The Human Work of Teaching: How EdTech Can Use AI Without Losing Its Soul

Every generation of educators faces a wave of innovation that challenges what it means to teach and learn. It’s generative AI’s moment — a moment filled with both excitement and unease. AI promises personalization, instant feedback, and limitless information. Yet it also raises the question that anchors every educational era: How do we keep what is uniquely human in the act of teaching and learning?

For educators, that question is not philosophical — it’s practical. And for EdTech leaders, it’s defining.

The purpose that anchors us — to nurture understanding, empathy, and agency — remains unchanged. The real opportunity of AI in education is not to automate that purpose, but to amplify it.

Beyond Efficiency: Designing for Empathy and Depth

AI’s promise in education has been efficiency. It can grade, analyze, personalize, and predict. But if efficiency becomes the goal instead of the byproduct, we risk creating systems that are fast — but not wise.

In classrooms, empathy is not a soft skill; it’s a teaching strategy. It’s the way a teacher reads a pause, senses discouragement, and decides to shift an approach. It’s how we build trust, interpret silence, and know when to push and when to pause.

AI can summarize a text, but it cannot sense the hesitation behind a student’s silence. It can predict who might struggle, but it cannot understand why.  It can suggest next steps, but it cannot read the look that says, “I almost understand this — don’t give up on me yet.”

These are the micro-moments where learning either sticks or slips away — and they are profoundly human.

That’s why the ethical imperative for EdTech creators is not simply to make tools that are capable, but tools that are compassionate in design.

When we build AI systems for learning, we are not just creating software. We are making choices about what counts as learning.

• If we reward speed, we teach that speed equals intelligence.

• If we optimize for completion, we imply that effort is secondary.

• If we measure only what’s visible, we overlook how thinking actually grows — through struggle, reflection, and collaboration.

  
  

The future of AI in education should not be about automating the teacher. It should be about amplifying the teacher’s reach and insight, while preserving the heart of teaching — the human connection that drives learning.

Designing for the Human Voice

AI can predict, categorize, and recommend — but it cannot feel.  It cannot notice the pride behind a student’s “I did it!”  It cannot sense when curiosity is fading and decide to ask the right question to bring it back.

So, when we design educational AI, we must ask: Does this tool make space for the teacher’s voice, or does it speak over it?

The “human voice” in education isn’t a metaphor — it’s literal. It’s the teacher’s tone of encouragement. It’s the phrasing of feedback that guides rather than grades. It’s the presence that communicates, “You can do this,” even before a student believes it.

AI can model language, but not belief.

And belief is what moves learning from compliance to curiosity.

Some EdTech tools are beginning to find this balance.

• Writing coaches like Grammarly for Education and Quill use AI to suggest sentence-level improvements, but they also prompt reflection and revision — encouraging students to think about why their choices matter.

• Adaptive learning platforms like Khanmigo and Century Tech use AI to provide personalized tutoring, but they also intentionally design space for human teachers to guide, interpret, and contextualize that data.

• Assessment tools such as Formative and Gradescope use AI to surface trends, not dictate judgments — allowing teachers to focus on analysis and feedback instead of manual scoring.

These tools work best when they position AI as a listener, not a lecturer. They show what happens when technology extends human capacity rather than replaces human intuition.

Designing for the human voice means creating systems that make it easier for teachers to teach with empathy, not harder.

For EdTech leaders, understanding that distinction — between simulating human interaction and supporting it — is where responsible innovation begins.

AI as a Tool 

The most powerful role for AI in education isn’t as a “co-teacher.”  It’s as a tool that supports the craft of teaching — one that removes friction, reveals insight, and creates more room for the real work of learning: connection and curiosity.

AI can manage the mechanical parts of instruction — organizing data, generating exemplars, offering practice, or flagging learning trends — but it cannot teach. It cannot interpret confusion, inspire confidence, or recognize the quiet triumph of understanding.

Teaching remains, and must remain, human work.

