EdTech, AI & Technology in Schools FAQ

Has EdTech actually improved student learning?

No. Despite massive investment, there's no evidence that any EdTech product works better than old-fashioned paper and pencil, textbooks, and teacher-directed instruction at scale. Things often work in the pilot because people are enthusiastic, but when you go big, the results disappear. Watch the video →

Why are you betting on a return to low-tech learning?

Because parents are pushing back, lawmakers in 16 states are pursuing legislation to roll back one-to-one device programs, and the evidence increasingly shows that pencil and paper is actually better for learning. The gamification and dopamine traps built into educational apps may be frying kids' attention spans. Schools are going to start putting devices away. Watch the video →

Should schools go back to paper-and-pencil standardized testing?

Yes. So much of the reason we have Chromebooks in elementary school comes down to testing. I bought Chromebooks as a principal expressly for testing, and it made sense at the time. But kids show what they know better on paper, and the shift to digital testing drives a cascade of technology purchases that don't improve learning. The efficiency gains aren't worth the educational downsides. Watch the video →

Are Chromebooks as bad as cell phones in the classroom?

They might be. Neuroscientist Jared Cooney Horvath's research shows that up to 38 minutes of every hour can be lost to distraction on school-issued devices. Kids get around every filter, get off task remarkably fast, and the net benefit of one-to-one devices is probably negative. If there's a place for these devices at all, it's a very limited one. Watch the video →

What about BYOD — Bring Your Own Device?

It's really Bring Your Own Distraction. There is no universe in which the net benefit of cell phones in the classroom is positive. A lot of the value of education is not access to technology or information — it's the structured opportunity to interact with other students, a teacher, and content in a way that can't be reproduced anywhere else. Watch the video →

Are kids really learning 10x faster with AI?

No. That claim is nonsense. The people making it are confusing self-motivated adults using AI to learn a topic they're passionate about with the reality of getting kids to learn everything they need to know, day after day, for an entire year. Learning is hard work, and human teachers are irreplaceable at motivating students to do that hard work over sustained periods. Watch the video →

What about Alpha School's "AI instead of teachers" model?

It's being overhyped. They charge $40,000 a year, have no teachers, and put kids on computers for two hours a day. The biggest constraint on learning isn't explanations — it's motivation. Eventually a kid on a computer runs out of steam, and Alpha's answer is extrinsic motivators and withholding snacks. That's grim. Most kids are motivated by participating in class with their peers, not sitting alone at a screen. Watch the video →

Is AI grading a good idea?

Only if it doesn't damage the teacher-student relationship. The problem with a lot of early AI grading is that the AI takes the teacher out of it completely — students get feedback that's clearly from a computer, passed along without teacher involvement. That breaks something important. AI can help with the workload, but the feedback has to pass through the teacher so professional judgment stays in the loop. Watch the video →

What's wrong with AI-generated SEL and mental health lesson plans?

Two things. First, schools shouldn't be doing mental health interventions in the first place — educators aren't mental health professionals, and the evidence base for school-based SEL interventions is very poor. Second, the idea that you can generate a lesson plan in a pinch with AI because you don't know what you're doing is not how a profession operates. Parents won't feel better knowing these lessons were AI-generated. Watch the video →

Should I use AI to write my cover letter?

You can use AI as a tool, but if you don't provide specifics about yourself — your career, accomplishments, and experience — your cover letter will sound identical to hundreds of others. Hiring committees can spot generic AI-generated language immediately. Use every tool in your arsenal, but the cover letter still has to be about you, with specifics. Watch the video →

When should we trust AI output?

Be very careful about what I call the "deference threshold" — the point at which we stop questioning a system's output and accept it blindly. We trust calculators and spreadsheets completely because they've earned that trust. Generative AI has not. There's always the possibility of hallucination or output you wouldn't put your name on. If you're using AI, check the output. Don't let the time savings push that deference threshold closer. Watch the video →

How bad is iReady?

I've never used it personally, but the complaints are consistent: kids hate it, problems don't progress in a helpful way, students can't skip ahead on audio, and massive amounts of instructional time are being spent on the platform without evidence that it catches kids up. This is the pattern with EdTech at scale — the pilot looks promising, then the results evaporate when you go big. Watch the video →

What about ClassDojo and behavior points systems?

I see some value in having an early warning system to identify students who are struggling across multiple classes. But the points-and-rewards approach has gone awry in a lot of schools — excessive focus on prizes rather than teaching expectations. And it's usually the same kids getting infractions every time, which raises the question of whether the system is actually helping anyone. Watch the video →

Should kids practice handwriting on an iPad?

No. The iPad is the problem. When kids don't color, don't use pencils and crayons, and don't develop their handwriting muscles through real physical practice, the solution is not to put them on a screen more. Writing on a screen with a stylus is not the same as writing with a pencil — the mechanics and musculature don't overlap. Get your kid a big box of crayons. Watch the video →

Can an iPad app teach a 2-year-old to read?

