It has been a few months since I last posted an article on here. In those few months I have had one of the busiest periods in my life. My wife and I welcomed our third child, I accepted a new job, I am helping my parents sell my childhood home, and I have finished the Brewing Innovation book that will be out this fall.

The world of educational technology, AI, and innovation have not slowed down, in fact there are several topics I will be posting about in the coming weeks and months:

• Pope Leo XIV’s Encyclical about AI

• AI Detectors (Substack’s included)

• The ongoing screen-time debate

• Staying human in the age of AI

• And much more

Stay tuned for more from Brewing Innovation in the coming weeks! Thank you all again for your support.

Every year, education gets another wave of predictions about technology.

AI will replace teachers.
Screens are destroying learning.
Cyberattacks are getting worse.
Budgets are getting tighter.
Students are either over-connected or under-supported.

But CoSN’s U.S. State of EdTech 2026 report gives us something better than panic. It gives us a clear snapshot of where schools actually are.

And the big takeaway is this: EdTech in 2026 is not about whether schools use technology. That debate is over. The real question is whether schools have the systems, staffing, training, and leadership to use it well.

CoSN’s report is based on 607 survey responses from U.S. school district education technology leaders. The report looks at the major issues shaping K–12 technology right now: AI, cybersecurity, connectivity, staffing, devices, procurement, and community concerns around screen time. It also makes clear that technology leaders are no longer just keeping the Wi-Fi running. They are protecting student data, supporting instruction, enabling innovation, managing devices, helping families understand technology, and guiding responsible AI use.

That is a much bigger job than “fix the projector.”

Cybersecurity Is Still the Front Door

For the eighth straight year, cybersecurity remains the top priority for education technology leaders. That should not surprise anyone working in schools. Nearly every part of the modern school day depends on secure systems: student information, learning platforms, communication tools, finance systems, transportation, safety systems, and more.

The report also highlights a serious tension. Districts understand the importance of cybersecurity, but many lack the staff and dedicated budget to address it fully. Nearly two-thirds of districts identify insufficient cybersecurity staffing and lack of a dedicated budget as top barriers. Add rising cyber insurance costs and AI-enabled cyber threats, and the challenge becomes even more complex.

This is where school leaders need to stop treating cybersecurity as an “IT issue.”

• Cybersecurity is instructional continuity.

• Cybersecurity is student safety.

• Cybersecurity is family trust.

• Cybersecurity is leadership.

A district can have the best curriculum, the best devices, and the best AI tools, but if systems are not secure, everything else is vulnerable.

AI Has Moved From Panic to Policy

One of the most encouraging parts of the report is that districts are getting clearer about AI.

In 2024, 40% of districts had not yet defined their approach to AI. In 2026, that number dropped to 19%. Nearly half of districts now define their AI strategy by use case, and only 1% report banning AI. That matters. It shows that most districts are moving beyond the false choice of “ban it or embrace it blindly.” They are starting to ask better questions: What is the purpose? Who is using it? What data is involved? What guardrails are needed?

The policy movement is also significant. The percentage of districts with no generative AI guidelines dropped from 43% in 2025 to 21% in 2026. More districts now have guidance connected to acceptable use, academic integrity, AI-specific policy, and data privacy.

This matches what I have believed for a while:

The goal is not to create the perfect AI policy and call the work done. The goal is to build a living system of guidance, modeling, professional learning, and reflection.

AI changes too quickly for a policy binder to be the whole strategy. Schools need guardrails, but they also need people who understand the road.

AI Is Being Used First for Productivity

The report shows that education technology leaders are increasingly optimistic about AI’s potential, especially around productivity and personalized learning. District initiatives are heavily focused on training instructional staff to use instruction-focused generative AI tools, productivity suite platforms for administrators and teachers, and training staff to use those tools effectively.

That makes sense.

For many teachers and administrators, AI’s first real value is not some futuristic classroom redesign. It is getting time back.

Time to create a differentiated reading.
Time to summarize a long document.
Time to draft a parent communication.
Time to build a rubric.
Time to generate practice questions.
Time to reduce the blank-page burden.

Some people dismiss productivity as “not real innovation.” I disagree. When teachers are drowning in paperwork, compliance, communication, planning, and feedback, giving them time back is not a small thing. It is a human thing.

But productivity cannot be the finish line.

The next step is using AI to strengthen learning: tutoring, feedback, accessibility, writing support, research support, executive functioning support, and personalized practice. The report shows that districts are growing in this direction, but instructional platforms still trail productivity tools. That gap is worth watching.

AI should not just help adults work faster. It should help students think deeper, access more, and build confidence.

The Screen Time Conversation Needs More Nuance

The CoSN report also confirms something many of us are hearing in board meetings, family conversations, and social media debates: communities are concerned about technology use.

Nearly two-thirds of districts report moderate to high community concern about technology used for teaching and learning. At the same time, the report makes an important distinction. Community concern is higher around personal device use than around education technology used to enhance teaching and learning.

That distinction matters.

A Chromebook used for research, writing, accessibility, feedback, or creation is not the same as a phone used for endless scrolling. A station rotation model with purposeful tech is not the same as passive screen time. An AI tutor used with teacher guidance is not the same as a shortcut machine.

We hurt the conversation when we flatten everything into “screens are bad.”

The better conversation is:

• What is the instructional purpose?

• What should be analog?

• What should be digital?

• What tools support access?

• What habits are we teaching?

• What boundaries are we setting?

• How are we communicating this clearly to families?

CoSN’s report shows that many districts are already trying to support healthier technology choices. Common approaches include banning social media access on school-issued devices, teaching online safety, using 1:1 device agreements, limiting personal devices, teaching self-regulation, and designing instruction to limit unnecessary screen time.

That is the right direction: not panic, not unlimited access, but intentional design.

Share

Access Is Still Not Finished

One of the quieter but most important findings is around home connectivity.

The report shows that districts are still dealing with gaps in broadband and device access outside of school. Even more concerning, many districts do not have complete data about whether students have devices, broadband, or sufficient bandwidth at home. At the same time, district support for off-campus broadband has declined over the last few years.

This should make every school leader pause.

A 1:1 program does not automatically create equity.
A device does not guarantee access.
A login does not guarantee opportunity.

If a student cannot reliably connect at home, or if the district does not know who can connect at home, then digital learning can widen gaps instead of closing them.

This is especially important as AI tools become more common. If AI tutoring, feedback, and study tools become part of the learning ecosystem, then access to those tools cannot become another privilege gap.

AI literacy cannot be only for the students with reliable Wi-Fi, updated devices, and tech-savvy adults at home.

Staffing Is the Hidden Constraint

One of the biggest messages in the report is that schools have built strong technical foundations, but many still lack the staffing to support teaching and learning with technology.

The majority of districts report adequate staffing for core technical functions, such as networks, applications, and maintenance. But 58% report being understaffed when it comes to supporting technology used for teaching and learning.

That gap explains a lot.

Schools often buy the tool, launch the platform, send the email, and assume implementation has happened. It has not.

Implementation happens when teachers have time to learn, try, reflect, adjust, and see the tool connected to their actual classroom goals. That requires instructional technology support, coaching, professional learning, and leadership.

Without that support, technology becomes one more thing on the plate. With that support, it can become a lever for better teaching and learning.

Procurement Needs to Grow Up

Another important section of the report focuses on procurement.

Districts are doing some things well. Many require vendors to provide product safety information. But fewer districts consistently require information about other quality indicators such as evidence-based design, inclusivity, usability, accessibility, and interoperability.

The report also shows that 24% of respondents say their districts do not have an established process to assess interoperability during procurement. That is a problem because tools that do not integrate well can create more work, more risk, and more frustration.

Accessibility is another major concern. Only 11% of districts report that accessibility and WCAG 2.1 AA compliance are formally embedded in procurement policies, even though accessibility requirements affect all local education systems.

This is where leaders need to be disciplined.

Not every tool deserves a pilot.
Not every shiny platform solves a real problem.
Not every free tool is actually free when you consider privacy, support, interoperability, and instructional coherence.

Procurement is not just purchasing. It is instructional strategy, data privacy, accessibility, sustainability, and system design.

What This Means for School Leaders

The CoSN report confirms that the future of EdTech is not about chasing the newest tool. It is about building better systems.

For school and district leaders, I see five big moves:

1. Treat cybersecurity as a leadership priority.
It belongs in budget conversations, cabinet meetings, board updates, and professional learning.

2. Build AI guidance that lives beyond the policy document.
Policies matter, but modeling, training, and routines matter more.

3. Invest in instructional support, not just technical support.
If teachers are expected to use technology well, they need coaching, time, and examples that fit their classrooms.

4. Communicate clearly with families.
The screen time conversation is not going away. Families need simple, repeated messages about purposeful use, healthy boundaries, and how technology supports learning.

5. Tighten procurement.
Every tool should be evaluated for privacy, accessibility, interoperability, instructional value, evidence, cost, and support.

Share

The Bottom Line

The state of EdTech in 2026 is not chaos, but it is complicated.

AI is growing.
Cybersecurity is urgent.
Budgets are tight.
Staffing is stretched.
Families are asking questions.
Devices are everywhere.
Access still matters.
Procurement needs stronger guardrails.

But underneath all of that is a simple truth: Technology is not the strategy. The system is the strategy.

The districts that thrive will not be the ones with the most tools. They will be the ones with the clearest purpose, strongest guardrails, best communication, and deepest investment in people.

That is the work.

Not banning everything.
Not buying everything.
Not panicking.
Not chasing hype.

Just leading with purpose.

Stay curious. Stay caffeinated. ☕

Be sure to visit the NEW Brewing Innovation shop for coffee, mugs, and apparel!

Brewing Innovation Shop

Every few months, the education world seems to swing back into the same familiar argument:

• Screens are bad.

• Devices are ruining kids.

• Technology has failed classrooms.

• We need to go back to paper, pencils, and “real learning.”

And honestly, I understand where some of that frustration comes from.

As a school administrator, instructional technology leader, former special education teacher, and parent, I have seen the problems firsthand. I have seen students distracted by devices. I have seen attention spans feel shorter. I have watched students try to shortcut thinking with technology. I have also seen schools roll out tools without enough training, without a clear instructional purpose, and without the systems needed to make the work meaningful.

So yes, the concerns are real, but the conclusion that “classroom technology is failing our kids” is too simple. The better question is not whether technology belongs in classrooms. The better question is this: Is the technology helping students think, create, practice, access, and connect, or is it simply giving them another place to be distracted?

That difference is important.

A recent Inc. article by Adil Husain responded to a viral C-SPAN video where neuroscientist Dr. Jared Cooney Horvath argued that the money spent putting laptops and tablets in schools has contributed to a generation that is less cognitively capable than their parents. The video resonated because it tapped into a growing anxiety many people already feel: that screens have taken something from kids. But Husain’s main point is important. The reality is more complex than the viral argument suggests. The issue is not simply “school technology.” It is how, when, why, and under whose guidance that technology is being used.

That nuance is where the real conversation needs to happen.

Share

We Keep Confusing Classroom Tech With Digital Distraction

One of the biggest problems in this debate is that we keep putting very different things into the same bucket.

• A student scrolling TikTok under the desk is not the same as a student using a Chromebook to write, revise, collaborate, research, or receive feedback.

• A phone buzzing with notifications is not the same as a teacher-led blended learning station.

• A student passively watching random videos is not the same as a student using a scaffolded tool to access grade-level text.

• An AI tool doing the thinking for a student is not the same as an AI tutor asking questions, giving hints, and helping the student practice.

These are not small differences, they are the entire point. Husain calls out what is often missing in the anti-tech argument: we are blaming structured, teacher-led educational tools for many of the problems created by unregulated, addictive personal smartphone use. That distinction matters because if we misdiagnose the problem, we will create the wrong solution.

And in education, we have a long history of doing exactly that.

We see a real problem, panic and then create a blanket rule. Then we wonder why the deeper issue remains.

If students are distracted, the answer cannot simply be “remove every device.” Sometimes that may be the right move for a specific lesson or classroom moment. But as a system-wide philosophy, it is too blunt.

The better answer is to teach students how to use technology with purpose, boundaries, and responsibility. That is harder than banning it. It is also much more important.

The Data Is Sloppy, But the Concern Is Real

This is where we need to slow down. A lot of the “technology is failing our kids” argument is built on data that gets flattened into a headline. Critics often point to declining test scores, drops in cognitive performance, and international assessment data as proof that classroom technology caused the problem. But that is not what the data clearly says. The problem is correlation being treated like causation.

For example, Husain points out that NAEP scores in math and reading peaked around 2013 and then began declining before the pandemic. That timeline lines up more closely with the rapid rise of personal smartphone ownership among teens than with the slower rollout of school-issued laptops and 1:1 programs. The article also notes that PISA data showing lower scores with more computer use is often misread. In some cases, struggling school systems invest in technology after problems already exist, which means the technology may be responding to the problem, not causing it.

That distinction matters. When people say, “The data proves screens are bad,” we need to ask better questions:

• Was the technology teacher-led or student-led?

• Was it connected to curriculum or just extra screen time?

• Was it used for creation, feedback, practice, and access, or passive consumption?

• Was the device a learning tool or a distraction machine?

• Was the teacher trained and supported, or simply handed another platform?

The data does not say, “All classroom technology is bad.”

• It says poor implementation is bad.

• It says unstructured device use is bad.

• It says personal smartphone distraction is bad.

• It says hardware without pedagogy does very little.

And it says that when technology is integrated into instruction with purpose, structure, and teacher guidance, it can support learning. That is not a reason to panic, but it is a reason to design better.

Hardware Without Pedagogy Is Just Hardware

One of the strongest points in the article is simple: Technology by itself does not teach. Putting a device in a student’s hands does not automatically create innovation. It does not automatically create engagement. It does not automatically create deeper learning.

• A Chromebook is not a lesson plan.

• An app is not a strategy.

• A platform is not a teacher.

• AI is not pedagogy.

Technology only becomes powerful when it is connected to strong instructional design.

The article points out that simply handing out laptops did not produce meaningful cognitive effects in one major long-term study. But when technology was paired with structured teaching, there were measurable gains. The takeaway is not “screens work” or “screens don’t work.” The takeaway is that implementation matters.

That has been my experience, too.

• When tech is used as a digital babysitter, it fails.

• When tech is used to digitize weak instruction, it fails.

• When tech is introduced without training, clarity, or purpose, it fails.

But when technology is used to support teacher-led instruction, expand access, create feedback loops, and give students more ways into the learning, it can be powerful. The issue is not the device, but the design.

Balance Is Not Anti-Tech or Pro-Tech. It Is Good Teaching.

The conversation around classroom technology often gets framed as a choice between two extremes. Either you are “pro-tech” and want students on screens all day, or you are “anti-tech” and want to remove devices completely.

That is a false choice. Good teaching has never worked that way.

• Some learning belongs on paper.

• Some learning belongs in conversation.

• Some learning belongs in hands-on work.

• Some learning belongs in movement, debate, drawing, building, and reflection.

• And yes, some learning belongs on a device.

The skill of teaching is knowing the difference. There are times when students should close the Chromebook and talk to each other. Then there are times when handwritten notes are better for processing and memory. There are also times when a printed text leads to deeper annotation and focus. Then, there are also times when technology gives students access they would not otherwise have:

• A student with a reading disability may need text-to-speech.

• A student with processing challenges may need to replay directions.

• A multilingual learner may need vocabulary support or translation.

• A student who struggles to get started with writing may need a scaffold, sentence frame, or model.

A student who is ready for extension may need a tool that helps them test an argument, simulate a problem, or create something more complex.

This is why I push back against the “no screens, no devices” crowd. Not because I think all screen time is good. It is not. But because I know that for many students, especially students with disabilities, technology is not a luxury. It is access. I wrote about this same idea in Enough With the Panic: What the “No Screens, No Devices” Crowd Is Missing: the problem is not technology itself, but how we use it, and when used well, technology can amplify student thinking rather than replace it.

That is the balance we need.

Not more screen time.
Not less screen time.
Better screen time.

The Real Risk Is Cognitive Offloading

The critics are not wrong about everything.

One of the biggest risks with modern EdTech, especially AI, is cognitive offloading. That happens when the tool does too much of the thinking for the student.

This is where we need to be honest.bSome tools are being built to make learning feel frictionless. But learning is not supposed to be completely frictionless. Students need productive struggle. They need to wrestle with ideas, make mistakes, revise their thinking, and build confidence through effort.

