Awe Without Surrender: Staying Curious But Grounded with AI

A recent talk

Sharing a recent talk that I did at The Generator, a hub at Babson University. It was a fun and different talk as its focus was this question of curiosity, creativity, and exploration while also being grounded in the challenge.

Some of what I discuss here is also kernels of a future post on agentic AI, and some of the things I’m doing and learning with it, and how that’s making me think about AI.

Of course, you can check out the resource document that I mentioned and the slides.

---

Alright, let’s start with a quick game of Jeopardy

Up here are the categories.

• Famous Unfriendly AI

• Famous Friendly AI

• Techbro Failed AI Promises

• Biggest AI Fears

• The AI We Want

• AI: Myth vs Reality

Someone pick a category.

[Proceeded to play the Jeopardy game for a few minutes]

The reason we were able to start this way, playing with this Jeopardy interface, is that I built it with the help of AI. I could have dived deeply into PowerPoint, figured out all the features, learned macros, maybe watched six YouTube tutorials…and about six hours later produced something vaguely resembling Jeopardy.

Or I could experiment and see if AI could close that gap. Both moments of play (creating Jeopardy and playing it with you) is actually where I want to begin today. Experiences like that capture something real about this moment with AI.

They can be surprising. They can be a little delightful. Sometimes they even feel a little bit magical.…right up until the moment they confidently explain something completely wrong.

I want to allow space for that feeling. I want awe. Especially in the world today, I want just a little bit of awe.

But I also want us to stay grounded. Grounded doesn’t mean fearful or rejecting the technology. I know these tools are powerful and potentially problematic. Grounded means keeping our judgment and maintaining a little bit of distance. It also means remembering that even when the tool can do remarkable things, we are still responsible for thinking about what we’re doing with it.

That’s the space I want to explore today. Awe without surrender. Staying curious about what these tools make possible and staying grounded in the idea that we are still the ones responsible for how they shape our thinking, our learning, and our work.

A short time ago, someone asked me a question that has been sitting with me ever since. The question was simple.

What is an ethical reason to use AI?

Oh boy. That’s the kind of question that makes you immediately wish you had chosen a different topic to talk about.

Just curious—has anyone here been asked that question before? Not what are the ethical reasons not to use AI, but what are the ethical reasons TO USE AI? It’s a harder question than we usually ask.

I wasn’t a deer caught in headlights, but it did cause me to pause. We have lots of conversations about ethical reasons not to use AI, but what is the case TO USE it?

Neither the do or do not dispositions have simple answers. A lot of the time, when we talk about AI, the conversation revolves around productivity. What can it do? How fast can it do it? How much more efficient can it make us? How can it make us better workers or earn more money?

Asking about ethics shifts the frame. Instead of asking what the technology can do, we start asking what we should be doing with it. And that question pushes us into a deeper kind of reflection.

But whenever we talk about new technology, I like to zoom out the lens a little bit. Because if you listen to the way some of the conversation about AI is happening right now, it can sound like we’ve entered some completely unprecedented moment.

It’s the end of thinking. The end of learning. The end of creativity. To mangle a REM quote, it’s the end of the world as we know it, and I DON’T feel fine. None of us feel fine.

I understand where those anxieties come from. But in addition to being a scholar of education, I’ve also been studying popular culture for over 20 years (And that’s more than just reading comics, I swear). And when you study popular culture, over the centuries, you start to notice a pattern.

About fifteen years ago, we were having a very similar set of conversations about social media. What it was doing to attention. What it was doing to discourse. What it was doing to democracy. We also worried about how Wikipedia was ruining education. How many of us fought in the Wikipedia wars? How many of us are still fighting the Wikipedia wars?

Ten years before that, the conversation was about the internet. What it was doing to knowledge. To expertise. To the idea of truth itself. Go back a little further, and the concern was video games. Before that, television made us into mindless fools. Before that, comic books were corrupting the youth. Before that, radio–that disembodied voice in our homes–OUR HOMES–was going to drive us mad.

Apparently, the safest form of media was silence.

But even silence was dangerous. Silent films were going to make it impossible for us to make sense of what’s real and what isn’t. Before that, dime novels were making us think impure thoughts–especially women.

Show of hands: how many people remember hearing at least one of these arguments when you were younger? Ok, the rest of us were the ones making these arguments?

And if you follow that lineage far enough back, you eventually arrive at writing itself.

