Is It Time for AI Immersion Learning?
In July of last year I intentionally injected novelty into my life. Jumped in with both feet. I tried to learn a new skill it wasn’t clear I was qualified for. It was terrifying. And confidence building. It had been too long since putting myself out there, feeling the imposter.
It was singing. With other people. Not in my car alone, choosing whatever octave suits me, imitating the style of the artist rather than expressing my natural tones, skipping the hard parts my voice can’t reach, or jumping to a different vocal part because it’s more fun.
My voice is decent; my musical kids gave me confidence that I could sing. But I had no training beyond the experiences of following the bouncing ball when bored in church as a kid and whatever I vicariously picked up from car-belting practice and my daughters’ musical journey.
I had joined a chorus in Western Massachusetts. For a week. Maybe 8 hours of singing a day, surrounding eating and chores, “campfire” singing in the evening, and sleeping in 100°F heat. Honestly I had no idea what it was going to be. Something about singing. The randomness was part of my intent.
But man was it intimidating. They all clearly knew what they were doing. I was trying to listen and remember my part (and avoid memory degradation while the other parts are practiced), relate what I hear to the sheet music, learn to pronounce the mostly non-English lyrics, hear my part well enough to hit the right notes, while hearing and integrating with the other parts.
My brain melted down the first day. A few times I just had to stop singing to ease the mental overload. But by day 2 I knew I could handle it. Others liked my voice. Eighty-something Wanda came over to tell me she never sang until in her sixties and that I’ll figure it out. Everyone in and directing the chorus were so patient, layering the learning in a way that each step felt manageable.
Emotionally it could have gone either way. I needed constant self-talk to get there. But I learned more in that week than a year in a conventional chorus. And gained a new perspective and hobby.
Immersion learning is psychologically tough. It works better for experiential learning than knowledge retention; immersion there is called cramming. Yet it can be a powerful learning tool.
Right now, schools and colleges need some immersion learning about AI. As a community. A coming together, where individual voices can still be heard but pragmatism toward quality solutions supersedes idealism.
This article is about what that might look like and what benefits and risks accompany the options.
The Current Approach Isn’t Working
Most schools haven’t changed much since Large Language Models (LLMs) hit prime time in November, 2022. There’s no shortage of short courses and media about AI since then. But most in schools have heard opinions about AI, often in siloed ways, more than are exposed to the more verified knowledge and issue complexity. Many in schools and colleges have responded to AI with outrage, but the responses of sending a few people to workshops, convening a committee, and letting the motivated few tinker isn’t commensurate with that disruption. There still seems a veneer of hope that if they just ignore it society will decide to make it go away. But their faces are turning blue; you can only hold your breath so long.
Workshops can serve real purposes, but understanding AI’s impact requires using it—or at minimum being shown how it’s used in ways beyond the obvious, and where it goes wrong. Passive formats don’t get you there. Neither does the typical committee lifecycle, useful as committees can be for surfacing perspectives and building policy: they meet episodically, they circulate documents, and the hard conversations get deferred until the next meeting, then the one after that. Nothing accumulates. Often each gathering restarts from roughly the same place.
Meanwhile, the underlying pressure hasn’t let up. AI is inside the research students submit, the feedback teachers give, the administrative communications that go home to parents. The model that schools have operated on for generations—information scarcity, human expertise as gatekeeper, performance assessed through demonstration of knowledge—is under serious strain. Fragmenting attention and spreading responsibility thin is a response to that pressure, but it isn’t a solution.
There’s some evidence for immersion learning. The National Writing Project’s Summer Institute—a multi-week intensive where teachers write, teach each other, and build community around a shared challenge—is one of the most rigorously studied professional development models in education. Research on the model consistently finds that transformation correlates with depth of immersion and the confidence that comes from intensive collaborative work, not with the number of sessions attended. It works not because of a single authoritative voice but because it’s a community of voices. The institutions that have made genuine progress with AI at some point stopped treating it as one more thing to manage alongside everything else and focused on it long enough to get institution-wide traction.
What Immersion is Useful?
An AI immersion inverts the normal priority structure temporarily. Normal instructional and administrative business would need to be made second-tier for a defined period—a week is a reasonable unit—and the institution focuses collectively on AI.