If AI writes the essay, solves the problem, summarizes the reading, and generates the final product without the student engaging deeply, that is not learning.

That is outsourcing. But again, the answer is not to ban the tool. The answer is to design better learning.

• Instead of asking AI to write the essay, students can ask it to give feedback on their thesis.

• Instead of asking it for the answer, students can ask it to explain the first step.

• Instead of using it to skip reading, students can use it to preview vocabulary, generate questions, or check understanding after reading.

• Instead of allowing students to use AI secretly, we can teach them to document how they used it.

That is AI literacy. And AI literacy is quickly becoming part of what it means to be educated.

AI Tutoring Works Best When It Is Human-Guided

This same lesson applies to AI tutoring.

The promise of AI tutoring is real, but we need to be careful with the hype. AI tutors are not magic. They are not replacements for teachers, interventionists, special educators, or actual human tutors.

The strongest model is not “student alone with a chatbot.” The strongest model is AI plus human support.

That means AI works best when it is connected to a teacher, tutor, or intervention model where the student’s actual needs are known. The student is not just getting random help from a machine. They are getting practice, feedback, hints, explanations, and scaffolds that fit into a larger instructional plan designed by a human being. That is where the effect-size conversation becomes much more important.

The University of Chicago Education Lab studied a high-dosage tutoring model that blended in-person tutoring with technology-supported practice. The model reduced cost and staffing needs while still improving math achievement. The related NBER paper found math gains of 0.23 standard deviations, nearly as large as a more intensive 2-to-1 tutoring model, while lowering per-pupil costs. (University of Chicago Education Lab)

That is the balance we should be paying attention to. Not AI instead of tutoring. Not AI instead of teaching. AI as part of a structured support system.

Brookings makes a similar point in its review of generative AI tutoring. The evidence is promising, but the best models rely on responsible design, safeguards, personalization, feedback, and a hybrid approach where human teachers and AI work together. Brookings also notes that generative AI tutoring still raises concerns around accuracy, pedagogical judgment, and possible dependence, which means the teacher’s role does not become less important. It becomes more important. (Brookings)

That aligns with what we already know about good teaching.

• Students need productive struggle.

• They need feedback.

• They need practice.

• They need encouragement.

• They need someone who knows when to push, when to scaffold, and when to step in.

AI can help with that. But it cannot replace the human judgment behind it.

For a student with an IEP, AI tutoring only works well if it is connected to their goals, accommodations, reading level, executive functioning needs, and confidence. For a student who is behind in math, AI support needs to reinforce the skill progression the teacher is already building. For a high-performing student, AI should extend thinking, not simply accelerate completion.

The model matters. The supports matter. The teacher still matters most.

What Schools Should Do Instead of Panic

If we want healthier technology use in schools, we need to stop treating this as a device problem and start treating it as a leadership and instructional design problem.

Here is what that looks like:

1. Start with the learning goal.
Before using any tool, ask: What are we trying to help students understand, practice, create, or demonstrate?

2. Define when technology helps and when it does not.
Some tasks are better analog. Some are better digital. Teachers and students need permission to make that distinction.

3. Teach students how to use tools responsibly.
Digital citizenship and AI literacy cannot be side conversations anymore. They are part of modern literacy.

4. Build routines for reflection and verification.
Students should be able to explain how they used technology, what they changed, what they checked, and what thinking remained their own.

5. Invest in teacher training.
EdTech companies need stronger evidence, but schools also need stronger implementation. Devices without professional learning are a recipe for frustration.

6. Protect attention.
We should absolutely create boundaries around phones, notifications, and unstructured digital distraction. Focus is a skill, and schools should protect it.

7. Keep humans at the center.
The teacher is still the most important technology in the room.

The Future Is Not Screen-Free. It Is Purposeful.

I do not want classrooms where students stare at screens all day. I also do not want classrooms where we remove powerful tools because adults are overwhelmed by the complexity of teaching students how to use them well. The future of education should not be screen-obsessed. It should not be screen-free either. It should be purposeful.

It should be a classroom where students know when to write by hand and when to collaborate digitally. Where AI is used as a tutor, coach, and feedback partner, not a replacement for thinking. Where teachers use technology to reduce friction, personalize support, and create more time for small-group instruction and human connection. Where students learn that devices are not toys, shortcuts, or distractions, but tools that require judgment.

That is the conversation we should be having. Not “Should we ban technology?” But: How do we design learning so technology supports the human work of teaching and learning?

Because technology is not failing our kids. Poor design is. Weak implementation is. Unstructured distraction is. The absence of clear expectations is.

The research is not telling us to ban technology. It is telling us to stop confusing access with impact. Devices alone do not improve learning. AI alone does not tutor well. But when technology is paired with strong pedagogy, human guidance, clear routines, and real supports for students, it can absolutely make learning more accessible, more personalized, and more human.

That is the balance we need.

Stay curious. Stay caffeinated. ☕

Works Cited

• Bhatt, Monica P., Jonathan Guryan, Salman A. Khan, Michael LaForest-Tucker, and Bhavya Mishra. “Can Technology Facilitate Scale? Evidence from a Randomized Evaluation of High Dosage Tutoring.” National Bureau of Economic Research, Working Paper No. 32510, May 2024.

• Burns, Mary. “What the Research Shows About Generative AI in Tutoring.” Brookings Institution, 27 Jan. 2026.

• Husain, Adil. “A Viral C-SPAN Video Claims Classroom Tech Is Failing Our Kids. But the Reality Is Far More Complex.” Inc., 16 Apr. 2026.

Be sure to visit the NEW Brewing Innovation shop for coffee, mugs, and apparel!

Brewing Innovation Shop

I decided to test it out by having NotebookLM create a video about designing a balanced blended learning lesson. What stood out to me was not just that it made a video. It was how quickly it turned a set of ideas into something clear, usable, and easy to share. Check out the video in this post.

That matters because in education, the challenge is often not coming up with good ideas. The challenge is finding the time to turn those ideas into resources that teachers and students can actually use.

That this feels promising. I have already seen NotebookLM as a valuable tool for organizing information, summarizing content, and helping educators make sense of complex material. But this new feature opens another door. It gives teachers, coaches, and school leaders a way to turn source material into something more dynamic and accessible.

Share

For blended learning, that has real potential.

A lot of people still misunderstand blended learning. It is not just students on devices. It is not just station rotation for the sake of saying you used tech. Good blended learning is intentional. It creates the right mix of teacher instruction, collaboration, independent work, reflection, and purposeful technology use.

A balanced blended learning lesson should still be grounded in strong pedagogy. The tech should support the learning goal, not distract from it.

That is why this NotebookLM update caught my eye. A teacher could use it to turn planning documents, notes, readings, or unit resources into a quick video overview for students. That video could be used to introduce a concept, support a station in a blended lesson, help absent students catch up, or even model a strategy during professional development.

And that is where I think AI is most useful in schools.

The best AI tools do not remove thinking. They remove friction. They help teachers spend less time packaging information and more time designing meaningful learning experiences. They give students more ways to access content without lowering the level of rigor.

That is the opportunity here.

• For students, this could mean more entry points into learning.

• For teachers, it could mean faster creation of useful supports.

• For school leaders, it could mean better ways to share models of instruction and scale good practice.

Of course, the tool itself is not the point. We still need to ask the same questions we should always ask with technology:

• What is the learning goal?

• Does this improve clarity or access?

• Does this save time in a meaningful way?

• Does this strengthen the lesson, or just make it look more high-tech?

If we lose sight of those questions, then this becomes just another shiny feature.

But if we use it well, it can help us design stronger blended learning experiences that are more flexible, more accessible, and more human.

This update is worth paying attention to. NotebookLM is starting to feel less like a simple note-taking tool and more like a real design partner for educators. And if that continues, it could become one of the most practical AI tools in education.

The future of AI in schools should not be about replacing good teaching.

It should be about helping great teachers plan better, explain more clearly, and create stronger learning experiences for students.

This feature feels like one more step in that direction.

Stay curious. Stay caffeinated. ☕

Also, be sure to visit the Brewing Innovation shop for coffee, mugs, and apparel!

Brewing Innovation Shop

Transparency note: This article was developed with the support of AI tools for drafting and organization. All instructional perspectives, examples, and final decisions reflect my professional experience and judgment.

Lately, there has been a growing chorus in education that sounds something like this:

• “We need to go back.”

• “Back to pencil and paper.”

• “Back to fewer screens.”

• “Back to the old ways.”

And look, I understand where that reaction comes from.

We have all seen classrooms where students are overexposed to screens. We have seen passive tech use. We have seen digital worksheets masquerading as innovation. We have seen students distracted, disengaged, and too used to quick answers. Those concerns are real.

But too often, people take a real problem and attach it to the wrong solution. The answer is not to run backward. The answer is to move forward more thoughtfully. Because let’s be honest. An 80-minute lecture is not the answer either.

If we are serious about student engagement, deeper thinking, and better learning, we have to stop pretending that old automatically means better and new automatically means worse. Neither one is true. A pencil is not automatically rigorous. A Chromebook is not automatically shallow. An AI tool is not automatically harmful. And a long lecture is not automatically high-quality instruction just because it looks more traditional.

Good teaching has never been about the tool alone. It has always been about the design.

The Real Problem Is Not the Device

The real problem is not that schools used technology. The problem is that too often, technology was used without purpose.

If all we do is replace a worksheet with a digital worksheet, that is not innovation. If students sit passively on a device clicking through low-level tasks, that is not meaningful learning. If AI is used to shortcut thinking instead of support it, that is a problem too.

But none of that means we should swing to the other extreme and pretend that going back to all paper, all lecture, and all old-school methods will somehow fix everything.

It will not.

Students do not need less opportunity to think, create, revise, collaborate, and access support. They need better-designed learning experiences. They need teachers who know when to use discussion, when to use direct instruction, when to put something on paper, when to leverage a device, and when AI can actually deepen the work instead of replacing it.

That is the job.

The False Choice Is Hurting the Conversation

Too much of this debate has turned into a false binary.

Either you are for screens or against them.
Either you are pro-AI or anti-learning.
Either you are innovative or you believe in basics.

That framing is weak, and it misses what good educators know from experience.

The best classrooms are rarely all one thing.

Sometimes students should write by hand. Handwriting can support memory, slow thinking down, and reduce distractions. Sometimes students should annotate on paper, sketch ideas, or work through math without a device in front of them.

Sometimes students should be in discussion, wrestling with ideas out loud.

Sometimes they should be using a Chromebook to research, collaborate, draft, revise, or create something that would be harder to do any other way. And sometimes AI can help give them the support they need to get unstuck, rehearse a skill, receive feedback, or access material at the right level.

That is not a contradiction. That is thoughtful teaching.

Let’s Stop Pretending the Old Way Was Always Better

There is a kind of nostalgia in education that sounds comforting, but it is often incomplete.

People talk about going back to pencil and paper as if the old model was always producing deep engagement, curiosity, and strong learning. But a lot of what some people want to return to was not especially effective either.

Rows of students sitting quietly for long stretches. One-size-fits-all instruction. Heavy reliance on lecture. Limited differentiation. Students who struggled quietly and fell behind without support.

That was not some golden age.

An 80-minute lecture does not become rigorous just because there is no screen involved. In many cases, that kind of passive learning is exactly what loses students. We cannot say we care about engagement and then defend instruction that asks students to sit and absorb information for long periods with little interaction, little feedback, and little ownership.

We know better than that.

Students learn more when they are thinking, discussing, practicing, creating, reflecting, and getting timely feedback. That is not a technology argument. That is a learning argument.

AI Can Help Teachers Design Better Learning

This is where I think the conversation gets more interesting.

AI is not here to replace the teacher. It is not here to do the thinking for students. But it can help teachers build more engaging lessons and stronger support systems.

• A teacher can use AI to quickly generate multiple versions of a text at different reading levels.

• A teacher can create checks for understanding in minutes instead of spending hours building them from scratch.

• A teacher can develop discussion prompts, exemplars, vocabulary supports, and retrieval practice more efficiently.

• A teacher can build a station rotation model where one station includes AI-supported tutoring or guided practice while the teacher works with a small group.

That is not lowering the bar. That is removing friction so the teacher can spend more time where they matter most. Instead of using all of their time creating materials, teachers can use more of their time analyzing student thinking, conferencing with learners, leading discussion, and giving human feedback.

That matters, because the goal should not be more screen time for the sake of it. The goal should be better teaching and stronger learning.

The Better Question

The question is not: Should we go back to the old ways?

The better question is: How do we build classrooms that are more balanced, more engaging, and more human?

That might mean students start class with handwritten retrieval practice. Then move into a short mini-lesson. Then rotate into collaborative work, teacher-led small group instruction, and AI-supported practice or feedback. Then close with reflection or discussion.

That is not screen obsession. That is intentional design. That is also a far cry from an 80-minute lecture or a classroom where every task is done on a laptop.

Balance is not weakness. It is wisdom.

What Schools Should Actually Be Fighting For

If we really want better classrooms, here is what we should fight for.

We should fight for better instructional design, not panic. Too much of this conversation has been driven by headlines, hot takes, and frustration instead of thoughtful teaching. The issue is not whether a device is in the room. The issue is whether the learning is well designed. A bad lesson on a Chromebook is still a bad lesson. A weak worksheet on paper is still a weak worksheet. What matters is whether students are being asked to think, wrestle with ideas, create, revise, and reflect. As I have written before, technology should be used in service of a goal, not as the goal itself. The right question is not, “Is there a screen involved?” The right question is, “What is the learning challenge we are trying to solve, and is this the best way to solve it?”

We should fight for more teacher training, not blanket bans. Schools do not get better when we remove tools out of fear. They get better when we help teachers use tools with confidence, clarity, and purpose. If we want stronger classrooms, we need to invest in the adults in the room. Teachers need time to collaborate, experiment, revise, and learn from one another. They need practical support on how to design blended lessons, how to use AI for scaffolding and feedback, and how to make thoughtful choices about when to go digital and when to stay analog. Without that support, even good tools can fall flat. With it, technology becomes a way to remove friction so teachers can spend more time where they matter most: in discussion, in feedback, in relationships, and in the human work of teaching.

We should fight for clearer AI guardrails, not avoidance. Avoidance is not leadership. Students are already encountering AI, both in and out of school. Pretending it is not there does not protect learning. It just guarantees that students will use it without guidance. Clear guardrails matter because they help us move from fear to responsibility. Students need to know what appropriate use looks like. Teachers need routines for modeling, reflection, and verification. Schools need language around transparency, ethical use, and accountability. That is how we build AI literacy. Not through one assembly, one policy, or one banned website, but through consistent habits that teach students how to question outputs, reflect on their choices, and use tools without surrendering their thinking.

We should fight for purposeful technology use, not digital overload. Being pro-technology should never mean advocating for nonstop screen time. It should mean advocating for the right tool at the right time for the right reason. Some learning belongs in conversation. Some belongs in handwriting. Some belongs in collaborative digital spaces. Some belongs in AI-supported practice, where students can get immediate explanations, targeted scaffolds, or help organizing their thinking. Thoughtful use is balanced use. It is not about filling every moment with a device. It is about making choices that deepen learning, increase access, and create more room for great teaching.

We should fight for a healthy balance of analog and digital experiences based on the learning goal. Students still need to annotate, sketch, read on paper at times, and write by hand. They also need to research, collaborate, create, revise, and access supports that paper alone cannot provide. A balanced classroom is not confused. It is intentional. It knows that handwritten notetaking may help in one moment, while a Chromebook may open doors in another. It knows that a small-group discussion may do more for learning than another slideshow. It knows that an AI tool can help a struggling learner get unstuck without lowering the level of the work. Great classrooms are not all analog or all digital. They are responsive, flexible, and designed around what students actually need.

Students still need to read deeply. They still need to write, think independently, and talk to one another. They still need strong teachers leading the room. None of that is old-fashioned. None of that is optional. In fact, those needs become even more important in a world full of distraction, automation, and easy answers. The goal of thoughtful technology use is not to replace those fundamentals. It is to strengthen them. AI can help a student rehearse before writing. A digital tool can help a struggling reader access grade-level content. A blended model can free the teacher to work with a small group while other students practice, collaborate, or receive scaffolded feedback. In the best classrooms, technology does not shrink the human side of learning. It creates more space for it.