Another show of hands: how many people have heard someone say AI will destroy thinking? Excellent. You’re participating in a tradition that’s about 2,500 years old. Socrates worried that writing would weaken memory. That once things were written down, people would stop truly knowing them.

Artisanal knowledge; truly earned knowledge and understanding about the world, would escape us and make us dumberer.

Now what’s interesting about that lineage isn’t that all of those concerns were wrong. In some cases, they were partly right. Social media really has changed our attention. The internet really has changed how we encounter and engage with knowledge. Television really did reshape culture. Every new technology arrives carrying tensions we have to navigate. New capabilities and new distortions. New possibilities and new problems.

I’m not here to say we should stop criticizing AI. The task is to think more carefully inside that criticism. Because what we’re experiencing right now is actually a familiar kind of moment. A moment when our tools are changing faster than our habits. Faster than our institutions. Sometimes even faster than our language for describing what’s happening. And that’s exactly why that earlier question matters so much.

What is an ethical reason to use AI?

Because if we treat this as just another productivity tool, we will miss something important. But if we treat it only as a threat, we will miss something important too. The challenge is learning how to stay curious about what these tools make possible, without surrendering our judgment about how they should shape us. Now, one possible response to all of this would be to say:

“Well, if the technology is this complicated and if the implications are this uncertain, maybe the safest thing to do is simply stay away from it.”

I get that instinct. In moments of rapid technological change, withdrawal can feel like the responsible move. But the longer I’ve sat with this question, the more I’ve come to feel that—especially in education—that option doesn’t feel right. Abstinence and censorship do not feel like the right tools for education.

Because our students are already walking into a world where these tools exist. In workplaces. In research environments. In creative industries. In everyday productivity software. They are already shaping the environments our students inhabit, whether we like it or not.

That creates a challenge for educational institutions. What does it mean to prepare students for a world where tools like this exist? How do we avoid celebrating them uncritically or banning them reflexively and build the ability to sit with them uncomfortably and actively? We have to figure out how to help people learn how to think alongside them and how to think about them.

Those are different things. If students only learn how to use the tool, we’ve failed them. And if we pretend the tool isn’t there or there is no value or meaning to it, we’ve also failed them. The real educational challenge is helping people develop judgment. Helping them ask questions like:

• What is this tool actually doing?

• Where does it work well?

• Where does it fail?

• Where does it distort?

• Who is doing the thinking here?

These questions have been shaping my own relationship with these tools. It began with curiosity and a willingness to tinker around. I did what lots of different folks have done. I tried different prompts that I found online. I tried things that felt relevant to me. I tried things that felt irrelevant. I tried silly and ridiculous things just to see, what it would come up.

Be honest—how many of you have asked AI something completely ridiculous just to see what it would do? Right, this is how most people begin using AI: we want it to make bad poetry, questionable recipes, and emails to our bosses.

But then, a trend I started to notice and learn from others was that I could use the tool to learn about the tool. This reminded me of the early days of the Internet, when I was building my first website on GeoCities

For those of you who remember GeoCities; welcome, fellow elders. You can expect to start receiving AARP recruitment letters any day now, if you aren’t already subscribed.

This moment was very much like that moment–I could use the internet to help me figure out how to make things. AI helps me figure out AI on a different level than what the internet did. Except, of course, when it doesn’t work.

Yet when it does, it’s fascinating. I can ask it how to help me develop better prompts for better outputs. I can ask it to challenge what I am thinking. I can ask it to critique itself and challenge its own outputs.

As I’ve come to learn these practices, I started another practice. Instead of just asking the system for answers, I started asking it to ask me questions. To push back. To ask me what I actually meant. To ask me to clarify my thinking.

Even this talk, it was developed by me sharing my thinking, and an AI asking me targeted questions to push my thinking and draw out assumptions and misconceptions. This dynamic changed my working and how I used it.

The tool stopped feeling like an answer machine and started feeling more like an interviewer. Which is great. Until it starts asking questions that reveal you didn’t think your idea through…or worse, it starts to ask about your relationship with your mother.

Still, someone—or rather something—pressing me to articulate what I was actually trying to do–that was intriguing. And once that shift happened, something opened up. It helped me to realize about this as a technology that can be flexible in relation to how we think. Yes, it can influence how we think if we are on autopilot, but it can also disrupt, challenge, or push our thinking.

It’s become a space of experimentation.

My thinking about these tools is less interested in the efficiency discourse, which I have some strong critiques about around this technology, and more about how these tools help me explore and experiment.

