highlight sections, label them, reflect on individual passages, then see a matrix of connections you’ve made while reading, all exported to you and your teacher once you’ve finished annotating and reflecting on it. That’s what I was able to create with this simple reading tracker app.

Creative Applications

Using AI to create practical applications is one clearly valuable use case for AI, but there are far more creative applications. Benjamin Breen’s experiments creating vibe-coded simulations for his world history students inspired me to see if I could create interactive games to augment the readings I planned for my fall students. I want my students to read a selection of classic science fiction short stories for students to explore the intersection of utopia/dystopia narratives and how these intersect with AI. So I’m building that section of my syllabus around close readings of Ursula K. Le Guin’s”The Ones Who Walk Away from Omelas,” Isaac Asimov’s “Nightfall,” and Harlan Ellison’s “I Have No Mouth, and I Must Scream.” To augment those readings, I tasked Claude 5 to generate simple gamified simulations interpreting aspects of each short story. It also walked me through how to host these games and set up a gallery for students to use them called Playable Texts:

The games aren’t great. They’re sketches of ideas, prototypes that need days, if not weeks, of work to make them functional for a classroom. Honestly, I’m not sure it matters that Claude 5 is down. I used up all the capacity I had on my $20 a month plan after it went down, switching between lower-tier models just to try and finish them, and I still didn’t have enough tokens.

Access to the most advanced AI models is increasingly based on what your wallet can afford and what region you were born in. What will that do to our classrooms and campuses? Industry is signaling that using AI coding agents will be extremely disruptive to our current digital world. Tim Cook stepped down from Apple recently, and several people have speculated that the reason had to do with the unclear place Apple’s app store plays in a world where anyone with a few thousand dollars and an idea can code an app and use their social influence to sell it as a product.

The messaging we’re hearing about AI is hurry! hurry! hurry!— move as fast as possible so that you aren’t left behind, but that’s not realistic for most of us. For a future arriving so quickly, many of us are trying to take stock of what these changes mean for our jobs, our daily routines, even our humanity. In that regard, there’s no reason we all shouldn’t slow down and process the big changes and challenges AI is having in our culture.

We may not have the resources, time, or strategies to preserve all learning from the uncritical adoption of AI by students in many of our educational spaces, especially those that are online, but we should likewise be mindful that we all don’t have the option or opportunity to use advanced models to pursue truly unique or creative experiments that transcend the academic integrity debates about AI. What we can do is lay out the stakes for students and ourselves and work to create opportunities where we all approach the technology with intention.

Preorder The Norton Guide to AI-Aware Teaching

Thanks to the wonderful team at Norton, The Norton Guide to AI-Aware Teaching is now available to pre-order! My coauthor, Derek Bruff, wrote the following in his newsletter. The ebook is expected to be available on July 1st, and print copies are expected to start shipping on September 24th. Here’s how you can get a copy:

1. Our publisher Norton is pleased to offer the guide as a free ebook for all instructors currently using a Norton textbook. If that’s you, you’ll receive access from the Norton team when the ebook is available July 1st and can contact your local Norton representative with any questions.

2. If you would like to pre-order the ebook so that you have it July 1st, you can now do so through Amazon and Barnes & Noble and perhaps other retailers.

3. If you would like to pre-order the paperback version of the book, you can now do so through Norton, Amazon, Barnes & Noble, and likely other retailers. If you go through Norton, be sure to use the code AIFREESHIP at check out to get free shipping!

4. If you would like to order multiple copies for a campus reading group or some other faculty development effort, Norton has an option for you: On orders of 10 or more print copies, we offer 50% off the list price and free domestic shipping. (Such orders must be on a nonreturnable basis.) To take advantage of this offer, contact Peter Wentz at pwentz@wwnorton.com with subject line “Norton Guide to AI-Aware Teaching.”