That is the fight worth having. Not screens versus paper. Not old school versus new school. Not panic versus hype. The real fight is for classrooms that are better designed, more intentional, more accessible, and more human. That is what students deserve. And that is where school leaders and teachers should keep their focus.

Final Thought

I am not interested in defending bad screen use. I am not interested in pretending every shiny new tool belongs in a classroom. I am not interested in replacing good teaching with automation.

But I am also not interested in confusing nostalgia with rigor. Going backward is easy. It feels safe. It sounds simple.

Leading well is harder. It means rejecting extremes. It means choosing the right tool at the right time for the right purpose. It means protecting what is timeless in education while also embracing what can make learning better, more accessible, and more engaging.

Schools do not need to go back. They need to get better. And that will not happen through fear, headlines, or over-correction.

It will happen through thoughtful, balanced, innovative teaching that keeps students at the center.

Stay Curious. Stay Caffeinated. ☕

Also, be sure to visit the NEW Brewing Innovation shop for coffee, mugs, and apparel!

Brewing Innovation Shop

Transparency note: This article was developed with the support of AI tools for drafting and organization. All instructional perspectives, examples, and final decisions reflect my professional experience and judgment.

Spend enough time in a school building and you realize something quickly: Kids are still kids.

They still forget their Chromebooks.
They still rush through assignments to get to lunch.
They still avoid raising their hands when they’re unsure.
They still need structure, repetition, feedback, and adults who believe in them.

Artificial intelligence hasn’t changed that.

What it has changed is the range of tools we have to support them.

Right now, the conversation around AI in schools swings between extremes. Some see it as a threat to learning. Others treat it like a silver bullet. Both miss the point.

At the building level, AI isn’t about replacing instruction or chasing innovation. It’s about access. It’s about rigor. It’s about consistency. And ultimately, it’s about leadership.

The real question for principals and assistant principals isn’t, Should we allow AI?

It’s this:

How do we leverage AI to support both struggling learners and high flyers, without losing what makes school human?

Let’s approach that from a building leader’s lens.

Start With Students, Not Subscriptions

The biggest mistake schools can make is beginning with the tool.

“What platform should we buy?”
“What filter should we install?”
“How do we block this?”

Those are compliance questions. Leadership starts somewhere else.

Where are students getting stuck?
Where are teachers overwhelmed?
Where are we unintentionally leaving learners behind—both those who struggle and those who soar?

AI is not the strategy. Access is. Rigor is. Clarity is. AI is simply a lever that, if used well, can strengthen all three.

Struggling Learners: Access Without Stigma

In every building, there are students who quietly fall behind.

The ones who shut down when the text becomes dense.
The ones who freeze when asked to write more than a paragraph.
The ones who stop trying after one wrong answer in math.
The ones who never ask questions publicly.

Traditionally, support for these students has required enormous teacher labor: modified materials, pull-out services, extra help blocks, individualized scaffolds. All necessary. All time-intensive.

AI doesn’t replace those supports. But it can extend them into the classroom in ways that are immediate and quiet.

Imagine a student who struggles with reading comprehension. Instead of waiting for the next support period, they can:

• Ask for a simplified version of the passage.

• Request vocabulary definitions in context.

• Break a multi-step prompt into manageable chunks.

That’s not cheating. That’s access.

When a student who struggles with writing can receive guided sentence starters or model paragraphs and then revise in their own words, that’s not avoidance. That’s scaffolding.

The key is structure. At the building level, we must be explicit: AI may assist with understanding. It may not replace thinking.

Simple shared routines can make all the difference:

• “Explain the concept in your own words after using AI.”

• “Highlight what you changed from the AI’s suggestion.”

• “Identify one error or limitation in the AI response.”

These aren’t tech strategies. They’re literacy strategies.

Used this way, AI keeps students in the room. It reduces the need for constant removal. It supports Tier 1 instruction so that students don’t disengage simply because they can’t access the material quickly enough.

Just as important, it protects teacher time.

If AI can help generate leveled texts, draft first-pass feedback, or create guided practice questions, teachers gain something invaluable: time for conferencing, encouragement, and relationship-building.

And relationships—not algorithms—are what change trajectories.

Share

High Flyers: Challenge Without Plateau

We talk frequently about struggling learners. We talk less about students who quietly plateau.

High-performing students often comply beautifully. They finish early. They earn strong grades. But too often, they are rarely stretched beyond the curriculum’s minimum expectations.

AI can change that, but only if we frame it correctly. In wrestling, there are no shortcuts. You don’t get better by avoiding hard reps. You improve by drilling, adjusting, refining, and pushing beyond comfort. AI should amplify that grind, not remove it.

For high flyers, AI can become an intellectual sparring partner. A student can draft an argument and ask AI to generate counterclaims. They can stress-test a thesis against alternative interpretations. They can simulate debate from multiple historical or scientific perspectives. But the critical step comes next: evaluating and refining.

One powerful routine looks like this:

1. The student writes independently.

2. AI critiques or challenges the work.

3. The student evaluates the critique.

4. The student revises and explains their reasoning.

Now we’re not using AI to shortcut learning. We’re using it to deepen metacognition. High performers don’t need easier work. They need harder thinking. AI, structured properly, can help provide it.

The Leader’s Role: Culture Over Compliance

At the building level, AI implementation cannot be about fear or frenzy.

Over-policing creates underground use.
Over-permitting creates chaos.

Leadership requires modeling. If we want teachers to integrate AI thoughtfully, we must show what thoughtful use looks like.

That might mean:

• Sharing how you used AI to draft a memo—and how you revised it.

• Demonstrating how to critique an AI-generated agenda.

• Publicly discarding portions of an AI response that don’t meet standards.

When leaders normalize discernment, teachers follow. Equally important is aligning around habits, not apps.

Tools change quickly. What should remain constant are the building’s expectations:

• Verify information.

• Reflect on revisions.

• Own your thinking.

• Use AI to strengthen—not replace—your work.

When these norms are consistent across departments, students receive a coherent message. And coherence builds culture.

Discipline Still Matters

AI does not remove the need for discipline. In fact, it demands more of it.

Students must learn when to use support and when to struggle. They must learn to question outputs, check sources, and revise independently.

Just like on a wrestling mat, you can’t outsource conditioning. You can’t fake reps. And you can’t blame someone else when you lose.

School should function the same way. AI may provide tools. But effort, accountability, and growth still belong to the student. Our responsibility as leaders is to build systems where that remains clear.

A Practical Framework for Building Leaders

If I were approaching AI integration as an assistant principal, I would focus on five pillars:

• Clarity: Establish clear expectations for ethical and instructional use.

• Consistency: Develop shared routines across grade levels so students receive aligned messaging.

• Coaching: Provide ongoing support for teachers—not one-off professional development.

• Communication: Engage families transparently. Explain how AI is used as a support tool, not a substitute.

• Culture: Ensure AI enhances what we already value: relationships, rigor, reflection, and growth.

When these pillars are in place, AI becomes manageable. Intentional. Sustainable.

Without them, it becomes noise.

The Bigger Truth

Artificial intelligence does not make students more mature. It does not make them less human. It does not eliminate struggle. What it can do is widen the range of access and extend the reach of effective teaching.

The struggling learner gets more reps.
The high flyer gets more challenge.
The teacher gains time for meaningful feedback.

But the heart of school remains unchanged.

Kids still need adults who hold high expectations.
They need boundaries.
They need consistency.
They need encouragement.
They need to feel known.

AI does not replace those needs. It amplifies our ability to meet them, if we lead wisely. Because no matter how advanced the tools become…

Kids are still kids.

And they deserve leaders who understand both the world they’re entering and the timeless principles that help them grow.

In the world of education, innovation often feels like being handed a shiny new espresso machine without being given the beans or the time to learn how to use it. We are currently standing in a digital landscape that is shifting faster than we can print a syllabus. For many educators, the reaction to new technology isn’t “Let’s go!”—it’s “Not one more thing.”

Leading in the age of AI requires more than just a tech budget. It requires a shift from chasing “shiny object syndrome” to building a culture of purpose-driven technology. To get people on board, you have to stop talking about the “future” and start talking about Tuesday morning.

1. Anchor Innovation in Real Pain Points

The quickest way to lose a staff is to introduce a tool that doesn’t solve a problem they actually have. Innovation should never be a standalone initiative; it must be grounded in need. When we focus on the “shiny” part of tech, we get resistance. When we focus on the “pain” part of the job, we get curiosity.

• The Compliance Test: Start where the administrative burden is heaviest. If a tool can turn “documentation fatigue” into clarity—reducing the time spent drafting complex, compliance-heavy documents like IEPs or progress reports from three hours to under 90 minutes—you aren’t just “innovating”; you are giving a teacher their Sunday back.

• The Differentiation Gap: We know our students have widening skill gaps, especially in high-stakes areas like writing and math. Use AI to provide the 1:1 scaffolding that a human teacher simply doesn’t have the bandwidth to provide for 30 students at once.

• Acknowledge the Exhaustion: Resistance is often just exhaustion in disguise. If a tool doesn’t save time or mental energy, it isn’t the right tool for right now. Our job as leaders is to clear the plate, not pile more on.

2. The Strategic Framework: Blueprint, Guide, and Toolbox

To get people on board, they need to see where they are going. I frame innovation through three critical lenses that align our “why” with our “how.” Without this, tech is just a series of random acts.

3. Integration Over Addition

The mantra for any leader should be: Integration, not addition. Our goal is to embed technology into existing routines so teachers don’t feel like they are starting over.

• Reduce Friction: Prioritize tools that live where your teachers already work—like Chrome extensions that appear directly in their existing document editors or LMS. If they have to open a new tab and remember a new password, you’ve already lost half the room.

• Protect the Time: You cannot ask for experimentation without giving people the space to try. Use “Directed Planning” sessions to allow staff to “play” with tools in a low-stakes environment, specifically focused on their upcoming units.

• Model the Workflow: If you want teachers to use AI, you must lead by example. Use it to draft your faculty meeting agendas, summarize complex state regulations, or streamline your community newsletters. When administrators model responsible use, staff see these tools as frameworks for efficiency, not just classroom gimmicks.

4. Stay Human, Stay Ethical

Innovation without responsibility is unsustainable. As we push forward, we must remain firm on privacy and the human element. The goal of AI isn’t to replace the teacher; it’s to automate the “robotic” parts of the job so they can be more human.

• The Privacy Guardrail: Lead with safety. Train staff to strictly avoid personally identifiable information (PII) in prompts. Use fictionalized composites or anonymized patterns. When people feel safe, they are more willing to be bold.

• Professional Judgment: Always remind your team that AI is a partner, not a replacement. The teacher’s voice, empathy, and professional intuition must always come first.

• Feedback Loops: Create a system where staff can give honest, unfiltered input. If a tool feels clunky, irrelevant, or adds more stress than it removes, have the courage to adjust the approach or walk away from it.

The Bottom Line

Change takes time to steep. Don’t wait for the “perfect moment” or the “perfect tool.” Identify a single pain point, test a solution, and build from there. Celebrate the small wins loudly—in meetings, newsletters, and hallway chats.

When we invest in building understanding rather than just introducing tools, our staff responds with curiosity instead of resistance. Let’s build better schools—one bold move at a time.

Stay Curious. Stay Caffeinated. ☕

Independent practice is often treated as the simplest phase of instruction. Students work on their own. Teachers circulate. Learning is assumed to occur.

In reality, independent practice is one of the most fragile moments in the learning process—particularly in classrooms where artificial intelligence is readily available. When poorly designed, independent practice becomes task completion rather than learning. When designed well, it is where understanding deepens, fluency develops, and ownership emerges.

Research from Rosenshine, Hattie, Sherrington, and Tucker is clear on one point: independence does not mean absence of structure. It means responsibility earned through intentional instructional design.

What Independent Practice Is Meant to Do

Rosenshine was explicit that independent practice should occur only after students demonstrate success during guided practice. Its purpose is consolidation, not exploration or discovery. Students apply skills they already understand to strengthen accuracy, automaticity, and retention (Rosenshine, 2012).

Tom Sherrington reinforces this by emphasizing that practice should be deliberate and aligned. Independent work should surface thinking, reveal misconceptions, and provide teachers with evidence to inform next steps—not simply fill time.

Independent practice, therefore, must be:

• tightly aligned to the learning target,

• within students’ current capability,

• and designed to produce usable evidence of learning.

AI complicates this phase—but it does not invalidate it.

What Hattie’s Research Adds to the Conversation

John Hattie’s Visible Learning research sharpens our understanding of independent practice by focusing on impact rather than activity. Hattie’s work highlights several influences that are especially relevant here:

• Practice has a moderate effect size when it is purposeful and feedback-rich.

• Feedback is one of the highest-impact influences on learning—but only when it is timely, specific, and actionable.

• Metacognition and self-regulation significantly influence student achievement when students are taught how to monitor their own learning (Hattie, 2009; Hattie, 2012).

From a Visible Learning perspective, independent practice must make learning visible—to students and teachers. If AI obscures what students actually know or can do, it undermines impact, regardless of efficiency.

Share

Where AI Can Support Independent Practice

When aligned carefully to instructional goals, AI can support independent practice in limited but meaningful ways.

1. Retrieval and Self-Checking (After Attempt)

Independent practice should strengthen memory and understanding. AI can support this after students attempt a task by:

• generating additional practice questions aligned to the objective,

• helping students check answers against criteria,

• or explaining why an answer is incorrect once the student commits to a response.

This aligns with Rosenshine’s emphasis on practice and Hattie’s focus on feedback—but only when AI follows student effort, not replaces it.

2. Feedback That Prompts Improvement, Not Submission

Hattie emphasizes that effective feedback answers three questions:
Where am I going? How am I going? Where to next?

AI can assist by providing rubric-aligned feedback that:

• highlights strengths,

• identifies gaps,

• and suggests next steps for revision.

However, Catlin Tucker consistently cautions that feedback should preserve student ownership. If AI rewrites student work, learning disappears. Feedback must prompt thinking, not completion (Tucker, 2016).

3. Reflection and Metacognition

Independent practice is also where students should develop self-regulation. AI can support this by prompting reflection:

• “Which strategy did you use?”

• “Where did you struggle?”

• “What would you try differently next time?”

These prompts align directly with Hattie’s findings on metacognition and help students internalize responsibility for learning.

Where AI Does Not Belong in Independent Practice

This boundary must be explicit.

There are specific moments in independent practice where AI should not be used at all, because its presence undermines the very purpose of the task.

1. First Attempts at Skill Application

When students are practicing a skill independently for the first time, AI use often replaces productive struggle with surface-level success. This prevents teachers from seeing misconceptions and denies students the opportunity to learn from errors.

Rosenshine was clear: errors must be detected early. AI masks them.

2. Tasks Designed to Show What Students Know

Any assignment intended to serve as evidence of learning—practice quizzes, written explanations, problem-solving tasks—loses validity when AI generates responses.

From a Visible Learning lens, if teachers cannot trust the evidence, they cannot respond instructionally. AI has no place in these moments.

3. Work Intended to Build Fluency or Automaticity

Fluency develops through repetition and recall, not generation. AI shortcuts this process by supplying answers instead of strengthening memory.

This is particularly critical in foundational skills such as:

• math procedures,

• vocabulary acquisition,

• writing mechanics,

• or historical reasoning frameworks.

AI actively interferes with learning in these contexts.

Tucker’s Contribution: Accountability and Ownership

Catlin Tucker emphasizes that blended and independent learning only succeed when accountability structures are clear. Students must know:

• what they are responsible for,

• how their learning will be checked,

• and why the task matters.

AI must be positioned as a support tool, not a workaround. This requires:

• clear norms for use,

• transparency about expectations,

• and tasks designed around process, not just product.

Independent practice is where ownership should increase—not disappear.

Designing Independent Tasks That Resist AI Shortcuts

AI exposes weak task design quickly. Tasks that focus solely on polished products are easily outsourced. Tasks that emphasize thinking are not.

Strong independent practice tasks:

• require explanation of reasoning,

• apply learning to novel contexts,

• connect ideas across lessons,

• or ask students to reflect on process.

This is not about “AI-proofing” assignments. It is about learning-centered design.

The Teacher’s Role Remains Central

Independent practice does not reduce the teacher’s role—it changes it. Teachers analyze patterns, respond to evidence, and decide what comes next.