That’s where the story of the building apps and play really becomes prominent.

What does that experimentation look like?

If you haven’t already figured it out, I’m from the Napster generation and I like to have music files rather than stream music. However, I’m also not a fan of most commercial music applications. There’s ads, there’s clunky interfaces, there’s a billion options–none of which I actually want.

I wondered if that is something that I can do with AI. Given that the extent of my programming ended decades ago building geocities websites, I thought this could be an interesting way to see if I can build software that I want, not just software that I tolerated.

I did not have a grand vision of how this could go. But I was curious whether I could close the gap between what I imagined and what actually existed.

First, I went to AI and had it interview me about what would be the features that I wanted from an Mp3 music program. Then, I worked with AI to start to build it, test it out, refine it, and complete it. There were hiccups along the way, but there was also things I started to learn about programming. I also started playing with and using programs that I never had before: Visual Studio, Xcode, and even hanging out in my computer’s terminal, regularly.

Harmony - MP3 app for Macs

I did get to a working MP3 program; I named it Harmony. And well, once that idea happened, it opened up new ideas and possibilities. I started to ask, “what else could I do?”

The next thing I built was an RSS reader. Again, I’m dating myself because not as many folks are familiar with RSS feeds as they should be; we’re all stuck in infinite scrolls on our social media platforms, but I wanted a space that I had more control over. Again, there are some RSS feed readers out there, but they’re all mediocre, cost money, or clunky.

I wanted one that allowed me to do some simple and consistent things and that I could count on not going away, changing prices, or adding more features. Once more, I started with AI to interview me for what I wanted. I used a summary of that conversation to work with another AI thread to build out what I was looking for. Reasonable progress, some hiccups, some new things learned, and now I have an RSS Feed Reader called “Fetch N Feed.”

It was about this time that I did something I never thought I would do, I opened up a GitHub account and began sharing the programs online. It helped that I could use AI to help me make sense of GitHub, post, and clean up my messes.

But after Fetch N Feed, I created a Podcast Listening tool, called PodVault, and then a random-artifact finder for the Internet Archive. I now find myself regularly thinking about what I can build, what I can make, how I can do things that wouldn’t usually be within reach.

PodVault, a podcasting listening app.

None of these programs are going to disrupt Silicon Valley. I’m fairly confident no venture capitalist is waiting outside to fund Harmony, Fetch N Feed, and PodVault Incorporated. BUT if you are…call me?

Still, every one of them does something that used to irritate me about existing software. That sparked a new insight. For most of my life, software was something I had to adapt to. Someone else built it, decided what the workflow should be, and what features mattered. And my job was just to live with those decisions.

That dynamic has shifted. The annoyances were no longer permanent. If I can imagine a different workflow and I can articulate what I want, I could at least try it.

Sometimes, it works. Sometimes it fails. Sometimes, it works better than you thought it could. Learning is happening regardless, and I’m gleaning the contours of what was possible with these tools.

Ok, let’s pause here for a moment. This is the moment where awe shows up and the moment where we have to remember not to surrender. This is the point where a certain kind of story about AI usually ends. The story where the technology suddenly feels magical, and the machine can do everything. That feels disingenuous because it’s more like a messy collaboration.

There’s iteration, there’s failure, there’s spending hours on trying to make a game for your partner, only to realize that’s way beyond your abilities, and the AI was really just sending you down useless paths. Or was I the only one that has happened to?

Still, there was something a little awe-inducing about it. The distance between idea and experiment had shrunk dramatically. Things that once felt impossible to even attempt suddenly became things I could try.

That is exciting. It’s also exactly why the question of groundedness matters so much. If we collapse the distance between imagination and execution, then the temptation becomes to just keep building, generating, producing. Not only our own inner temptation but, unfortunately, the external expectation of our world. Producing more becomes the point.

But we always need to stop in order to ask what we are learning in the process.

Over time, one question I keep coming back to in my usage of AI, and as I think about what it means to be a person and a professional in today’s world. Who is doing the meaning-making here?

Just out of curiosity—how many of you have used AI to help think through something in the past week? [The majority of the room raised their hands.] That’s why we need these conversations. Is the AI generating something that I am simply accepting, or is it helping me clarify something that I am still responsible for thinking through?

Now, there will be times when it might be ok to accept it. Building an MP3 player for me to enjoy music on is a fair example. Other times, I need to be more thoughtful about what it offers.