Of course not. One company launched with that promise at $500 a month and quickly realized it doesn't work. Two-year-olds vary too much, and most aren't developmentally ready. The key ingredient in teaching a kid to read is teaching them — especially phonics — and you don't need any fancy technology to do it. My mom taught me with bread bag tags. Watch the video →

Why is skepticism so important in education?

Because skepticism is our only defense against marketing. EdTech companies spend millions selling unproven solutions to schools, and when you successfully market to one group, their testimonials replace actual evidence. Programs can survive for 30 years without producing results if the marketing is good enough. We have to demand evidence before adoption — that's professional responsibility, not negativity. Watch the video →

Is personalized learning actually better?

Not automatically. We take it for granted that personalization is always good, but most kids learn more in the social experience of a regular classroom. Look at the history of online schools — when you individualize everything and say "it's all on you," most students just do nothing. A small minority thrives, but large-scale personalization via technology has consistently underperformed well-designed whole-class instruction. Watch the video →

Does technology make learning too easy?

Yes, and that's the problem. "The person who does the work is the only person learning." Any technology that removes the effort from students is probably reducing learning. If the device is there to make work easy, give dopamine hits, and remove all struggle, we shouldn't expect much learning to come from it. Education is for the student, not for the profit of the companies selling the technology. Watch the video →

What about the argument that school should be fun and engaging?

Learning should be interesting — and a lot of what we teach is inherently interesting. But if we send kids the message that learning should always be easy, fun, and give a rush of dopamine, we're setting them up to fail. School cannot compete with Nintendo or social media for entertainment value, and it shouldn't try to. The work itself produces the learning. Watch the video →

How can technology help me get into classrooms more?

By removing friction from the parts of your job that keep you at your desk. Every time you streamline how you process email, document an observation, or write feedback, you're buying time that goes back into classrooms. Technology doesn't create time — but it eliminates the inefficiencies that steal it.

Specific examples: text expansion tools let you write common phrases and email responses in a fraction of the time. Digital walkthrough tools let you capture evidence, share notes, and track visits without paper systems that pile up. Task management apps let you process your to-do list in minutes instead of the constant re-sorting that happens with sticky notes and legal pads.

The best technology for instructional leadership is invisible to teachers and students. If you're visibly typing on a device during a classroom visit, you're undermining the relational quality of the visit. The tool should serve the visit, not compete with it. For richer observation beyond what a quick visit allows, Sibme enables video-based coaching — teachers record lessons and receive asynchronous timestamped feedback, extending your coaching reach without requiring both schedules to align.

When should school leaders use AI writing tools, and when shouldn't they?

AI is powerful for drafting — generating a first version of an email, a newsletter section, or evaluation language that you then edit in your own voice. It's especially useful for overcoming the blank-page problem on writing tasks that aren't your primary expertise.

Where it gets risky is in high-stakes judgment. AI can't observe a classroom. It can't assess whether a teacher's practice has genuinely improved. It can't weigh the political dynamics of a difficult staffing decision. When you use AI to draft evaluation narratives, you must ensure the result sounds like you, reflects your actual observations, and represents your professional judgment — not a generic summary.

The test is simple: would you be comfortable telling the teacher that AI helped you write this? If the answer is no, you're relying on it too heavily. The tool should extend your capacity, not replace your judgment. One area where AI works well for instructional leaders: Sibme generates AI Insights reports from observation video automatically, reducing documentation time while keeping your professional judgment in the loop.

How should schools approach technology adoption and training?

The same way they should approach any change initiative: start with early adopters, build success, and let that success convince the majority. Whole-staff technology training is the equivalent of bulk change — it treats everyone as if they're at the same starting point, which they never are. Some teachers are already proficient. Others are anxious beginners. A one-size session helps neither group.

The most effective approach is on-demand support and piloting. Let interested teachers try the tool, provide responsive help when they get stuck, and create opportunities for them to share what they've learned with colleagues. The most important questions about any technology emerge during actual use, not during training.

For school leaders specifically, model the technology yourself first. If you want your teachers to use a shared platform, use it visibly. If you want them to embrace digital communication tools, be excellent at using them yourself. Technology adoption in schools follows the same social dynamics as any other change — people adopt what they see working for people they trust.

What technology do you recommend for video-based instructional coaching?

Sibme is our recommended platform for video-based coaching. Unlike general video tools, it's purpose-built for professional learning — with timestamped annotations, AI-generated observation reports, coaching workflows, and goal-tracking built in. Teachers own their videos and control what they share, which supports the professional trust that makes coaching effective. The asynchronous model means coaching conversations happen in context, not just in a conference room.

See Tools We Recommend →

---