AI may help organize data or surface trends, but it cannot interpret learning in context. That responsibility remains human.

Why This Matters

Independent practice is where students learn to own their learning. When AI is aligned to research-based instruction, it can:

• enhance feedback,

• support reflection,

• and strengthen metacognition.

When misaligned, it quietly erodes learning.

Independence must be taught.
Evidence must be trusted.
Instruction must still lead.

References

Hattie, J. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to achievement. Routledge.

Hattie, J. (2012). Visible learning for teachers: Maximizing impact on learning. Routledge.

Rosenshine, B. (2012). Principles of instruction: Research-based strategies that all teachers should know. American Educator.

Sherrington, T. (2018). Exploring Barak Rosenshine’s seminal Principles of Instruction: Why it is THE must-read for all teachers. Teacherhead.

https://teacherhead.com/

Tucker, C. (2016). Blended learning in grades 4–12: Leveraging the power of technology to create student-centered classrooms. Corwin.

Transparency note: This article was developed with the support of AI tools for drafting and organization. All instructional perspectives and final decisions reflect my professional experience and judgment.

Brewing Innovation Shop

As artificial intelligence becomes more visible in classrooms, one concern continues to surface among educators and school leaders: Will AI replace direct instruction?

The answer, grounded firmly in research, is no. In fact, the rapid expansion of AI makes explicit, well-designed instruction more important than ever.

Decades of cognitive science research, synthesized by Barak Rosenshine and thoughtfully extended by Tom Sherrington’s work at Teacherhead, show that students learn best when instruction is clear, sequenced, and responsive. AI does not change these principles. It either supports themmor undermines them, depending on how it is used.

Share

Direct Instruction: What the Research Actually Says

Rosenshine’s Principles of Instruction synthesize findings from cognitive psychology, classroom observation, and studies of effective teaching (Rosenshine, 2012). His work emphasizes that novice learners benefit most when teachers:

• clarify learning goals,

• model new material explicitly,

• present content in small steps,

• check for understanding frequently,

• guide practice before independence,

• and revisit learning through review.

Tom Sherrington, through his writing on Teacherhead, has consistently reinforced these ideas by emphasizing learning as a cycle, not a one-off event. He frames effective instruction as a process of explanation, checking, practice, feedback, and refinement, what he often refers to as responsive teaching (Sherrington, 2018).

This framing matters because AI often enters classrooms at the wrong point in the cycle. Students need the foundational knowledge and skills in place before they begin to use AI to support learning.

The Real Risk: Skipping the Instructional Foundations

One of the most common missteps in AI integration is asking students to generate, explore, or create before they understand the underlying concepts.

Research is clear: minimally guided instruction is ineffective for novice learners because it overloads working memory and encourages shallow strategies (Kirschner, Sweller, & Clark, 2006). Sherrington echoes this concern, warning against premature independence that leads to “performance without learning.”

In AI-rich environments, this risk is amplified. When explicit instruction is weak or absent:

• students rely on AI-generated answers instead of understanding.

• misconceptions are masked rather than revealed.

• and surface-level completion is mistaken for mastery.

AI does not solve these problems. Strong instruction does. Students need sound instruction and foundational knowledge from teachers to build up on their learning toward mastery. AI can support that, but it cannot replace it.

Where AI Fits in Rosenshine’s Models

When aligned intentionally, AI can strengthen several core instructional practices highlighted by both Rosenshine and Sherrington, without replacing the teacher.

1. Explanation and Modeling

Rosenshine emphasized the importance of clear teacher explanations and worked examples. Sherrington extends this by stressing that explanations should be tested through questioning and application.

AI can support this phase by:

• generating additional worked examples.

• offering alternative representations of a concept.

• or helping teachers anticipate misconceptions.

However, AI should follow teacher explanation, not replace it. The teacher remains the primary model.

2. Checking for Understanding (Responsive Teaching)

One of both Rosenshine and Sherrington’s most consistent themes is the importance of checking for understanding, not as a formality, but as a driver of instructional decisions.

AI can assist teachers by:

• generating hinge questions.

• creating quick diagnostic prompts.

• or organizing student responses for faster analysis.

Used this way, AI supports responsiveness, not automation. The teacher still decides what to reteach, slow down, or extend. Again, this shows where AI can be used with intention.

3. Guided Practice: Where AI Can Strengthen Learning Without Replacing Thinking

Guided practice occupies a critical position in Rosenshine’s instructional framework. It is the phase where students begin applying new learning with support, allowing teachers to correct errors, reinforce success, and gradually release responsibility. Rosenshine emphasized that students should not move to independent practice until they demonstrate a high level of accuracy during guided practice, typically around an 80 percent success rate (Rosenshine, 2012).

Tom Sherrington reinforces this idea through his emphasis on responsive teaching. In his writing, guided practice is not simply “time to work,” but time for teachers to actively observe, question, and adjust instruction based on student responses. Practice is only effective when it is deliberate, monitored, and aligned to the learning goal.

This is precisely where AI can add value—if it is used intentionally.

What Guided Practice Is (and Is Not)

Before discussing AI, it is important to clarify what guided practice actually entails.

Guided practice is:

• closely aligned to the teacher’s explicit instruction,

• completed immediately after modeling,

• structured to prevent large errors,

• supported through prompts, cues, and feedback,

• and monitored continuously by the teacher.

Guided practice is not:

• independent work,

• open-ended exploration,

• or production of final products.

AI should only be introduced into guided practice when it reinforces these conditions—not when it shortcuts them.

How Teachers Can Use AI During Guided Practice (Rosenshine-Aligned)

1. Scaffolding Practice Tasks Without Giving Answers

One of the strongest uses of AI during guided practice is as a scaffolding tool, not an answer generator. Teachers can design AI interactions that:

• restate directions in simpler language,

• prompt students with guiding questions,

• or break multi-step tasks into manageable chunks.

For example, instead of asking AI to solve a problem, students might be prompted to ask:

• “What is the first step I should consider?”

• “What does this question want me to demonstrate?”

• “What information from the lesson applies here?”

This aligns directly with Rosenshine’s emphasis on presenting material in small steps and guiding students through initial application.

2. Immediate, Criteria-Based Feedback

Rosenshine identified feedback as most effective when it is immediate and specific. During guided practice, AI can help provide rapid feedback tied to teacher-defined criteria. For example:

• Teachers can provide AI with a rubric or checklist aligned to the lesson objective.

• Students can submit a response and receive feedback focused on accuracy, clarity, or structure—not completion.

Importantly, this feedback should highlight what is correct and what needs adjustment, not rewrite student work. Sherrington cautions against feedback that does the thinking for students; AI feedback must prompt revision, not replacement.

3. Supporting Error Analysis and Misconception Detection

Another powerful use of AI during guided practice is helping students analyze errors. Rosenshine emphasized the importance of identifying and correcting mistakes early, before they become ingrained. AI can be used to:

• explain why an answer is incorrect.

• compare a student response to a model example.

• or highlight where reasoning diverged from the expected process.

This mirrors Sherrington’s focus on responsive adjustment: noticing misconceptions quickly and responding before students practice them repeatedly.

How Students Should Use AI During Guided Practice

Clear norms are essential. Without them, guided practice quickly turns into AI-assisted completion. Students should be taught explicitly that during guided practice:

• AI may help clarify understanding, not generate final answers.

• AI may support revision, not replace effort.

• AI should be used to ask better questions, not avoid them.

A helpful framing for students is:

“Use AI to improve your thinking, not to replace it.”

Teachers should model this explicitly by showing:

• appropriate prompts.

• inappropriate shortcuts.

• and examples of productive AI-supported practice.

The Teacher’s Role Remains Central

Even with AI support, guided practice is not automated. Rosenshine and Sherrington both make clear that teacher monitoring is essential during this phase. Teachers must still:

• circulate.

• observe student responses.

• ask probing questions.

• and decide when to reteach, slow down, or move forward.

AI can make guided practice more efficient, but it cannot replace instructional judgment. In fact, guided practice becomes more effective when AI frees teachers to focus on small-group instruction and targeted intervention.

Protecting the Transition to Independent Practice

One of Rosenshine’s strongest cautions is against moving students to independent practice too quickly. AI can obscure this boundary if teachers are not intentional. Before releasing students to independent work, teachers should ensure that:

• students demonstrate consistent accuracy.

• misconceptions are addressed.

• and AI support is gradually reduced.

Sherrington often emphasizes that independence should be earned, not assumed. AI should follow the same rule.

Why This Matters

When guided practice is designed well, it is the bridge between instruction and independence. When AI is aligned to Rosenshine’s principles, it can strengthen that bridge by:

• increasing feedback frequency,

• supporting scaffolding,

• and reducing unnecessary struggle.

When AI is misused, it collapses the bridge entirely. Guided practice remains a teacher-led, student-centered process. AI’s role is to support learning, not to replace it.

Protecting Cognitive Demand and Academic Integrity

Sherrington frequently writes about the danger of “task completion masquerading as learning.” AI makes this danger more real. Without clear boundaries, students may:

• submit AI-generated work they do not understand.

• confuse fluency with mastery.

• or avoid the struggle necessary for durable learning.

This is where instructional norms matter:

• AI may support thinking, not replace it.

• AI may clarify, not complete.

• AI use must align explicitly with the learning goal.

These boundaries protect learning, integrity, and the teacher’s role as the instructional leader. Teachers must acknowledge that we now live in an AI driven world. Even if we do not agree with the promise of AI, we need to understand that students will try to use it, whether it is an approved tool or not. We need to understand that our practice needs to adapt and make assessments either AI-proof or AI-supported, with the later being the more effective method.

Direct Instruction in a Blended, AI-Supported Classroom

When paired with blended learning, particularly station rotation, direct instruction becomes even more effective. Teachers can:

• deliver explicit instruction to small groups.

• check understanding in real time.

• and direct students to AI-supported practice that reinforces specific skills.

This aligns perfectly with Sherrington’s emphasis on responsive adjustment and Rosenshine’s focus on guided practice. In this model, AI lives primarily in the “We do” and early “You do” phases, never replacing the instructional core.

Why This Matters for Teacher Efficacy

Teacher efficacy improves when educators feel in control of instruction, pacing, and learning outcomes. Poorly implemented AI erodes this. Well-aligned AI strengthens it. When AI supports Rosenshine- and Sherrington-aligned practices:

• planning becomes more efficient.

• differentiation becomes more manageable.

• and feedback becomes more timely.

Teachers remain the architects of learning. AI becomes an assistant—not a substitute.

Looking Ahead

The next post in this series will focus on guided practice, independent practice, and feedback, examining how AI can support practice cycles without shortcuts or dependency.

Direct instruction is not outdated. It is the foundation upon which effective AI integration must be built.

References

Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 41(2), 75–86.

Rosenshine, B. (2012). Principles of instruction: Research-based strategies that all teachers should know. American Educator.

Sherrington, T. (2018). Exploring Barak Rosenshine’s seminal Principles of Instruction: Why it is THE must-read for all teachers. Teacherhead. https://teacherhead.com/

Transparency note: This article was developed with the support of AI tools for drafting and organization. All instructional perspectives, examples, and final decisions reflect my professional experience and judgment.

Stay Curious. Stay Caffeinated. ☕

Also, be sure to visit the NEW Brewing Innovation shop for coffee, mugs, and apparel!

Brewing Innovation Shop

Once schools establish that instruction must lead technology, and that safety must be intentionally designed, the next challenge becomes practical implementation. Even educators who believe in the potential of AI often ask the same question:

How do we integrate these tools with intention and without overwhelming teachers, diluting instruction, or losing control of the classroom?

The answer is not another tool or platform. It is structure.

Blended learning, when implemented intentionally, provides the organizational framework necessary to integrate technology. I first worked with it when I was in the classroom reading the work by Dr. Catlin Tucker, who has sound research-based strategies for over a decade. As we now educate in a world that becomes ever tech-driven, her work has become instrumental in how we can integrate technology and AI in ways that strengthen instruction rather than compete with it.

Share

Why Structure Is the Missing Piece in Many AI Initiatives

Many technology initiatives falter not because the tools are ineffective, but because they are introduced without a clear instructional role. Teachers are asked to manage whole-class instruction, individualized learning needs, behavior, assessment, and technology use simultaneously. The result is often increased cognitive load for teachers and fragmented learning experiences for students.

Blended learning addresses this challenge by deliberately organizing instructional time, space, and teacher attention. Rather than expecting technology to “enhance learning” in abstract ways, blended models define when and how tools are used in service of specific learning goals.

As Catlin Tucker emphasizes, blended learning is not about putting students on devices; it is about designing learning experiences that balance teacher-led instruction, collaboration, and independent practice (Tucker, 2016).

Station Rotation as a Research-Aligned Model

Among blended learning models, station rotation is particularly well suited to classroom-based instruction. In this model, students rotate through multiple stations during a class period, typically including:

• a teacher-led instructional station

• a collaborative station

• and an independent or online station

Tucker’s research and classroom-based work highlight that station rotation succeeds when each station is intentionally designed with a clear purpose and aligned to learning objectives (Tucker, 2016). This intentionality is what makes the model powerful—and what makes it compatible with research-based instruction.

Alignment With Cognitive Science and Effective Instruction

When designed thoughtfully, station rotation aligns closely with established instructional research.

At the teacher-led station, educators can apply principles consistent with Rosenshine’s work by:

• explicitly modeling new content

• presenting material in manageable steps

• guiding practice

• and checking for understanding

This station protects the role of direct instruction and ensures that foundational knowledge is developed before students move into more independent or exploratory work.

The collaborative station supports deeper processing through discussion, application, and problem-solving. This aligns with research on elaboration and retrieval practice, as students articulate their thinking, respond to peers, and apply concepts in new contexts.

The independent or online station is where technology, and AI in particular, can be integrated most effectively. When bounded by clear expectations and instructional goals, AI can support:

• scaffolded practice

• clarification of misunderstandings

• organization of ideas

• and formative feedback

Crucially, AI is not positioned as a replacement for instruction or thinking. Instead, it functions as a support tool within a controlled instructional environment.

Why Blended Learning Makes AI Integration Safer

From a safety perspective, blended learning offers built-in guardrails that unstructured technology use does not.

Because students rotate between learning modes:

• AI use is time-limited rather than continuous

• teacher proximity and supervision are increased

• and expectations for appropriate use are repeatedly reinforced

This structure directly supports the Classroom AI Safety Loop introduced in the previous post. Monitoring, redirection, and reflection become part of daily instruction rather than reactive responses to misuse.

In other words, blended learning transforms AI from an open-ended risk into a managed instructional resource.

Supporting Teacher Efficacy Through Instructional Design

Blended learning also has a significant impact on teacher efficacy. Research on teacher efficacy consistently shows that educators feel more effective when they can:

• manage classroom routines

• respond to individual student needs

• and see evidence of learning

Station rotation increases teacher efficacy by allowing educators to:

• work with small groups rather than managing the entire class at once

• observe student thinking more closely

• and make timely instructional adjustments

When AI is embedded within this structure, it can further support efficacy by reducing repetitive workload such as generating practice variations or organizing feedback, while preserving professional judgment and instructional authority.

Blended Learning Is Not a Shortcut

It is important to be clear: blended learning is not a shortcut to innovation. It requires planning, intentional design, and ongoing reflection. Poorly implemented blended learning can be just as ineffective as poorly implemented technology.

Blended learning, when designed with purpose, provides the necessary conditions for the responsible, ethical, and instructional use of AI. When I taught in the classroom using the co-teaching model, this was a valuable method we used to address learning needs and leverage the teaching expertise of both teachers. My background in social studies allowed me to facilitate highly effective instruction when co-teaching in a social studies classroom.

Overall, blended learning allows schools to innovate without abandoning proven teaching practices.

Looking Ahead

The next post in this series will narrow the focus further, examining explicit instruction, modeling, and guided practice, and how AI can support these high-impact strategies without reducing cognitive demand or undermining academic integrity.

Innovation becomes sustainable when it is structured.
Blended learning provides that structure.

References

Tucker, C. (2016). Blended learning in grades 4–12: Leveraging the power of technology to create student-centered classrooms. Corwin.

Transparency note: This article was developed with the support of AI tools for drafting and organization. All instructional perspectives, examples, and final decisions reflect my professional experience and professional judgment.