As mentioned, for this talk, I had AI interview me and get my ideas out about how I wanted it to go. I gave all the details; it organized them. It came up with a decent enough draft, but it was one that had a lot of limitations. In fact, in the resource document, you can see the “AI Draft” vs. the Final Draft. Even that final draft isn’t the final draft, as I was editing the text on the slide this morning.

I had to do a reasonable amount more to make sure the meaning-making in this talk genuinely reflected what I wanted to say and how I wanted to say it.

Because if the tool replaces our thinking or just embeds too deeply as our thinking, we will lose something important. The goal is to use the tool, not have the tool use us. Have it challenge, push, or extend your thinking and see what else becomes possible.

Yet again, it would be easy to tell the story I just told and leave it there. Look what we can build and experiment. Holy smokes, the distance between imagination and execution has collapsed. We can’t give up our critical thinking. Peace out, have a good ride home.

That too would be negligent. There’s another set of tensions that we have to sit with. These technologies never arrive as neutral tools. They arrive wrapped inside systems.

• Economic systems

• Labor systems

• Environmental systems

• Systems of ownership and power.

That means engaging with AI responsibly requires us to hold more than one thought at the same time.

The possibilities and the problems.

On one hand, these tools can genuinely expand what individuals are able to attempt. We can explore ideas faster. Prototype ideas faster. Experiment in ways that would have been far outside our reach before. But on the other hand, we also have to ask harder questions.

• Questions about where these systems come from.

• Questions about whose labor helped produce them and whose labor is lost.

• Questions about the environmental cost of running them.

• Questions about bias embedded in the data they were trained on, and how it impacts us in our own thinking.

• Questions about where the material came to train the models and who owns the outputs.

• Questions about who benefits most as these tools become integrated into our lives.

And many more questions. Each question has a plurality of nuanced and complicated answers.

Take the environment as an example.

We know that they have an environmental cost, and we want to be aware of that. Still if we only understand the environmental cost of an AI prompt without a deeper understanding of all the other costs of our social media, streaming video, device charging, and just having our numerous devices that we have always-on and always connected–then we’re not really focused on the environment.

These questions and how we go about exploring the answers are not unique to AI. They are questions that have accompanied nearly every major technological shift. These are perpetual problems–important issues to grapple with as we have with every technology. The difference in this moment is the speed at which this technological change makes us face those questions much sooner than our institutions are used to processing them. We can’t resolve those tensions immediately, and some we may never resolve. Others may actually be resolved through the process of exploring and further developing the technologies.

The challenge is learning how to live thoughtfully inside them. To stay curious about what the technology can do while also remaining attentive to the systems that surround it.

One practice I’m cultivating that certainly isn’t new in the world, but we need helpful reminders about is the practice of the pause. Working with AI benefits from an intentional pause. A moment where you step back and ask a few simple reflective questions. Questions like:

• What am I actually using this tool for right now?

• Is it helping me think, or replacing thinking I should probably still be doing myself?

• What am I learning through this interaction?

• What habits am I developing as I work with it?

These tools change what we do, and they also change what habits and actions we get used to doing or expecting from our technology. Most importantly for professionals and scholars, they reshape our habits of attention, inquiry, and patience. We should be mindful since education has always been, in part, about cultivating attention, inquiry, and patience.

Of course, none of these answers the question I continue to think about: What is an ethical reason to use AI?

I don’t know that I have a clear and accurate answer; certainly not one that will be useful to purists. As an educator, though, I can see its presence and know that the better I can understand it and use it, the better I can bridge students’ learning to whatever it is I’m teaching. That being literate in AI is important in the same ways as being literate in other consequential focal points in our world. As important for me is to just keep carrying that question with me.

I can see in many ways that AI becomes most valuable when it helps us explore, articulate, test, and extend ideas without relieving us of the responsibility for thinking. It allows us to further examine ourselves and our worlds, and that is part and parcel of a key part of our intention in higher education–to continue to make sense of a complex and diverse world.

The tool can extend what I do. Yet to do that well, I need the balance. We all deserve awe without surrender.

I also need help to figure out AI–we all do. Something else I’ve been thinking about since early on with ChatGPT is that we all think differently. Therefore, how we engage with AI will look different.

AI isn’t a thinking tool, but it is a tool that mimics and intersects with thinking. Because of that, learning how different people engage and use it is incredibly important and valuable. It can help us better understand the borders of what AI is and can be.

Even more important, the sharing of how we each engage and the learning we each take from it can help us meet artificial intelligence with human collective intelligence. An intelligence that helps us understand and connect better with one another, and also better understands the place of these new tools.

