This week I sat down with two Grade 7 students to conference on a math performance task, and we did something I want to make a habit of: we all read the feedback together and each of us annotated the documents as we went — their observations alongside mine.
The task and the evaluation were both AI-generated — a rooftop garden design problem spanning fractions, algebra, data analysis, and budgeting, aligned to all four categories of the Ontario Growing Success Achievement Chart. The feedback was detailed and criterion-referenced, with specific commentary for each question and conference prompts prepared in advance. Our focus was on that feedback loop: what did the assessment reveal, what did it miss, and what did the students themselves have to say about it.
As the three of us read through the documents together and something stopped us. The students had originally submitted a colour-coded grid diagram, but the AI had assessed the PDF without picking up the colour, effectively missing the diagram entirely and assuming it hadn’t been completed. Reading through the feedback together, the students flagged it.
We looked at the original, counted the squares section by section, and it became clear that every area matched their proposed redesign from Part B exactly — vegetables at 24 squares, herbs at 16, pollinator at 14, seating at 6, totalling 60 squares and 240 m² without a remainder. To produce that, you have to know your target area, divide by 4 m² per square, then design a compact rectangular region with that exact integer count, and do it for all four sections at once. That is spatial and proportional reasoning applied deliberately to a real constraint.
So we did what made sense: we redid the diagram with proper labels and resubmitted it for re-evaluation. The AI had flagged their use of the redesigned areas throughout the rest of the task as a cascading error — with the labelled diagram now in front of us, the re-evaluation told a different story.
Their answer was essentially: we proposed a new garden design in Part B, so we applied that new design everywhere — to the tray calculation, the budget, the diagram, all of it. If the plan changed, the whole plan should change.
The AI had read that as an error — careless copying of the wrong values. But when you heard them explain it, the logic held up. They weren’t being careless. They were being consistent. They made a design decision and followed it through. That is a different thing entirely, and it took the three of us together to surface it. That reasoning is coherent. It also cost them marks because the task instructions pointed back to the original areas, and we talked through exactly why, and what they’d watch for next time. Then they wrote their own feedback into the document alongside mine — what made sense, what they’d do differently, what they knew they’d got right.
The AI produced feedback precise enough to be worth conferencing over, but it was the three of us together — reading, questioning, counting squares — that turned the assessment into actual learning. One of the students explained their design decision and she was right about the logic, even though the task had different expectations. No model caught that. We did.
AI in the classroom isn’t about replacing the teacher. It’s about raising the floor of what’s possible so the ceiling — the conversation, the moment of genuine inquiry, the student explaining their reasoning to an adult who is actually listening — can happen more often, with better material underneath it.
✓Designed a rich, multi-strand performance task aligned to the Ontario curriculum in a fraction of the time
✓Generated criterion-referenced feedback for every question with strengthening and extending comments
✓Prepared conference prompts and next steps in advance
✓Re-evaluated when the diagram was resubmitted with labels — revised the score and the interpretation
✓Raised the floor: made a high-quality assessment process possible in the time available
What only we could do
✓Noticed the colour-coded diagram had been missed because the AI read a PDF and couldn’t see colour
✓Counted the squares together and understood what the precision of the layout actually meant
✓Asked the question: was this a deliberate design decision, not an error?
✓Understood that what the AI called a cascading error was actually a deliberate design decision — once they changed the plan, they applied it everywhere, and that only became clear when we talked
✓Co-authored the record: students wrote their own feedback into the document alongside the teacher’s
The ceiling is the conversation. AI raised the floor — better material, more detailed feedback, faster turnaround. But the learning happened when three people sat in a room, read the document together, and one student said something that made the teacher pause. No model catches that. That’s the point.
AI-Assisted Assessment & Feedback Conference
How a performance task, AI evaluation, and a three-way conference produced better learning than any one alone
1
AI-Generated Performance Task
A rooftop garden design problem — fractions, algebra, data analysis, budgeting — aligned to all four Growing Success Achievement Chart categories and the Ontario Math curriculum (2020).
AI builds the floor
2
Students Complete the Task
Two Grade 7 students work through all four parts, submitting written responses, calculations, and a colour-coded grid diagram of their garden layout.
Student work
3
AI-Generated Criterion-Referenced Evaluation
Detailed feedback aligned to each success criterion — what was done well, where marks were lost, descriptive strengthening and extending feedback, and conference prompts prepared in advance.
AI raises the detail
4
Three-Way Feedback Conference
Teacher and both students read the evaluation together, out loud. Each annotates the document. The students flag a colour-coded diagram the AI had missed. Squares are counted. The reasoning behind a design decision is surfaced that no model had identified.