Stay Curious. Stay Caffeinated. ☕

Also, be sure to visit the NEW Brewing Innovation shop for coffee, mugs, and apparel!

Brewing Innovation Shop

As artificial intelligence becomes more visible in classrooms, the conversation around safety often swings between two extremes. On one end is fear, calls to ban AI outright. On the other is unstructured enthusiasm, tools to proliferate without clear guardrails. Neither approach serves students or educators well.

Responsible AI integration requires a third path: intentional, continuous, and instructional safety practices. In other words, safety is not a one-time checklist or a static policy. It is a loop.

Why AI Safety Must Be Ongoing

Recent research and reporting underscore why safety must be central to any AI strategy. Surveys show that AI use among teachers and students has increased rapidly, often outpacing training, policy development, and instructional guidance. While many educators report benefits such as time savings and personalization, concerns persist around data privacy, academic integrity, misinformation, bias, and the erosion of critical thinking skills.

Notably, students frequently encounter AI without sufficient guidance. Many report limited understanding of how AI works, how to evaluate its outputs, or how school expectations apply to its use. This gap places an unfair burden on teachers to manage risk reactively rather than proactively.

The need for AI Literacy is as critical as ever in the last two years. This alone is a skill-set, along with Digital Fluency. The reality is that students and staff are going to use AI, it is essential that EdTech leaders provide guidance on how to use it in an appropriate and efficient way. The more we know, the better we can address concerns and clear the path for the 21st century where an AI-driven world is becoming more of a reality than ever before. To put it in layman’s terms: Know better, do better.

The takeaway: AI safety cannot be addressed through prohibition or optimism alone. It must be designed into classroom and everyday practice.

Subscribe now

The Classroom AI Safety Loop

A useful way to conceptualize responsible AI use is through a safety loop rather than a rulebook. This loop consists of four interconnected stages that repeat over time as tools, models, and usage evolve.

1. Establish Clear Expectations

The first step is clarity. Students and teachers need explicit guidance on what AI can and cannot be used for, grounded in academic integrity and learning goals. This includes distinguishing between using AI to support thinking—such as brainstorming or organizing ideas—and using AI to replace thinking, such as submitting generated work as one’s own.

Clear expectations also require transparent communication about privacy. Students should understand what types of information should never be shared, how data may be stored or used, and why school-approved tools matter.

2. Teach AI Literacy Explicitly

Safety improves when students understand the limitations of AI. Instruction must address issues such as hallucinations, bias, and the difference between pattern generation and human reasoning. When students are taught to critically evaluate AI outputs, verify information, and question accuracy, AI becomes a learning tool rather than an authority.

This instruction cannot be assumed. It must be modeled, practiced, and revisited across grade levels and subject areas.

3. Monitor and Respond Thoughtfully

Even with clear expectations, misuse and misunderstandings will occur. Effective AI safety systems include monitoring and adult oversight—not as surveillance, but as instructional support. When issues arise, responses should focus on redirection, reflection, and learning rather than punishment alone.

This approach mirrors existing best practices in digital citizenship and behavior management: intervene early, clarify expectations, and guide students toward better choices.

4. Reflect and Refine

The final—and most often overlooked—stage is reflection. AI tools change rapidly, as do student behaviors and needs. Schools must regularly revisit their practices, evaluate what is working, and adjust accordingly. Feedback from teachers and students is essential in this process.

Safety, in this sense, is never “finished.” It is strengthened through continuous refinement.

Share

Aligning Safety With Instruction

Importantly, the AI safety loop is not separate from instruction—it supports it. When safety practices are embedded into daily routines, they protect cognitive demand, reinforce academic integrity, and preserve the human relationships at the center of learning.

This aligns closely with broader principles of effective teaching: clarity, modeling, guided practice, feedback, and review. AI safety works best when it is taught the same way we teach any complex skill. It is a skill-set that is applicable across all disciplines and is going to be as important as ever as we continue to grow in today’s society. The more we discuss reports about the job outlook by 2030 by the World Economic Forum, the more we understand why this skill-set is critical for today’s students.

What This Means for Schools and Classrooms

For schools, this means moving beyond static policies toward living frameworks that evolve with practice. For teachers, it means having clear guidance and support rather than being left to navigate AI challenges alone. For students, it means learning how to use powerful tools responsibly, ethically, and productively. Administration and leaders in districts need to take the type to explicitly explain data privacy policies and procedures, plus why they are important.

Safety is not the barrier to innovation. It is the condition that makes innovation sustainable and attainable.

Subscribe now

Coming Next

Post #3: Blended Learning That Works — Using Structure to Make AI Manageable

We’ll explore how blended learning models, particularly station rotation, provide the structure needed to integrate AI intentionally without overwhelming teachers or students.

References:

• Google for Education. (2023). Teaching responsible use of AI: Lessons and activities for students. https://edu.google.com/intl/ALL_us/ai/education/

• Google. (2023). AI principles. https://ai.google/principles/#our-ai-principles-in-action

• MagicSchool. (2024). The AI safety loop for students. https://www.magicschool.ai

• Vilcarino, J., & Langreo, L. (2025, October 8). Rising use of AI in schools comes with big downsides for students. Education Week. https://www.edweek.org/technology/rising-use-of-ai-in-schools-comes-with-big-downsides-for-students/2025/10

Transparency note: This article was written with the support of AI tools for drafting and organization. All instructional perspectives, examples, and final decisions reflect my professional experience and judgment.

Stay Curious. Stay Caffeinated. ☕

Across schools and districts, the conversation around educational technology and artificial intelligence has accelerated rapidly. New tools are introduced, pilot programs are launched, and professional development sessions are scheduled with the promise of innovation and efficiency. Yet, despite this momentum, one critical question continues to surface: Is this improving instruction and learning?

That question must remain central. Decades of educational research consistently demonstrate that student learning improves when instruction is intentional, structured, and responsive. Technology, regardless of its sophistication, does not improve learning independently. Instead, its effectiveness depends on how well it aligns with sound instructional practices (Hattie, 2009).

This series begins with a simple premise: instruction must lead, and technology must follow.

Share

Why This Series Matters

Having worked as a classroom teacher, special education teacher, and instructional leader, I have seen educational technology used in markedly different ways. When thoughtfully implemented, technology expands access, supports differentiation, and allows teachers to focus more deeply on feedback and relationships. When poorly aligned, it introduces complexity without improving outcomes.

The difference between these outcomes is not the tool itself, but the instructional framework guiding its use. This series examines how artificial intelligence and educational technology can be integrated responsibly by grounding them in research-backed instructional practices rather than trends or novelty.

Instructional Research as the Foundation

The integration strategies explored throughout this series are anchored in the work of several influential educational researchers whose findings continue to shape effective teaching.

Barak Rosenshine’s Principles of Instruction synthesize decades of cognitive science research and classroom observation. His work emphasizes the importance of clear learning goals, presenting new material in small steps, modeling through worked examples, frequent checks for understanding, guided practice, and ongoing review (Rosenshine, 2012). These principles are especially relevant in technology-rich classrooms, where cognitive overload can easily occur. When AI tools are aligned to these instructional moves such as generating exemplars or supporting scaffolded practice, they can strengthen instruction rather than distract from it, think about NotebookLM in this situation. Rosenshine’s principles, widely regarded as a foundational synthesis of cognitive science and effective classroom practice, continue to be cited as essential reading for educators seeking to improve instructional clarity, student success rates, and retention (Sherrington, 2018).

John Hattie’s Visible Learning research further reinforces the need to evaluate instructional decisions based on their impact on student learning rather than intuition or popularity. By analyzing over 1,400 meta-analyses, Hattie identifies the practices most likely to influence achievement, emphasizing clarity of learning intentions, success criteria, and feedback as high-impact influences (Hattie, 2009; Hattie, 2012). Within this framework, the value of AI is determined not by its novelty, but by its measurable effect on learning.

Robert Marzano’s research on high-yield instructional strategies complements this work by identifying classroom practices that consistently support student achievement. Strategies such as identifying similarities and differences, summarizing and note-taking, reinforcing effort, and providing feedback remain effective regardless of the presence of technology (Marzano, Pickering, & Pollock, 2001). Artificial intelligence can enhance these strategies by improving efficiency and consistency, but it does not replace the instructional judgment required to implement them well.

Catlin Tucker’s work in blended learning provides a practical structure for integrating technology while maintaining strong instructional control. Her station rotation model intentionally balances teacher-led instruction, collaborative learning, and independent practice, ensuring that technology serves a defined purpose within the learning process rather than dominating it (Tucker, 2016). AI tools fit naturally into these models when their role is clearly aligned to instructional goals.

The Non-Negotiable Role of Safety

Any discussion of AI integration must have safety as a focal point. Responsible implementation requires explicit attention to data privacy, age-appropriate use, bias, accuracy, and academic integrity. Research and emerging guidance on educational AI emphasize the importance of ongoing evaluation, clear expectations, and instructional oversight rather than one-time policy statements.

Throughout this series, safety will be treated as a foundational requirement rather than a compliance exercise. Effective AI integration depends not only on what tools are used, but on how thoughtfully boundaries and norms are established for both students and educators.

What This Series Will Provide

This series is designed to move beyond abstract theory and tool promotion. Each post will connect research-based instructional practices to realistic applications of AI and educational technology. The focus will remain on improving clarity, increasing instructional efficiency, supporting teacher decision-making, and strengthening student learning experiences.

By grounding technology use in established research, the series aims to provide educators and leaders with practical guidance they can apply immediately while maintaining a clear focus on instructional quality and student impact.

Looking Ahead

The next post in this series will focus on safety as the starting point for AI integration. It will explore how schools can move beyond fear-based responses or unstructured adoption by establishing clear norms, safeguards, and instructional expectations.

Instruction remains the foundation of learning. When technology is aligned to that foundation, it has the potential to meaningfully enhance teaching and learning. When it is not, it risks becoming another initiative that promises change without delivering impact.

References

Hattie, J. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to achievement. Routledge.

Hattie, J. (2012). Visible learning for teachers: Maximizing impact on learning. Routledge.

Marzano, R. J., Pickering, D. J., & Pollock, J. E. (2001). Classroom instruction that works: Research-based strategies for increasing student achievement. ASCD.

Rosenshine, B. (2012). Principles of instruction: Research-based strategies that all teachers should know. American Educator.

Sherrington, T. (2018). Exploring Barak Rosenshine’s seminal Principles of Instruction: Why it is the must-read for all teachers. Teacherhead. https://teacherhead.com/2018/06/10/exploring-barak-rosenshines-seminal-principles-of-instruction-why-it-is-the-must-read-for-all-teachers/

Tucker, C. (2016). Blended learning in grades 4–12: Leveraging the power of technology to create student-centered classrooms. Corwin.

Transparency note: This article was written with the support of AI tools for drafting and organization. All instructional perspectives, examples, and final decisions reflect my professional experience and judgment.

Stay Curious. Stay Caffeinated. ☕

Lately, my feeds have been full of the same message:
“Ban laptops. Ban screens. Ban devices.”
Some posts go so far as to suggest that giving students technology “took away their brains.” Sound familiar? Matt Miller captured this wave perfectly in his newsletter this morning.

And look — I get it. I’m an assistant principal, a K–12 instructional technology leader, and a parent. I’ve seen the doom-scrolling. I’ve seen the fractured attention spans. I’ve seen the effects of unhealthy digital habits.

But here’s the thing the discourse keeps missing:

If we put a device in front of our students with no intention or systematic solution, it is no different than the parent that gives their child a tablet at the dinner table in a restaurant to keep them quiet. There’s a massive difference between giving kids devices… and giving kids purposeful, well-designed, teacher-led learning experiences that leverage technology to deepen thinking, expand access, and personalize support.

Let’s talk about that difference and what we lose when the louder voices push us toward all-or-nothing solutions.

Share

Yes, There Are Real Concerns, But That’s Not the Whole Story

Matt lays out a mountain of evidence: mental health trends, sleep disruption, attention issues, and the rise of addictive platforms engineered to keep kids hooked.

None of that should be ignored.

But when critics take that research and jump straight to:
“Ban devices in classrooms,” they’re making the same mistake we made during remote learning and assuming the tech caused the problem rather than recognizing the absence of intentional design.

Unstructured screen time ≠ instructional technology.
YouTube doom loops ≠ guided learning.
TikTok ≠ blended instruction.

And anyone who has actually taught in a modern classroom knows the difference immediately.

Access Is Equity — Especially for Students With Disabilities

Before I was an administrator, I was a special education teacher in 15:1 classrooms.
Technology wasn’t “nice to have.” It was access.

• Students who struggled with writing could speak their ideas through voice typing.

• Students with processing disabilities could replay instruction, not just rely on one-time delivery.

• Students who needed scaffolded text could get it instantly, without embarrassment or stigma.

• Students with limited working memory could use organizational tools that kept them afloat.

I didn’t use tech to replace instruction. I used it to remove barriers. When people say “ditch the devices,” I hear:

“Let’s take away the tools that help students who need them most.”

That’s not equity. That’s nostalgia masquerading as pedagogy.

Catlin Tucker Showed Us the Blueprint for Doing This Well

If you’ve ever implemented Catlin Tucker’s blended learning strategies, especially station rotation, you know exactly how powerful instructional technology can be when framed through sound pedagogy.

Blended learning didn’t make my classroom more digital; it made it more human, because technology handled the tasks that didn’t need me:

• Differentiated practice at one station

• Teacher-led small groups at another

• Collaborative work that required students to think, create, and reflect

Tech wasn’t the star of the show. It was the infrastructure that made personalization possible. Catlin’s work has become foundational for educators across the country for a simple reason:

Blended learning gives students more voice, more choice, more agency — and gives teachers more time where it matters most.

We don’t need less of that. We need more.

Binary Thinking Helps No One

Matt calls this out directly: the loudest arguments fall into all-or-nothing thinking — “Ban it all” or “Tech is the enemy.”

But teaching has never been all-or-nothing. It’s always been artful decision-making.

Just look at my own practice (and his):

• Some things belong on paper.

• Some things belong on Chromebooks.

• Some things belong in conversation.

• Some things belong in AI-supported tools.

Good teaching is, and has always been, the act of choosing the right tool at the right time for the right purpose. Devices don’t change that. AI doesn’t change that. Pandemic-era trauma doesn’t change that. In fact, the nuance matters now more than ever.

What the Viral Arguments Get Wrong — Let’s Look at the Actual Examples

Matt Miller highlighted several of the exact arguments fueling this “ban the tech” argument and when you look closely, they all share the same flaw: they oversimplify complex problems into a catchy headline.

Here are the big three he calls out, and the nuance they ignore:

1. “Ban laptops — handwriting leads to deeper learning.”

Matt cites Adam Grant’s viral post arguing that laptops should be removed from classrooms because students supposedly “learn more and get better grades” when taking notes by hand.

But this argument makes a huge leap: Better retention with handwritten notes does NOT mean laptops have no place in learning.

Notetaking is just one small slice of classroom practice. Laptops also support:

• accessibility tools

• project-based learning

• collaborative writing

• guided practice with instant feedback

• AI-supported revision and coaching

The conclusion shouldn’t be “ban laptops.” It should be “teach students when to hand-write and when to type.” That’s called instruction, not elimination.

2. “We gave students laptops and took away their brains.”

Matt includes a screenshot from a TikTok video claiming that screens have destroyed cognitive stamina, citing the book Digital Delusion as evidence. Digital Delusion does raise important concerns about passive screen consumption and we should take those seriously. But here’s the nuance missing from the viral take:

Digital Delusion critiques unstructured, unsupervised, entertainment-driven screen use — not purposeful instructional technology.

There is a world of difference between:

• an hour of TikTok

• and an hour of blended learning where students rotate between small-group instruction, targeted practice, and collaborative problem-solving.

Screens aren’t the villain. Unstructured consumption is. And it’s our responsibility to teach students the difference.

3. “AI will destroy critical thinking.”

This one is everywhere — and Matt hits the nail on the head. The viral argument assumes students will only ever use AI to shortcut thinking. But that assumes the worst, not the possible.

In reality, when used intentionally, AI can:

• scaffold struggling readers

• provide alternative explanations

• help students rehearse ideas before writing

• act as a study partner

• free teachers to spend more time in small groups

• differentiate instruction in ways impossible before

AI isn’t replacing thinking. It’s replacing busywork so students can think.