Collectively means with everyone. Teachers and administrators, yes, but also students, parents, and at the college level, staff and community partners. Each constituency brings something the others can’t. Parents who use AI in their professional lives or are seeing their fields change can have conversations with students, administration, and other parents about it. Students—especially older ones—often have more hands-on AI experience than their teachers and deserve a real seat at the table rather than being the passive subjects of a policy debate that’s ostensibly about them. Teachers and administration can demonstrate to students in tangible ways how learning suffers when they offload to AI, but might be empowered if they use it differently. Teachers who’ve been quietly experimenting can surface what they’ve learned in a context that gives it weight.
The agenda is custom to the institution, but the key is to identify the AI questions that matter to the community, create conditions where people learn enough to move past binary positions, and assign each adult in the system a specific task with a real deliverable at the end of the week—not months later. That requires clearing time for those adults to actually do the work, which means other support will be needed to keep students occupied for part of the week. The goal is to roll up what’s been learned into commitments with real names attached to them.
What does that look like in practice? At a K–8 school, a meaningful thread might be parents demonstrating how AI has entered their fields, followed by structured conversation with students about what that means for what they’re learning. At a high school, it might be a genuine debate—with real disagreement tolerated—about AI use in assignments, followed by small groups drafting actual classroom policies, or a guided discussion about how the nature of work is changing and to what degree learning should follow. At a college or university, it might include faculty working alongside instructional designers and students to prototype AI-integrated assignments and stress-test them together.
The through line is doing. Building something, deciding something, disagreeing about something until a workable position emerges. A week of that produces more durable change than a year of workshops because it requires everyone to actually engage with the hard questions rather than observe someone else engaging with them.
What Makes It Work—and What Can Derail It
A week of intensive community focus on AI doesn’t automatically produce good outcomes. Done badly, it can calcify the wrong consensus, or produce the appearance of buy-in that evaporates by the following Monday.
The most important precondition isn’t agreement; it’s nuance. The goal isn’t to get the AI skeptics to become enthusiasts, or the enthusiasts to become more cautious. It’s number one to put real attention on what is a gigantic societal and educational issue. But beyond that it’s to get everyone past their binary positions. There’s more complicated but also more productive territory in between “AI is going to ruin education” and “AI is going to transform education.” Binary thinkers can’t problem-solve together.
One way to get there is to give skeptics real problems to work on that matter to them. An educator who’s worried about AI undermining academic integrity isn’t going to be argued out of that concern by a keynote speaker. But perhaps if they draft some ways to measure the impact of AI uses on learning and have to socialize it to get metric and process agreement, that skepticism is put to productive use. The socialization required to reach agreement could also force tough conversations that otherwise get ignored. They’re no longer defending a position; they’re solving a problem. That’s the shift immersion can make possible, because it creates the time and the structure for that kind of engagement rather than assuming it will happen spontaneously.
The other thing that derails immersion efforts is poor facilitation. The chorus in Massachusetts worked because the people running it were patient, layered the learning deliberately, and made space for the terrified newcomers without letting the most advanced participants set the pace for everyone. An AI immersion week needs facilitation that holds the tension between different perspectives long enough for something real to emerge, rather than letting the loudest voices foreclose the conversation early, or letting AI rumors and use shaming turn it into a mosh pit.
The week doesn’t change minds on its own. What it can do is create enough shared experience, shared vocabulary, and shared accountability that the work can continue after it ends. A good use of AI itself might be to help plan something this complicated—adjusting course schedules so that some aspect of the curriculum still gets covered, but in the context of AI. The real output is a community that’s been through something together and knows how to keep talking.
Real change in education has always required more than information. It requires changed relationships between people, and changed expectations about who gets to shape the decisions. An AI immersion week is a bet that focused, whole-community engagement can do what years of workshops and committee meetings have not. Most of the community has never been involved in the important AI-related decisions and conversations. It’s hard to feel part of the solution when you’ve only ever been imposed upon.
I joined the chorus clueless, and I left feeling part of something. I was capable of it, as many more are in the education community when it comes to AI. I just hadn’t given myself a chance.
Give your community a chance. I bet they will be far more constructive than you think once they feel they can contribute one way or another. They’re not resistant because they don’t care. They haven’t had the experience of being in a room where the question is real, the conversation is honest, and something is actually expected of them when it’s over.
©2026 Dasey Consulting LLC