We just don't know.

My mid-year crystal ball on AI in education.

As your resident organized, structured, and step-by-step expert, I’m here to say: when it comes to AI in education, we just don’t know.

The unique problem AI in education presents is that so much of understanding something is rooted in learning. And learning, en masse, happens at schools. Meaning, in theory, that students should learn OR it’s expected that they learn about AI in the classroom. But that’s not what’s happening. At least right now. And it’s because we’re in a very tangled transition phase that involves so many variables and so much nuance, that no one can confidently speak to what AI in education should look like.

I wrote down a few notes on what I think this phase should look like, what it shouldn’t look like, and maybe find a way to ground us in something.

What it does look like:

For tech companies

• Better product marketing and packaging: this probably isn’t a hot take, but outside of tech, AI is a problem. It fell victim to its own version of the horn effect, and its only positioning to the rest of the world is students cheating on essays and lazy people who can’t write an email. All of the #ThoughtLeadership discussion on automation workflows, team collaboration, and reducing manual tasks is only happening on LinkedIn (where students and teachers are not). The messaging generative AI needs to earn buy-in is either absent or buried in things like frustration, fear, and disinterest

• Data & privacy as non-negotiables: Probably don’t want to end up in a hacker saga like PowerSchool 😬

For teachers:

• ~Being the change~: Because it’s still new, it’s likely that your school and district have limited/zero AI regulations and/or best practices established. Sharing your insights at the school level can influence district-wide decisions

• Assuming students are using AI: At this point, it’s safe to say the majority of students are probably using AI (whether we like it or not). The opportunity now is to guide when, how, and why to use it, shifting it from a cheating tool to a thought partner, and helping students meet the moment with some sort of intention

• Educators becoming designers of AI experiences: AI in education needs more people who understand their students enough to ask things like: “where can AI scaffold, and where should it step back?” I imagine this would be in lock step with the point above

What it doesn’t look like:

For tech companies

• The classroom isn’t a sandbox: Students and teachers aren’t beta testers. Rolling out half-baked tools into schools under the guise of "innovation" is reckless, especially when there’s little oversight or support

• …speaking of support: Offloading risk to educators: When tools come with unclear guidance, bugs, or inconsistent results, teachers are left to clean up the mess. On top of everything else they already do

For districts + admin

• One-size-fits-all AI training: Revamping PD is a topic for another day, but we have to be honest and know that a 60-minute webinar on “AI in the classroom” isn’t enough to make teachers feel supported and prepared

For students

• Framing AI as the end-to-end solution, not a tool: relinquishing everything to the generative gods and not understanding the input or output is *not* how to use AI

For everyone:

• Avoiding nuance: AI isn’t an apocalypse or a miracle! It’s a tool that needs infrastructure, boundaries, and thoughtful experimentation. This moment, this phase, demands nuance and not absolutism

• Constant discussions: Don’t explore AI in a vacuum. Ask your peers what’s working. Share your use cases. Compare notes across schools, roles, and contexts

What does it maybe look like (aka, my cop out for not bucketing it in the above)?

Redefining what education even is.

You've probably seen this conversation a few times, like maybe here. AI is accelerating fast, while education remains slow, which isn’t surprising because designing a public system that serves 50 million K–12 students is Herculean.

The problem is that everyone wants something different. Students and parents often see school as a pipeline to good grades and college. Teachers want to inspire and support their students. District leaders need funding. Meanwhile, tech companies are launching AI tools like they’re Memorial Day sales. We’re not aligned. And without alignment, transformation is an uphill battle.

It’s all messy, and worse, there’s no immediate solution. We need urgency and patience. We need intentionality within uncertainty.

I’m curious (and honestly slightly terrified) to revisit this post in December to see what still holds up, and what we have to add to keep up.

---

Study Notes

• The turns have tabled! Students are frustrated that professors are using AI, after scolding them for doing so

• More states are allowing community colleges to offer bachelor’s degrees to expand access for rural students

• Major win: “Mississippi has become the fastest-improving school system in the country.”

  Mississippi Schools Are Better Than Yours