The Pattern Is Clear: The Issue Isn’t Tech, It’s Oversimplification

Each of these arguments:

1. Starts with a real concern

2. Applies all-or-nothing thinking

3. Ends with a blunt solution that removes tools instead of improving teaching

As Matt writes, we don’t need sledgehammers — we need scalpels. We need intentional, thoughtful design, not bans. Even Digital Delusion, when read in full, doesn’t argue for eliminating instructional technology. It argues for better habits, clearer boundaries, and healthier relationships with screens, exactly what schools should be teaching in the first place.

What We Should Be Fighting For

This isn’t a moment to ban devices.
It’s a moment to invest in teacher skill, instructional design, and healthy tech habits.

Here’s what our schools actually need:

1. Pedagogy-First Tech Integration

Tech should never be the driver. Pedagogy leads; tools follow.

2. Clear Routines and Guardrails

Students need structure around screen use, just like they need structure around everything else in school.

3. Media Literacy and AI Literacy for All

Because students don’t need less technology, they need smarter, healthier, more empowered relationships with technology.

4. Professional development that treats teachers as designers, not device managers

We should be helping teachers make intentional choices, not mandating tools or banning them.

5. A culture where tech is a scalpel, not a sledgehammer

Matt’s metaphor is spot on: we need precision, not blunt-force rules.

Let’s Stop Pretending That Banning Devices Solves the Problem

Removing laptops doesn’t magically:

• improve mental health,

• fix instructional gaps,

• rebuild attention spans,

• or prepare students for the world they’re walking into.

But empowering teachers to design learning that is:

• active

• reflective

• collaborative

• blended

• human-centered

…that does make a difference.

We don’t need a tech ban.
We need better teaching supported by thoughtful technology use.

That’s the work. And it’s work worth doing.

Stay Curious. Stay Caffeinated. ☕

Share

NYSCATE 2025 reinforced something very clear to me: AI isn’t a wave we’re preparing to ride someday. It’s a current already reshaping the work we do today, right now.

Students are using AI before the first bell. Teachers are experimenting before PD catches up. Leaders are trying to build the plane while flying it. And as uncertainty grows, the demand for meaningful, human-centered innovation grows with it.

This moment doesn’t require tech champions.
It requires leaders who can create clarity in a fog of change.

That’s why I’m sharing this: a detailed, actionable AI Leadership Playbook for 2025—designed for school administrators, EdTech leaders, teacher-leaders, and anyone tasked with shaping the future of learning.

It’s not about tools. It’s about systems, culture, clarity, safeguards, and humanity. Let’s get brewing.

Subscribe now

1. Begin With Mission, Not Machines

AI should amplify your district’s purpose, not distract from it.

Here’s the biggest implementation mistake I see districts make:
They begin with, “Which AI tools should we use?”

The districts that thrive start somewhere else entirely.

Try this exercise with your leadership team:

1. Pull up your district mission statement.

2. Highlight one phrase that defines what you believe about learning.

3. Ask: How could AI help us do this better, faster, or more equitably?

This simple move reframes everything.

Suddenly AI is not about: automation, shortcuts, novelty

It becomes about: belonging, equity, student voice, access to rigorous learning, personalization that honors human strengths, reducing teacher burnout, empowering deeper thinking

Example:

If your mission emphasizes “empowering independent learners,” then AI becomes a tool for: research coaching, immediate feedback, curiosity-driven inquiry, metacognitive reflection.

If your mission emphasizes “equitable access,” then AI becomes a tool for: differentiating reading levels, scaffolding writing, creating multimodal representations, providing translation and language support.

Mission → Strategy → Tools
Never the reverse.

2. Build Teacher LLM Literacy Before You Build AI Initiatives

If your staff only knows how to click shortcuts, you don’t have AI integration—you have AI confusion.

At her NYSCATE session, Holly Clark put it bluntly:

“Shortcut tools are terrific training wheels, but teachers need to actually use LLMs.”

This is the truth most districts miss. PD often focuses on apps, not thinking. And tools are changing too fast for tool-based PD to keep up.

Teachers need a working mental model of how AI thinks.

Not just:

• Which button to press

• Which menu to navigate

• Which prompt to paste

But:

• Why the model responded this way

• What blind spots it has

• What it’s confident about—and what it’s not

• How to cross-check accuracy

• When to trust it (and when not to)

Introduce the “Eager Intern” model.

This analogy transforms teacher anxiety into curiosity.

AI is: eager, fast, well-read, helpful.

But also: inexperienced, overconfident, prone to mistakes, lacking context, not a teacher.

It needs guidance, direction, correction, and feedback.

Your PD focus in 2025 should shift from:

❌ “Here’s a list of tools” ➡️ to ✅ “Here’s how to think with AI—critically, creatively, responsibly.”

When teachers understand AI as a thought partner, not an answer engine, everything changes.

3. Craft an AI Vision That Is Short, Clear, and Human

If your AI vision takes more than 20 seconds to explain, no one will remember it.

Every district needs a concise AI vision that answers three questions:

1. Why are we using AI?

Example: “To expand access, deepen inquiry, and save teachers time.”

2. How will AI help students and teachers?

Example: “AI will support personalized learning, accelerate planning, and provide instant feedback.”

3. What are the guardrails?

Example: “AI use will always center human judgment, transparency, and student dignity.”

That’s it.

Your AI vision should:

• be timeless

• avoid trendy buzzwords

• anchor in student experience

• focus on learning over tools

• be simple enough for teachers to reference daily

Too many districts hide behind jargon. A real vision acts like a compass. It gives direction—even when the landscape shifts.

4. Redesign Learning So Students Don’t Want to Cheat

AI hasn’t created cheating—it’s revealed which tasks were weak to begin with.

One of the most important shifts from NYSCATE was this idea:

“Stop asking if students will cheat. Start asking what needs to change so they don’t want to.”

Students cheat when : tasks lack meaning, teachers gate keep curiosity, assignments feel like compliance, feedback is delayed or absent.

AI invites a better question:

What would learning look like if curiosity—not compliance—drove the work?

When students can explore: alternative interpretations, role-play with characters, test hypotheses, ask “what if?,” debate AI’s mistakes.

…they stay engaged. And integrity becomes a byproduct of meaningful learning.

Where Blended Learning Meets AI: Catlin Tucker’s Station Rotation Model

This is also where Catlin Tucker’s blended learning framework becomes incredibly powerful—especially the Station Rotation model, which remains one of the most flexible, student-centered structures in modern instruction. Station Rotation naturally integrates AI without losing the human heart of teaching.

Imagine a class where students rotate through:

• A teacher-led station for direct instruction, questioning, and coaching

• A collaborative station for discussion, peer feedback, or problem-solving

• An AI-enhanced station where students use tools like ChatGPT or Gemini to explore ideas, analyze texts, get instant feedback, or test scenarios

Instead of hiding AI use, students learn to leverage it—ethically, transparently, and reflectively. It turns AI from a shortcut into a learning accelerator. It also gives teachers structured windows for high-quality feedback, all while developing student agency and independent thinking.

I’ll share a full sample AI-enhanced Station Rotation lesson template—including prompts, setup, and reflection routines—in an upcoming post.

Adding “Truth Statements” Builds Integrity

A simple move: ask students to add a Truth Statement at the bottom of every assignment:

“Here is how I used AI to support this work…”

This creates: transparency, accountability, and metacognitive awareness.

Cheating becomes irrelevant because learning becomes meaningful.

5. Build Guardrails That Empower, Not Restrict

Fear-based policies create fear-based cultures.

The goal of AI policy is not to restrict AI.
It is to guide safe, ethical, aligned use across your district.

Every district needs: a teacher AI policy, a student AI policy, a data privacy checklist, clarity around high-stakes tasks (grading, IEPs, evaluations), clear consequences for misuse—but also safe pathways for experimentation.

The best guardrails are:

• specific

• supportive

• role-based

• scenario-driven

• written in plain language

Example policies:

• “Teachers may use AI for first drafts, but final versions must reflect professional judgment.”

• “Students may use AI for brainstorming but must document their use in a truth statement.”

• “AI cannot replace teacher evaluation or grading of high-stakes assessments.”

Policies should reduce fear, not cause paralysis.

6. Make AI a Leadership Tool Before You Make It a Classroom Tool

If leaders aren’t modeling the work, implementation will fail. Here is the uncomfortable truth most districts avoid: If leaders aren’t using AI daily, teachers won’t either.

AI should be part of leadership workflows:

• drafting faculty emails

• designing agendas

• summarizing meeting notes

• creating walkthrough rubrics

• developing family communication templates

• generating PD outlines

• preparing board presentations

• analyzing survey data

• drafting vision statements and strategic plans

This does two things:

1. It normalizes AI as a thinking partner.

2. It sets the tone: “We’re learning this together.”

The worst leadership message is:
“I want everyone using AI… except me.”

The best leadership message is: “Here’s how AI helped me think through this decision. Try it.”

7. Adopt Multi-Model Reasoning: The New Essential Leadership Skill

This is the most powerful workflow I brought home from NYSCATE. This year, I started treating Gemini and ChatGPT as consultants who critique each other. It’s game-changing.

Here’s how the workflow works:

1. Ask Gemini to create a draft.

2. Ask ChatGPT to critique Gemini (strengths, weaknesses, blind spots).

3. Ask ChatGPT to produce its own draft.

4. Ask Gemini to critique ChatGPT’s version.

5. Merge the best thinking from both.

Why this is transformative:

• It reduces hallucinations.

• It exposes different perspectives.

• It generates more robust thinking.

• It mitigates bias.

• It accelerates clarity.

• It creates higher-quality work with less effort.

This is where AI leadership is heading: not single-model prompting, but multi-model reasoning. In 3–5 years, this will be a baseline skill.

8. Build a Culture Where Innovation Feels Safe, Not Scary

AI transformation is 20% strategy and 80% psychology. Innovation dies in cultures where teachers fear making mistakes.

Leaders must create cultures where experimentation is expected, celebrated, and shared.

That means saying:

• “Try something small this week.”

• “Tell us what worked and what didn’t.”

• “Share your prototype with the team.”

• “Your first attempt won’t be perfect—and that’s the point.”

Leadership ≠ hierarchy.

Leadership = humanity. One of the most powerful messages at NYSCATE was this:

“People don’t wake up wanting to do a bad job.”

Teachers want to grow. Students want to explore. Leaders want to support. Your job is to remove friction: simplify, clarify, encourage, model, protect, celebrate.

This is the work. This is the leadership AI demands.

The Bottom Line: AI Doesn’t Replace Leaders—It Reveals Them

2025 is the year districts stop “trying AI” and start building systems around it.

The leaders who thrive will be the ones who:

• anchor their work in mission

• prioritize teacher literacy

• design human-centered policies

• encourage curiosity-driven learning

• model real workflows

• use multi-model reasoning

• build cultures of experimentation

This is how we build future-ready schools. Not through apps—but through leadership. Meaningful, human, bold leadership. And we build it one intentional step at a time.

Stay Curious. Stay Caffinated. ☕

Subscribe now

AI Transparency Statement

Parts of this post were developed using AI tools, including ChatGPT, Gemini, and NotebookLM, as thought partners in the drafting and revision process. All ideas, interpretations, and final decisions reflect my own professional judgment and experience. AI supported the writing; it did not replace the human thinking behind it.

This week’s post runs longer than my usual 5–7 minute reads, but that feels appropriate. NYSCATE wasn’t a “quick insight” kind of conference. It was the kind of gathering that forces you to slow down, rethink assumptions, absorb ideas, and reimagine what leadership needs to look like in 2025. I left with pages of reflections, several new frameworks, a renewed sense of urgency—and honestly, a sense of excitement that’s hard to describe.

Because here’s the truth: AI isn’t coming “someday.” It’s here. Students are using it daily. Teachers are experimenting. Leaders are thinking, planning, guessing, and hoping they’re making the right decisions. And the districts that figure out how to lead this moment with clarity, courage, and humanity are going to define the next decade of public education.

Share

AI Won’t Wait—And Neither Should We

One of the opening sessions set the tone immediately: students are already building AI habits, identities, and relationships—long before schools have developed coherent frameworks. When a presenter shared that one in three teens already have a “chatbot friend,” the room went silent. Not because people were afraid, but because the implications were suddenly impossible to ignore. Students aren’t waiting for district guidance. They’re exploring, creating, and sometimes stumbling through these technologies on their own.

That reality reframed the entire conference. The question was no longer whether AI belongs in schools. The question was whether districts can lead with enough clarity to shape student AI use into something responsible, equitable, and empowering. Inaction isn’t neutral—it’s abdication. And in 2025, schools can’t afford to sit back.

The Framework That Anchored My Thinking

The infographic I’m including in this post does an incredible job capturing what many presenters echoed throughout the conference: meaningful AI leadership requires districts to develop governance and instructional readiness at the same time. One without the other is incomplete. You can’t build teacher confidence without policies, and you can’t build policy without understanding classroom impact. Strategy and practice grow together, like braided roots of the same tree.

The sessions reinforced this point again and again. Effective district AI planning doesn’t begin with the tools—it begins with the mission. Several speakers asked us to take one phrase from our district’s mission and ask: How should AI support this? It was such a simple exercise, but it snapped the work into focus. When AI becomes an extension of your values—not an add-on or a novelty—clarity comes naturally. You stop chasing shiny objects. You stop launching random pilots. You stop reacting. You start leading.

That’s meaningful innovative leadership.

From Tools to Thought Partnership

Another major theme was the shift from treating AI as an assistant to treating it as a thought partner. In the session on building AI-infused classrooms, presenters stressed that teachers need direct exposure to large language models—not just the “wrapper apps” built around them. Those apps are helpful, but they hide the actual reasoning engine. If teachers never learn how LLMs think, question, hallucinate, or refine ideas, we risk turning AI into another shortcut tool that undermines rather than elevates the profession.

This shift—from tool user to AI-literate educator—is foundational. It’s a move toward understanding AI’s capabilities, limitations, quirks, and biases. It’s learning to critique outputs, not just consume them. It’s recognizing that AI can accelerate thinking, but humans must still guide the cognitive process.

To make this point stick, several presenters used an analogy I’ll now repeat often: AI is an eager intern. It has read every textbook on earth but has never taught a class. It needs direction, structure, correction, and expectations. That analogy reduces fear while reinforcing the value of human judgment. It gives teachers permission to try, adjust, and refine—and positions AI exactly where it belongs: supportive, not authoritative.

Rethinking Academic Integrity: Curiosity Over Compliance

Some of the most compelling conversations centered on academic integrity. Instead of agonizing over whether students will cheat with AI, presenters challenged us to ask a different question: What needs to change so students don’t want to cheat?

That shift—from policing to motivation—opened up a richer discussion about curiosity. Research shared during the sessions showed that learning sticks when curiosity drives it. AI, when used thoughtfully, can fuel that curiosity by helping students explore alternative interpretations, challenge assumptions, create hypothetical scenarios, and receive immediate feedback. Several teachers described students becoming more engaged when AI helped them question characters in a novel or test out “what if” scenarios in social studies.

Academic integrity isn’t a monitoring problem. It’s a culture problem. Schools that cultivate curiosity—and teach students to use AI transparently and reflectively—build far stronger integrity than those that focus solely on detection.

A Future Arriving Faster Than We’re Ready For

Across keynotes, classroom demonstrations, and leadership conversations, there was a shared understanding that the pace of AI is accelerating faster than school systems traditionally adapt. Speakers projected that by the end of 2025, AI-driven tools will be present in the overwhelming majority of classrooms. Another presenter emphasized a more startling point: by the time today’s kindergarteners graduate, AI systems may be operating at levels of capability we can barely conceptualize today.

The takeaway wasn’t fear—it was urgency. Schools must become adaptive organizations, not reactive ones. Policies must evolve. Procurement must evolve. Teacher development must evolve. And leadership—perhaps most importantly—must evolve.

The Tools That Truly Surprised Me

I’ve spent the last year diving into every major AI platform, and I still left NYSCATE blown away.

NotebookLM feels like a true breakthrough. It doesn’t just “use AI”—it builds a knowledge engine from your own materials. Lesson plans become more coherent. Curriculum becomes more accessible. Research becomes more actionable. Documentation becomes easier to manage. It respects sources, cites accurately, creates learning materials instantly, and acts like a genuine collaborator rather than a generic assistant.