If that’s the posture we’re trying to cultivate, then the final questions become practical ones.

What does that actually look like in our everyday work? Whether it’s in our classrooms, in our labs, or in the world at large. How do we build and engage with these tools, and how do we learn together with them?

Earlier today, we started with a game of Jeopardy. And for a few minutes, we were just playing. Guessing. Laughing. Trying things out. That captures something important about how we might approach AI. We need curiosity, experimentation, and community. A willingness to try something and see what happens.

But what matters is what happens after the moment of awe or that surprise. What do we do after the tool does something impressive? That’s the moment where judgment comes back into the room. Where we should ask:

• What did we just do?

• What did we learn?

• What role did the machine play

• What role did we play?

• How do I share these insights?

Some worry about AI suddenly becoming intelligent or sentient. The actual risk is that we slowly become less reflective about our own intelligence. We become less attentive or curious. We choose not to wrestle with difficult questions.

That’s why I keep coming back to who is doing the meaning-making here? If the answer is always the machine, then we’ve surrendered something important. But if the answer is still us, then these tools may actually expand what we’re capable of exploring.

And that practice is the counterbalance to artificial intelligence: collective intelligence. Individuals sharing how they experiment and pause with these tools. People sharing notes and challenging one another’s assumptions about what AI is and isn’t. Learning together what these systems can do and what they shouldn’t do.

Because none of us fully understands this technology yet, no matter what experts say. And that’s okay. Moments of technological transition have always required something from us. To find our way through this, we need deep curiosity, a good deal of humility, and sustained conversation. The simple takeaway is to go and thoughtfully engage with AI. The deeper task is to learn how to engage meaningfully and share with others–engage together as you have today.

That, for me, is what awe without surrender looks like. Allowing ourselves to be surprised by what these tools can do. But never forgetting that the most important intelligence in the room is still human intelligence. And especially the intelligence that emerges when we think together.

---

The Update Space

Upcoming Sightings & Shenanigans

• Writing with AI: Making Informed, Defensible Choices Before You Submit. Free Webinar with Paperpal. Wednesday, April 22 at 3pm (ET).

• EDUCAUSE Online Program: Teaching with AI. Virtual. Facilitating sessions: ongoing

Recent Recordings, Resources, & Writings:

• Damm, C., & Eaton, L. (2026, March). From prompt to practice: A framework for transparent GenAI use in higher education. EDUCAUSE Review.

• Eaton, L., Nemeroff, A., & Sun, X. (2026). AI-assisted course design and development. In K. S. Ives, M. Cini, & R. Schroeder (Eds.), AI applications in online higher education administration: Strategies for maximizing returns and improving outcomes. Routledge.

• Margin of Thought with Priten: Season 1, Episode 5: How Can We Center Pedagogy During the AI Tech Wave? (February 2026)

• Online Learning in the Second Half with John Nash and Jason Johnston: EP 39 - The Higher Ed AI Solution: Good Pedagogy (January 2026)

• The Peer Review Podcast with Sarah Bunin Benor and Mira Sucharov: Authentic Assessment: Co-Creating AI Policies with Students (December 2025)

• David Bachman interviewed me on his Substack, Entropy Bonus (November 2025)

• The AI Diatribe Podcast with Jason Low (November): Episode 17: Can Universities Keep Pace With AI?

• The Opposite of Cheating Podcast with Dr. Tricia Bertram Gallant (October 2025): Season 2, Episode 31.

• The Learning Stack Podcast with Thomas Thompson (August 2025). “(i)nnovations, AI, Pirates, and Access”.

• Intentional Teaching Podcast with Derek Bruff (August 2025). Episode 73: Study Hall with Lance Eaton, Michelle D. Miller, and David Nelson.

• Dissertation: Elbow Patches To Eye Patches: A Phenomenographic Study Of Scholarly Practices, Research Literature Access, And Academic Piracy

• AI Syllabi Policy Repository: 200+ policies (always looking for more- submit your AI syllabus policy here)

Finally, if you are doing interesting things with AI in the teaching and learning space, particularly for higher education, consider being interviewed for this Substack or even contributing. Complete this form, and I’ll get back to you soon!

We periodically host small-group workshops and leadership sessions for higher ed teams. You can learn more about our current offerings here.

---

AI+Edu=Simplified by Lance Eaton is licensed under Attribution-ShareAlike 4.0 International