Introducing: Tell Me Your Terrors: Joys and Fears of AI & Education

A revamped newsletter, and (coming soon!) podcast

Hi everybody, this is Hugh McGuire. I am relaunching this long-dormant newsletter (now on Substack!) You may have subscribed ages ago. Welcome back. I’m also planning a podcast, coming soon with great guests—stay tuned! The focus will be AI and learning, and I will of course be touching on the open web, reading, creation, publishing, technology, and collaboration. I hope you’ll join me. Below is a little intro to what I’ll be covering.

AI has come for us all

For many years, I have worked on projects exploring the power of the open web, open collaboration, the writing and reading of books on the web, and the sharing of knowledge. I started LibriVox in 2005 (free public domain audiobooks created by volunteers), Pressbooks in 2011  publishing platform mainly used in higher ed for publishing free/open digital textbooks) , and The Rebus Foundation in 2016 (support for groups of educators publishing open educational resources).

Tied together in all of these activities was a desire to help people create and share knowledge with the world, through the medium of the web. I’m still animated by these ideas, but something ground-shifting has happened that profoundly changes the human relationship to knowledge: AI has (truly) arrived as a tool that you and I can use.

I believe that AI, even in its present form, will transform the human acquisition, sharing, and use of knowledge. We are still in early days, but AI is getting more powerful every hour. If you’ve really played with ChatGPT 4 and some of the more advanced consumer-facing AI tools like Claude and Gemini, you can see glimmers of a transformation.

I expect AI to be a bigger driver of change than the web has been. Possibly a bigger driver of change than the printing press. Perhaps I am wrong, but I don’t think so.

It’s an exhilarating, and terrifying, time. Below you’ll find a small introduction to some of the things I will be exploring in this newsletter and podcast. I hope you will join me.

Existential risks, AI’s recent history, and the things GPT is good at

Back in September 2023, I had lunch at a little Persian/Italian restaurant in Montreal with some old friends who are very smart about technology. Other than the excellent hummus, there was only one topic of conversation: ChatGPT. Things got hot and heavy, and we probably frightened the other customers—such was our excitement and animation.

We talked about existential risks of AI taking over the universe, the speed of evolution of AI tools, the ways in which ChatGPT gives everybody an easy interface to interact with AI, which has, in fact, been working away in the background of our digital lives for years. We talked about the cool things ChatGPT was doing for each of us—from suggesting controversial conference topics to organizing trips. And we talked about all the cool things it surely would be doing sooner than many expect.

But, what do humans add?

I kept coming back to one question though: if you project forward, what can humans add to whatever ChatGPT does and will do? If we project out six months, six years, 16 years, as ChatGPT and its ilk get better and better at doing so many of the things humans have been uniquely good at in the past, what will be the value add that humans can provide? What is the nature of the unique value humans might add to a world with powerful generative AI everywhere?

Educators, tell me your terror

I have spent the past number of years (at Pressbooks and Rebus) focused on higher education. And higher education, let me tell you, is grappling with something very, very big with the arrival of ChatGPT.

Not everyone answers this way, but if you ask a senior administrator at a higher ed institution what their biggest terror is right now it often boils down to:

“If ChatGPT can write essays and do exams as well as our students, how are we going to assess what our students have learned? If we can’t assess our students with confidence, how can we stand behind our credentials? If we can’t stand behind our credentials, what service are we delivering?”

Educators, tell my your excitement

It is by no means all doom and gloom among educators. So many of them see unlimited possibilities for AI to help in teaching and learning. Personalized tutors that know everything about you; personalized materials that are shaped to optimize your learning needs. AI is polite, helpful, awake 24 hours a day, and knows everything (ish). Surely we can put such a tool to amazing use in education?

But at what cost? Does everyone benefit equally? Do teachers still have jobs? What about the systemic biases built into the models?

We are at a crossroads in education. Terror and possibility are unveiling themselves to us, faster than we can keep up, and while we face a scary, bumpy time ahead, it is also invigorating.

AI, the (sort-of) all-knowing entity, everywhere

In generative AI, we have among us now this incredibly powerful entity that is knowledgeable on just about every topic. It is able to converse, provide guidance, and also make mistakes. It can answer easy questions and write long essays, and summarize the most complex scientific concepts. It can simplify and simplify. It can parse and format data. It can write clean code and find bugs. It can suggest areas of inquiry and suggest ways to rethink questions. It can translate into languages, and restate ideas through multiple philosophical frameworks, explained through the voice of a 5-year old or in the style of Shakespeare.

In short, we have at our fingertips a revolution in human access to knowledge unlike anything we’ve seen before. It’s terrifying and exciting.

Humans are learning machines

Humans, someone recently told me, are learning machines. I like that framing, because it helps focus on a critical question: How can we use AI to help us all learn? How must learning change, knowing we will have AI at our fingertips in the future? How can we use AI to help teachers be better at what they do? How can educational institutions adapt to become better at delivering the learning that people want and need? What should a classroom look like in 10 years? What about educational institutions?

I’ll be talking about all these questions, and more, in this newsletter and podcast. If you know someone whose life touches learning and education, who might be interested, please pass this along to them.

---