Gemini Gems opened up even more possibility. The idea that we can build custom micro-models—“specialist brains”—for specific tasks completely changes how districts might scale support. Imagine a Gem designed specifically for IEP writing, or Regents prep, or parent communication, or analyzing student data. This is the beginning of personalized AI infrastructure for schools.

But the biggest shift for me was realizing that my best workflow now involves letting Gemini and ChatGPT evaluate each other. I’ve begun treating them like consultants. For any major task—PD planning, policy drafting, writing, analysis—I have Gemini produce a draft, let ChatGPT critique it, then reverse the process. The models expose each other’s blind spots. They sharpen each other’s ideas. They help me see nuance faster. The result is consistently stronger than anything either tool produces alone.

This isn’t “prompt engineering.”
It’s AI reasoning—and it’s where the profession is headed.

What Meaningful Innovative Leadership Really Looks Like

After hearing from district leaders, technology directors, coaches, and keynote speakers, a clear picture of modern leadership emerged. Meaningful innovative leadership isn’t about being the smartest person in the room or the tech expert with all the answers. It’s about creating coherence where others see chaos, designing guardrails while encouraging experimentation, and leading with humanity in a time when it would be easy to retreat into hierarchy.

One leadership session captured this beautifully. The speaker reminded us that leadership is built on humanity—not titles. People need clarity and feedback. They need someone who sees them. They need someone who can get a room full of strong voices moving in the same direction. They need compassion and courage in equal measure. And sometimes, they need a leader who can laugh at the “low-key panic” moments that come with doing hard, meaningful work.

That’s the kind of leadership AI requires—not robotic efficiency, but deeply human clarity.

Share

Where Do We Go From Here?

One question closed multiple NYSCATE sessions, and I’ve been thinking about it constantly since returning home:

Where is your district on the AI readiness roadmap—and what is the next step you need to take?

Not the five-year plan.
Not the perfect system.
Not the ideal rollout.

Just the next right step.

That’s how meaningful innovation begins.
And that’s how it grows.

We don’t need to have everything figured out.
We just need to start leading with clarity, strategy, and humanity—one intentional step at a time.

Stay Curious. Stay Caffeinated. ☕

Note: This post, as all my posts are curated with the assistance of AI

A few weeks ago, I was in Pittsburgh, and the city’s vibe resonated with everything we talk about here at Brewing Innovation. Pittsburgh is often synonymous with steel, but it’s a city defined less by the metal it once produced and more by the fierce, innovative spirit the sheer grit of the people who built it, and then rebuilt it.

That transformation—from the furnaces of the Steel City to a hub for tech and robotics—holds the perfect blueprint for school leaders and educators today. We’re not waiting for change; we’re forging it ourselves, one bold move at a time.

Subscribe now

Why Every Innovative Educator Needs “Pittsburgh Grit”

Pittsburgh’s history isn’t just about production; it’s about pivoting. When the steel industry collapsed, the city didn’t crumble; it diversified, reinventing itself as a leader in healthcare and technology. This reinvention wasn’t luck; it was grit in action: the courage to fail fast, embrace new tools, and lead with a clear, forward-looking vision.

This is the mindset we need in education right now. We face our own “furnace challenges”—burnout, funding pressures, and a rapidly changing technological landscape. We can’t just maintain the status status quo. We have to become masters of the pivot.

Pittsburgh Grit for Educators Means:

1. Embracing the Furnace: Innovation starts with friction. Just as steel is forged in extreme heat, truly effective EdTech solutions are forged by testing them against real classroom pain points. Don’t chase the shiny new tool; use AI to solve the problem that’s causing burnout or widening learning gaps. (Think of our work streamlining IEPs with Gemini—that’s anchoring innovation in the friction point.)

2. The Rooney Legacy: Building a Dynasty of Excellence: The Steelers, guided by the Rooney family for generations, never rebuilt by accident. They focused on culture and consistent performance, creating two dominant dynasties: the “Steel Curtain” of the 70s and the hard-nosed teams of the 2000s.

   • From Chuck Noll to Mike Tomlin: Like legendary coaches Chuck Noll, Bill Cowher, and Mike Tomlin, effective school leadership is about fostering core values that endure, regardless of the technology changes. Consistency in vision (innovation as efficiency, not novelty) creates psychological safety for teachers to experiment.

   • The Blueprint of Grit: Think about players like “Mean Joe” Greene, Jack Lambert, Jack Ham, Mel Blount, and Franco Harris of the 70s, or the defensive savagery of Troy Polamalu, James Harrison, and T.J. Watt. These players didn’t just win; they were relentless. In EdTech, this translates to building systems, not running pilots. Our goal isn’t just to try an AI tool for a semester; it’s to embed AI literacy and efficient workflows into the DNA of the school, just like the Rooney family built a dynasty on enduring values.

3. The Three Rivers Strategy: Connecting Data, Tools, and People: Pittsburgh is where the Allegheny and Monongahela Rivers meet to form the Ohio River. This confluence is critical. In innovation, we need a confluence of:

   • Data (Allegheny): What problem are we actually solving? (e.g., 75% of our special education teachers report burnout during IEP season.)

   • Tools (Monongahela): What AI/EdTech tool can precisely address this? (e.g., Google Gemini for drafting PLPs.)

   • People (Ohio): The system we build to connect them. (e.g., Clear ethical guardrails, role-specific guides, and protected time for PD.)
     Innovation fails when any of these three elements flow separately. True progress happens at the junction.

Practical Strategy: Forging Your Own EdTech Steel

To apply this Steel City mindset, focus on the process of creation, not just the finished product.

1. Find the Iron Ore

Pittsburgh Analogy: Locating the raw material.

Education Application: Identify the Pain Point. Where is the biggest time sink or the most frustrating instructional bottleneck? (e.g., differentiated assignments, giving writing feedback).

2. The Blast Furnace

Pittsburgh Analogy: Applying intense heat to purify the metal.

Education Application: Test with Intentionality. Isolate the problem and use the simplest tool possible (like Brisk Teaching) to address it. Measure time saved or student engagement. Fail fast, learn faster.

3. Tempering

Pittsburgh Analogy: Strengthening the metal through heat and cooling.

Education Application: Build Trust and Culture. Model the tool’s use yourself. Share successes and failures openly. Only ask staff to adopt a tool after it has proven its worth and fits into their existing workflow.

4. Structural Steel

Pittsburgh Analogy: The product is ready to build the city.

Education Application: Embed into System. Once proven, anchor the tool and its associated literacy into district policy, PD plans, and the Portrait of a Graduate. It becomes a foundational part of how you operate.

If we stop viewing AI as a threat and start seeing it as a tool, we can save hundreds of hours of busywork—time that can be reinvested into the high-impact, human connections that truly build better schools. AI is such a powerful tool and is a large reason why I am able to create this Substack by helping me synthesize my ideas and edit my writing. It is such a powerful way to amplify ourselves and spread positive change.

Pittsburgh rebuilt itself with grit, creativity, and a willingness to embrace a new future. Educators are doing the same right now. Let’s lead with that same bold, blue-collar spirit.

Your Weekly Shot of Insight:

• Actionable Takeaway: Think of a legendary defensive player like Jack Lambert or Troy Polamalu. They mastered their one job with relentless focus. Identify one tool (like Gemini) and one administrative task (like drafting emails) to master this week. Focus equals efficiency.

• Reading Corner: Look up the history of the “Art Rooney Rule” in the NFL. Its intention—to foster equitable opportunity—mirrors the equity goal of using AI to level the playing field for all students and teachers.

Let’s build better schools—one bold move at a time.

Stay Curious, Stay Caffeinated ☕

— Steve

Share

The big picture: more students, worse outcomes

This month Google published an extensive Future of Learning report. It is a blueprint for how AI could reshape education. The report opens by noting that 90 % of primary‑school children are enrolled worldwide, yet learning outcomes are falling and 272 million children remain out of school. It emphasizes that education faces multiple pressures—pandemic learning loss, teacher shortages, mental‑health challenges and socio‑economic disparities—and that the world will need 44 million additional teachers by 2030. Simply doing more of the same will not close the gap.

In explaining why AI matters, the report argues that technology alone won’t save us; instead, AI should bring learning science to everyday teaching. AI systems can spark active participation, encourage deep practice and embed spaced repetition. When well designed, they lighten the burden on educators by making powerful strategies more accessible and move students from passive consumption to active application. The report outlines opportunities to personalize teaching and tutoring—adapting content to a learner’s context, identifying knowledge gaps and providing tailored feedback—and to make almost anything learnable by stripping away jargon and connecting new information to prior knowledge. AI tutors should complement human teachers, serving as non‑judgmental, always‑available helpers, not replacements.

Subscribe now

How Dan Fitzpatrick frames it

Dan Fitzpatrick’s recent newsletter distills similar themes. After keynoting multiple education conferences, he describes the Future of Learning report as a vision statement and reality check: AI could make learning more personal, accessible and even more human. Fitzpatrick highlights the paradox that a machine‑born force might restore the human side of education we’ve been losing. He summarizes five promises of AI—embedding learning science into everything, personalizing at scale, making anything learnable, removing barriers and freeing teachers to teach—but he also stresses the messy truth that every promise comes with an equal and opposite peril. Google’s own authors concede that AI models hallucinate, replicate biases and raise privacy concerns, that they can erode critical thinking if they simply deliver answers, and that they risk widening inequality for the “5 %” of highly motivated learners. Fitzpatrick urges educators to rethink assessment and design assignments that reward process over product.

Google’s tools: turning research into practice

The Keyword blog post accompanying the report sketches how Google intends to translate research into tools. Gemini offers “guided learning” paths that coax learners to reason rather than copy answers; NotebookLM can turn your own sources into quizzes, flashcards or immersive audio/video experiences and new features in YouTube and Search make these platforms more conversational. In Google Classroom, a suite of no‑cost AI tools helps with lesson planning and administrative tasks so teachers can focus on inspiring students. Google stresses that AI should cultivate understanding and curiosity, not just provide quick answers, and that tools must be grounded in best‑in‑class learning science and built with educators in mind. Importantly, Google notes that academic integrity and equitable access are top priorities—they are experimenting with assessments such as oral exams and portfolio projects that AI cannot easily replicate.

A cautionary tale: AI and IEP writing

While these visions take shape, educators are already using AI to solve immediate pain points—sometimes without clear guidance. An Education Week investigation found that 57 % of special‑education teachers surveyed in the 2024‑25 school year used AI to help draft individualized education programs (IEPs), up from 39 % the prior year, and 15 % used AI to write an entire plan. Teachers say AI platforms help them manage paperwork and free up time for instruction.

However, this convenience comes with risks. Large‑language models are trained on corpora that may not accurately represent people with disabilities, raising concerns about bias; educators can inadvertently violate student‑privacy laws if they paste sensitive data into consumer tools; and there is little official guidance—only two U.S. states currently require districts to have AI policies. Research by Olivia Coleman and Danielle Waterfield shows that AI‑generated IEP goals are comparable in quality to teacher‑written ones, and training improves outcomes. But they emphasize that AI should be a writing partner, not an autopilot; drafts must be reviewed and revised; bias must be checked; AI usage should be disclosed to parents and documented; and the effectiveness of prompts should be tracked. As educator Elizabeth Bettini notes, special‑education teachers often work 10 hours a week outside school on paperwork; AI can relieve some of that cognitive load, but ethical use and transparency are non‑negotiable.

Share

Navigating this new chapter with intention

What I love about Google’s report and Dan Fitzpatrick’s commentary is their balanced tone. They acknowledge that AI’s greatest breakthroughs in education will be societal, not merely technical—a sentiment that aligns with Brewing Innovation’s philosophy. The real opportunity isn’t to chase shiny tools but to design learning experiences that are human‑centered and future‑ready. Here are a few guiding principles I’m taking away:

• Focus on pedagogy first. Whether you’re experimenting with Gemini or ChatGPT, start with sound learning objectives and use AI to support deep practice and reflection.

• Treat AI as a co‑teacher, not a cheat sheet. Use generative tools to scaffold thinking and pose questions, not just to provide answers. Redesign assessments toward debates, portfolios and oral exams that emphasize reasoning.

• Stay transparent and ethical. When using AI for administrative tasks like IEPs, disclose it to families, document edits and check for bias. Develop AI literacy so you can monitor outputs critically and teach students to do the same.

• Invest in equity. Remember the “5 % problem”. Without deliberate investment in local languages, accessible devices and teacher training, AI could widen gaps rather than close them.

By keeping these principles front and center, we can turn the Future of Learning from a blueprint into a reality that uplifts every learner. AI can help us brew a more innovative, humane and efficient education system—if we pour with care.

How are you using AI tools in your classroom? Share your wins and cautionary tales in the comments—let’s learn together.

Stay Curious. Stay Caffinated ☕

Further Reading:

• AI and the Future of Learning

• AI and learning: A new chapter for students and educators

Author’s note: I have been learning Agent Mode with ChatGPT and I thought this would be a great opportunity to try it out. I have used the chat to help with posts in the past, but this is a great way to synthesize information. Let me know your thoughts in the comments!

Subscribe now

In many districts, educators had a choice: block AI or embrace it.

When ChatGPT first made its way into public hands, the instinct was to restrict it and to protect learning. But quickly, it became clear that blocking wasn’t protection; it was avoidance.

Students were already using AI on their phones, in their homework, and (yes) sometimes on Snapchat. The question wasn’t if they’d use it, but how we’d guide them to use it well.

Many innovative educators made a different choice: they decided to teach access.

Subscribe now

Why Access Matters

As a former special education teacher, I’ve seen how access changes everything.
When a student struggling with reading can use Gemini to summarize a passage or highlight key ideas, that’s not cheating, it’s scaffolding.

When a student with anxiety can rehearse a presentation using ChatGPT’s role-play feature before delivering it live, that’s not avoidance, it’s growth.

And when a high-performing student uses AI to test an argument or refine their writing, that’s not automation, it’s acceleration.

AI, when used responsibly, helps level the playing field. It gives every student, from those catching up to those racing ahead, a coach, a tutor, and a creative partner.

Understanding the Machine Behind the Magic

Before we can talk about access to AI, we need to talk about understanding it.

In his recent article, “How AI Tools Like Copilot & Gemini Actually Work — for Non-Techie Educators”, author and education leader Al Kingsley offers one of the clearest explanations I’ve seen of what’s really happening under the hood of tools like ChatGPT and Gemini.

Kingsley compares these systems to very advanced autocomplete engines. They don’t think, reason, or know in the human sense, they predict. They analyze patterns from billions of examples of text and use those patterns to generate likely next words, ideas, or explanations.

That might sound simple, but it’s foundational to AI literacy. When teachers and students understand that AI isn’t pulling from a single webpage or “knowing” an answer, it’s making a probability-based prediction their entire relationship with the tool changes.

They stop asking, “Can I trust it?” and start asking, “How do I guide it?”

Kingsley goes on to explain how educators can anchor AI’s power to school-specific knowledge through Retrieval-Augmented Generation (RAG), linking the model to trusted documents like curriculum maps or district policies so it produces grounded, reliable responses. He also offers practical guidance on avoiding bias, protecting data, and designing effective “prompt briefs” that mirror lesson planning.

In other words, Kingsley gives us the technical backbone for what AI literacy looks like in schools:

• Demystify how AI works — prediction, not perception.

• Model responsible skepticism — question outputs, check sources.

• Anchor to trusted context — use school data, not random internet text.

It’s the perfect complement to access. Because if access is the doorway to learning with AI, understanding how it works is the key that unlocks it.

What Access Looks Like in Practice

Giving students access to AI doesn’t mean giving them free rein. It means designing learning environments with guardrails and intention.

Here’s what we’ve found works best in classrooms:

1. Structured Environments
Use school-managed accounts like Gemini for Education or ChatGPT Edu. These platforms meet privacy requirements, limit exposure to harmful content, and give educators visibility into student use.

2. Transparent Modeling
Teachers show how to use AI openly, not secretly. They think aloud:

“Let’s see what Gemini suggests. Now, how can we verify that?”

This normalizes curiosity and teaches digital skepticism, not dependency.

3. Reflective Routines
Students explain how they used AI. A short line like,

“I used ChatGPT to brainstorm ideas but wrote my own final version,”
turns a potential shortcut into a moment of metacognition.