When designed and implemented well, AI becomes an extension of teacher expertise, not a replacement for it. It helps educators:

• See patterns across student work.

• Offer individualized feedback faster, freeing time for discussion and reflection.

• Access curated content, scaffolds, or exemplars that spark new approaches to instruction.

For example, when an AI tool analyzes hundreds of student responses and highlights misconceptions, it doesn’t correct those students — it equips teachers to engage them with sharper insight. When it drafts a lesson structure or assessment idea, it doesn’t determine what’s taught — it gives teachers a starting point for creative decision-making.

In other words, AI can prepare the conditions for great teaching, but it cannot create great teaching. That distinction matters.

When we call it a teacher, we risk giving it moral weight it can’t carry. When we call it a tool, we keep the focus where it belongs — on the humans who teach, design, and learn.

AI should serve the classroom — extending the teacher’s reach, honoring their judgment, and reinforcing the relationships that make learning possible.

The most meaningful future for AI in education isn’t as a teacher, but as a tool that clears the routine so educators can do the transformative: connecting, motivating, and inspiring.  It can clear the noise so educators can do what they do best, teach.

It gives teachers back time to notice — to look up, listen, and lead learning rather than manage logistics.

Empowering teachers is only half the equation. Building tools that truly honor learning begins with how we design them

Building AI That Honors Learning

As AI becomes embedded in EdTech design, leaders must move from experimentation to intentionality. The question is no longer “What can AI do?” It’s “What should it do — and why?”

Thoughtful EdTech teams are starting to take this seriously. They are building ethical guardrails into their development cycles:

• Including educators and students in testing phases.

• Auditing for bias in datasets.

• Designing dashboards that illuminate, rather than obscure, the complexity of learning.

• Prioritizing data privacy as a foundation, not an afterthought.

  
  

Just as we once taught students how to research critically or cite responsibly, we now must teach them how to leverage AI thoughtfully — how to ask for feedback, identify what’s missing, and use the tool as a springboard for deeper thinking rather than a shortcut to an answer.

We’re beginning to see positive shifts in how AI supports instruction. For example:

• Generative lesson-planning tools like Curipod or MagicSchool.ai can generate materials, but they also allow teachers to edit tone, adjust rigor, and adapt content to their classroom — keeping professional judgment at the center.

• AI tutoring systems like ChatGPT’s educational versions are being built with guardrails to encourage metacognition — asking students, “What do you think next?” instead of simply providing answers.

• Equity-focused tools are being designed to close gaps in access, offering multilingual support, text-to-speech features, and adaptive reading levels — making learning more inclusive without assuming one-size-fits-all.

These examples show that balance is possible when human purpose leads technological power.

The Educator’s Seat at the Table

For AI in education to succeed, educators must have a voice in how it’s built. Not as end users testing a finished product, but as co-designers shaping what learning should look like in an AI-powered world. Educators bring a kind of intelligence that no algorithm can replicate — the lived understanding of how learning feels for a student, and how growth unfolds in real time.

If EdTech companies want AI to truly enhance education, they must listen to that expertise. Invite teachers, instructional designers, and curriculum leaders into the design process — early and often. Let them define the problems worth solving, not just react to the solutions already coded.

The most transformative educational technology has always come from collaboration — not between humans and machines, but between educators and engineers.

Humanity at the Heart of Learning

Technology will continue to evolve, as it should. AI will get faster, smarter, and more predictive. But the heart of education — the human work of seeing, guiding, and believing in learners — endures. AI can generate words, summarize texts, and predict patterns. But it cannot replace the spark that happens when a learner feels seen. That spark — the one that turns confusion into curiosity — is still human work.

The future of AI in education isn’t about teaching machines to act human. It’s about teaching humans to design machines that honor humanity.

The EdTech leaders who thrive in this moment will be the ones who remember: technology may transform learning, but only humanity can transform lives.  

AI is powerful, but educators give learning its humanity, intention, and direction. The future of learning isn’t artificial — it’s deeply, deliberately human.

Winnie O'Leary October 2025