The Guardrails Are Catching Up

Until recently, student AI access was a gray area. But 2025 has brought real progress:

• ChatGPT Teen Experience: OpenAI launched a version for ages 13–17 with parental controls, safer content filters, and transparent data use.

• Gemini for Education: Google’s student-facing Gemini runs in secure, district-managed environments under FERPA compliance.

• Responsible AI Teaching Framework: Google’s new guide supports teachers in introducing AI ethics, bias, and digital literacy alongside skill development.

The message is clear: the era of “ban it and hope” is over. We’re entering the era of teach it and guide it.

Share

Building Habits, Not Just Skills

AI literacy doesn’t happen through one PD day or a flashy app. It happens through habits.

James Clear’s Atomic Habits reminds us that small, consistent actions compound into transformation. The same principle applies here.

Every time a teacher says,

“Check what the AI got wrong,”
they’re building a micro-habit of critical thinking.

Every time a student documents how they used AI ethically, they’re building a micro-habit of accountability. And every time a classroom compares human and AI writing side-by-side, they’re building a micro-habit of discernment.

These daily, low-stakes routines build the muscle of AI literacy , not just for students, but for teachers, too.

Future-Proofing Students

Kevin Roose, in Future Proof, says that to thrive in the age of automation, we need to focus on what makes us most human: creativity, empathy, and originality.

That’s what true AI access is about. It’s not about mastering the tool; it’s about developing the judgment to know when not to use it.

One of Roose’s “rules for humans” is to be surprising, social, and scarce. That’s our job as educators, to help students do what AI can’t: to collaborate, empathize, and innovate in real contexts.

A New Definition of Access

When we talk about “access,” we’re not just talking about logins or devices.
We’re talking about equity.

We can’t allow AI fluency to become a privilege for only the tech-savvy or affluent. Access means giving every learner — in every zip code — the opportunity to understand, question, and shape AI’s role in their world.

The real question isn’t if students should use AI.
It’s who gets to learn how to use it well.

Leadership in the Age of AI

For administrators, this isn’t about chasing shiny tools. It’s about building systems that protect teachers’ time, promote innovation, and model responsible experimentation.

Here’s what we’ve learned works:

• Anchor AI in your district’s instructional priorities, not your tech plan.

• Build pilot teams of teachers to test and iterate on real workflows.

• Model AI use yourself, in emails, agendas, feedback loops, and reflection.

• Create safe spaces for curiosity and failure.

When leaders model learning with AI, it gives everyone else permission to explore.

Access Without Abdication

Access doesn’t mean letting AI take over. It means teaching students how to think alongside it.

When I see a student use Gemini to plan an essay outline — and then improve it with their own insights — that’s not automation.
That’s amplification.

When I watch a teacher use Gemini to reword an IEP goal for clarity and save 10 minutes that’s not replacement. That’s restoration.

AI, used with intention, gives teachers back time and gives students back ownership.

Try This Week: Building AI Habits in Your Classroom

Here are a few small, doable routines to build AI literacy in your school:

✅ AI Audit: Ask students to evaluate an AI-generated summary for accuracy or bias.
✅ Prompt Reflection: After using an AI tool, have students write one sentence about how they used it and what they changed.
✅ Compare & Contrast: Use Gemini or ChatGPT to analyze a topic, then have students research a human source and compare findings.
✅ Teach the “Why”: Ask, “Why would you trust (or not trust) this output?”

One conversation, one reflection, one habit at a time, that’s how we build a culture of responsible AI use.

The Horizon Ahead

We’re still writing the rules. But one thing is clear: AI access is no longer an experiment: it’s a literacy.

If we get it right, AI won’t make learning easier. It will make it more human.

Our job now isn’t to predict the future.
It’s to help students shape it — with integrity, creativity, and courage.

Further Reading

• How AI Tools Like Copilot & Gemini Actually Work — Al Kingsley (LinkedIn)

• Teaching Responsible AI (Google for Education PDF)

• The AI Teacher’s Decalogue — Ioannis Anapliotis (LinkedIn)

• Education in the Gemini Era — YouTube

• Future Proof by Kevin Roose

• Atomic Habits by James Clear

  Share

Stay Curious. Stay Caffinated ☕

There was a time when tutoring meant finding a retired teacher who could squeeze you in on a Tuesday at 6:30.
Now, every student can have a tutor anytime, anywhere.

AI tutors like Google Gemini Gems, Brisk Teaching’s Brisk Boost, Google Notebook LM, and ChatGPT’s Learning Mode are reshaping how students learn. They’re not replacing teachers. They’re extending what’s possible.

As someone who started my career in special education, I’ve spent years trying to close learning gaps for students who struggled to keep up—not because they weren’t smart, but because they needed time, repetition, and feedback that the traditional system couldn’t always provide. The promise of AI tutoring is that those supports are no longer reserved for the lucky few who can afford them. They can now exist for every learner, every day.

Subscribe now

The Big Advantage: Access and Equity

AI tutors are an equalizer. Every student, regardless of background, ability, or zip code can get immediate feedback, explanations, and extra practice tailored to their needs. That’s transformative.

When I was teaching, I had students who needed instructions read aloud multiple times before they could get started. Others needed step-by-step modeling to make sense of a math problem. Some needed reminders to slow down and check their work before moving on. AI tutors can now provide that layer of patient, individualized support.

For students with learning disabilities, this means real-time scaffolding.

• A student with a reading disability can use Brisk Student Boost to simplify complex text or generate questions that break down meaning, sentence by sentence. With the teacher as the guide of the assignment.

• A student with ADHD can use Notebook LM to summarize key ideas from class notes and ask clarifying questions about what they missed when attention drifted.

• A student with processing delays can use Gemini Gems to revisit classroom material in a conversational way, without feeling embarrassed to ask again.

And just as important—these tools can help teachers differentiate without doubling their workload. A teacher can create a Gemini Gem or Brisk Boost once and reuse it across classes, customizing the level of support for different learners.

How Each Tool Expands the Learning Environment

Google Gemini Gems
Gems let teachers or students design custom AI tutors—essentially digital learning partners trained on classroom materials. You can make a Gem act like a coach, explainer, or Socratic guide.
For example, a Global Studies teacher can upload key documents and create a Gem that helps students practice sourcing and contextualization questions. A student can interact with it on their own time, asking, “Why was this document significant?” or “How would this connect to another event we studied?”
That’s individualized inquiry without pulling the teacher away from 25 other students.

Brisk Teaching’s Brisk Boost
Brisk Boost integrates directly into Google Docs and Classroom. Students can highlight any text and instantly get support definitions, summaries, scaffolded prompts, or example responses.
For a struggling reader, that means they can clarify meaning as they go instead of stopping in frustration. For a teacher, it means less time reteaching basic comprehension and more time guiding analysis and reasoning.
Special education teachers can create differentiated Boosts aligned to IEP goals like simplifying word problems or providing sentence starters for writing.

Notebook LM
This tool is like a digital study coach. Students upload notes, slides, or readings and can chat with their materials.
It’s ideal for students who need help organizing thoughts or connecting big ideas. Notebook LM can answer questions like, “What were the three main causes of the French Revolution according to my notes?” or “Explain this concept using simpler language.”
For students with executive functioning challenges, it’s like having a virtual co-pilot that keeps learning structured and interactive.

ChatGPT Learning Mode
This version of ChatGPT acts more like a mentor than a machine. It remembers what you’re learning, adapts its tone, and personalizes its approach.
Students can ask it to quiz them, walk them through math problems, or challenge their reasoning. Teachers can model how to prompt responsibly, turning AI into a tool for thinking rather than shortcutting.
When guided well, Learning Mode becomes a sandbox for curiosity, one where students can explore ideas deeply, safely, and creatively.

The Pros and Cons

Like any innovation, AI tutoring brings both promise and pitfalls. But the advantages far outweigh the challenges, especially when schools lead with purpose and pedagogy.

The Pros

• Equity and Access: Every student, regardless of ability or income, gains access to high-quality, personalized feedback.

• Individualized Support: Students with disabilities receive immediate scaffolds that align with their needs—without waiting days for feedback.

• Teacher Empowerment: AI tutors handle the repetitive reteaching, allowing teachers to focus on higher-order feedback and relationships.

• Student Confidence: Learners who might hesitate to ask for help in class can engage privately and build confidence at their own pace.

• Deeper Learning: High-performing students can use AI tutors to extend beyond the curriculum—debating, testing, and refining their understanding.

• Scalable Differentiation: Teachers can customize supports once and apply them across groups, saving hours of planning time.

• Promotes AI Literacy: Students learn how to use AI responsibly—questioning, verifying, and refining rather than copying answers.

The Cons (and How to Manage Them)

• Some students will try to let the AI do the work for them. This isn’t new, it’s just a new form of academic shortcutting. The key is explicit modeling of how to use AI as a learning partner, not a replacement.

• Misinformation and bias can still occur, especially with open-ended AI systems. Teachers must build verification routines into lessons (“How do you know that’s true?”).

• Data privacy remains an ongoing concern. Schools should prioritize tools with educational safeguards like Brisk and Google’s offerings.

A Special Education Lens

For years, I’ve seen special educators fight to bridge the space between a student’s potential and their performance. We’ve built scaffolds, accommodations, and relationships to make that bridge stronger, but the gap often persists because time and attention are finite.

AI tutors extend that bridge. They don’t replace individualized instruction or human connection; they make them more sustainable. A student can use AI to rehearse a skill, receive immediate corrective feedback, and then bring questions to their teacher the next day. That cycle of practice, reflection, and reinforcement accelerates growth.

And the best part? It keeps students in the learning environment. Instead of being pulled out for constant remediation, they can stay in the flow of class, with AI providing just-in-time support. It’s a form of inclusion that works quietly but powerfully.

What About High Performers?

High-achieving students need challenge as much as struggling learners need support. AI tutors can become their intellectual sparring partners.

• A student in AP English can use ChatGPT Learning Mode to test interpretations of a text or refine argument structure.

• A science student can use Gemini to simulate data experiments or generate hypothesis scenarios.

• A history student can use Notebook LM to cross-analyze sources or draft historical arguments, then refine their reasoning through guided prompts.

These students learn grit and discipline when AI is used to push, not pacify.
The focus shifts from “Can I get the right answer?” to “Can I build a stronger one?”
That’s AI literacy at its best, students learning how to think with technology, not through it.

The Real Work: Guided Independence

The danger of AI tutoring isn’t in the technology—it’s in the temptation to skip the struggle. But when schools and teachers frame AI as a thinking companion, not a crutch, students develop independence, confidence, and clarity.

This is especially true at home. When students are on their own, many will turn to unsupervised tools like Snapchat’s AI to do the work for them. That’s where our role as educators comes in. We have to intentionally teach them to use the right tools, the right way, building habits that promote understanding over convenience.

There’s no perfect system, but that’s the world our students are entering: one where every choice is mediated by technology, and integrity must be learned as a skill.

Brewing the Takeaway

AI tutors don’t make schools less human, they make them more humane.
They ensure no student slips through the cracks because they needed more time, more explanation, or more feedback than the system could provide.
They empower teachers to focus on relationships, coaching, and creativity rather than repetitive reteaching. And they push high performers to deepen their reasoning, curiosity, and persistence.

The future of learning isn’t about replacing teachers with technology. It’s about equipping both teachers and students to think boldly, learn deeply, and grow intentionally, with AI as a partner in the process.

Share

Stay Caffeinated. Stay Curious. ☕

Every week I read the rebuttals, scroll through the hot takes, and hear the concerns about AI in schools. The headlines don’t help: one week it’s “AI will revolutionize education,” the next it’s “AI will destroy learning as we know it.”

Caught in the middle are teachers who already feel stretched thin and leaders who are just trying to make smart choices for their schools.

Here’s the problem: most of the conversation is built on myths. And while we waste energy worrying about the wrong things, we risk missing the real opportunities (and challenges) that AI brings to the classroom.

So let’s bust three of the biggest myths about AI in schools and focus on what we should actually be paying attention to.

Share

Myth #1: “AI will replace teachers.”

Why people believe it:
It’s easy to see why. Automation has disrupted entire industries: factory jobs, retail, even journalism. So the fear is natural that if a chatbot can answer questions or write a lesson plan, why would we need teachers?

The reality:
Teaching is not just about delivering information. It’s about relationships. It’s about noticing when a student is struggling, coaching them through frustration, and celebrating their growth. No AI tool does that as a human connection.

What AI can do is lighten the workload that keeps teachers from doing the most important work. Imagine:

• Instead of spending Sunday night tweaking lesson slides, a teacher uses AI to adapt materials for different reading levels in minutes.

• Instead of staring at a blank screen to write IEP goals, AI drafts a starting point that the teacher personalizes.

• Instead of manually creating comprehension questions, AI generates options so the teacher can spend their time refining and choosing the best ones.

AI doesn’t replace teachers. It gives them back the most precious resource they’re losing: time to be human with their students.

What to really worry about:
The real risk isn’t teachers being replaced. It’s teachers being left behind. Educators who refuse to engage with AI may find themselves drowning in busywork while their colleagues are freeing up hours for creativity, connection, and coaching.

Myth #2: “AI makes cheating unstoppable.”

Why people believe it:
Type “write my essay” into ChatGPT and boom — instant product. TikTok is full of “AI hacks” for homework. It feels like academic dishonesty has gone nuclear.

The reality:
Cheating has always existed. AI didn’t invent shortcuts — it just gave students a new one. What’s changed is the visibility and speed.

Here’s the uncomfortable truth: if an assignment can be fully completed by AI, maybe it wasn’t that strong an assignment to begin with. This doesn’t mean we throw out essays or traditional tasks. It means we double down on teaching process, thinking, and ownership.

Examples from classrooms I’ve seen:

• A teacher uses Gemini to show three different thesis statements on the same topic, then has students critique which one is strongest and why.

• Another has students generate a draft outline with AI, but then requires a handwritten reflection explaining their revisions and reasoning.

• A science teacher asks students to fact-check an AI explanation of a lab experiment, which often leads to uncovering mistakes and deepening their understanding.

AI isn’t the end of academic honesty. It’s an opportunity to redefine it, but to teach integrity, discernment, and the ability to work with new tools responsibly.

What to really worry about:
If schools ignore AI out of fear, students will still use it, just poorly and secretly. And worse, they’ll graduate without the critical literacy to use these tools ethically in college or the workplace.

Myth #3: “AI will dumb down learning.”

Why people believe it:
If a student can ask a chatbot for the answer, why would they ever wrestle with tough material? It feels like thinking could be outsourced.

The reality:
AI doesn’t have to replace thinking. When guided well, it can actually deepen learning by scaffolding complex ideas, modeling examples, and giving instant feedback.

Picture this:

• A middle schooler struggling with fractions gets step-by-step explanations in plain language, with practice problems adjusted to their exact level.

• A high schooler in Global History can use AI to generate multiple perspectives on a historical event, then analyze bias and reliability.

• An English Language Learner can translate a passage, ask clarifying questions, and generate vocabulary practice tailored to their current fluency.

How it helps build foundations:
For struggling learners especially, AI can be a lifeline. It can break down intimidating concepts into smaller, digestible pieces. It can reteach content in different ways until something clicks. It can provide unlimited practice without stigma.

Instead of dumbing things down, AI can actually help students build the foundational knowledge they need to engage in higher-level thinking, the very skills we want them to master.

What to really worry about:
The danger isn’t students using AI. It’s teachers adopting it without intention. A shiny tool without purpose doesn’t elevate learning, it just distracts. But purposeful integration? That’s where the magic happens.

So… What Should We Really Worry About?

Not whether AI is coming for our jobs.
Not whether it’s making cheating easier.
Not whether it’s turning kids into robots.The real question is this: Will we use AI with purpose, or let it use us?

Educators who model wise use, teach discernment, and design classrooms where tools serve learning (not replace it) will prepare students for the world they’re actually entering. Those who ignore it risk leaving both themselves and their students unprepared.

Coming Next Week: The Toolbox

This week was about clarity — cutting through the noise and busting myths. Next week, we’ll get practical.

I’ll be sharing The AI Toolbox Every Teacher Needs This Year:

• A short list of the essential tools worth your time.

• Step-by-step classroom routines you can plug in tomorrow.

• Workflows that save time while strengthening student ownership.

Because AI in schools isn’t about hype. It’s about helping teachers do more of what only humans can do: connect, inspire, and lead learning.

Share

Stay Curious. Stay Caffeinated. ☕