The conversation the document made possible
5
Student Re-Submission
Students redraw the diagram with proper labels and resubmit for re-evaluation. The AI revises its assessment. The score changes. More importantly, the students understand exactly why.
Student agency
6
Assessment as Learning
Students write their own feedback into the document alongside the teacher’s. The record of the learning is co-authored. Growing Success fulfilled — not as compliance, but as genuine practice.
Assessment as learning · Growing Success
Rich Baxter is a Grade 7 teacher at John McCrae Public School (TDSB), Founder & CEO of Innovation in Education, and creator of the Lern2ern platform. He is shortlisted for the 2025 QS Reimagine Education Award in Nurturing Employability.
Introduction: Reconceptualizing Education in a Digital Era
The advent of artificial intelligence has initiated a profound reimagining of education’s purpose and methods. While much discussion centers on AI’s potential to streamline assessment or personalize learning, we must look beyond efficiency to consider more fundamental questions: Who owns knowledge in an AI-enhanced world? How might we reconfigure education systems to prioritize student agency? And how can emerging technologies help democratize learning rather than further entrenching existing power dynamics?
This piece explores how the convergence of AI with Web3 technologies could foster an education paradigm where students are not mere consumers of knowledge but active creators and owners of their intellectual contributions. Drawing from practical classroom experiences and emerging technological frameworks, I propose that this integration could address critical challenges in contemporary education while supporting democratic values that underpin healthy societies.
The Problem: AI Risks Reinforcing Traditional Power Structures
Current educational models often position students as passive recipients of knowledge, with power and authority concentrated in institutions, curricula designers, and assessment bodies. This imbalance risks being amplified rather than diminished by AI implementation. When AI tools are deployed within traditional frameworks, they frequently reinforce existing hierarchies – centralizing decision-making, standardizing outputs, and prioritizing measurable outcomes over creativity and critical thinking.
The typical application of AI in education today follows a distressingly familiar pattern: edtech companies develop proprietary algorithms that analyze student data, teachers receive analytics dashboards to track performance metrics, and students become subjects whose behaviors are increasingly monitored and shaped. In this scenario, AI serves not to liberate learning but to enhance surveillance and control, widening the power gap between technology providers and education stakeholders.
Meanwhile, students create content that feeds these systems without retaining ownership or receiving recognition for their contributions. Their essays, projects, and discussions become training data for AI models owned by corporations, with no attribution or compensation flowing back to the original creators.
The Possibility: A Decentralized Model of Learning
Web3 technologies—particularly blockchain, decentralized autonomous organizations (DAOs), and non-fungible tokens (NFTs) and tokenization—offer mechanisms to realign these power dynamics when thoughtfully integrated with AI systems. Together, they can create an education ecosystem that recognizes student agency and creativity while fostering community-centered learning.
In a classroom experiment called “Lern2ern” conducted over two years in my Toronto school, we implemented a primitive version of this vision. Students earned tokens for demonstrating key competencies like collaboration, critical thinking, and creativity. These tokens were initially tracked on an analog ledger, then digital spreadsheets, with plans to move them onto a blockchain. Local businesses became partners, providing tangible rewards for accumulated tokens.
This system represented more than gamification—it created a microeconomy where learning behaviors gained recognized value. Students who previously disengaged from traditional assessment models became enthusiastic participants when their contributions earned visible acknowledgment. More importantly, they began viewing themselves as creators rather than consumers of educational content.
The next evolution of this approach would integrate AI as a facilitator rather than controller of the learning process. AI systems could help students document their work, provide feedback on developing competencies, and connect them with relevant learning opportunities and collaborators. Meanwhile, blockchain would ensure that students retain ownership of their intellectual property, with smart contracts automatically attributing their contributions and managing permissions for how their work is used.
Practical Implementation: Redefining Educational Value
Imagine a future classroom where:
Students maintain digital portfolios authenticated on a blockchain, with AI helping curate and present their work for different audiences.
Learning artifacts created by students become NFTs, establishing provenance and enabling attribution even as works are remixed and built upon by others.
AI systems serve as learning companions that help students identify interests, set goals, and connect with mentors and collaborators globally.
Classroom communities function as DAOs, with students and teachers collectively governing how resources are allocated and how group projects evolve.
Assessment shifts from centralized testing to community validation, where demonstrated competencies earn tokens that represent recognized skills and knowledge.
This model acknowledges that in an age where information is abundant and AI can generate content on demand, education’s value lies not in knowledge transmission but in fostering human creativity, critical thinking, and collaborative problem-solving. It positions technology as an enabler of human potential rather than a replacement for human judgment.
Challenges and Ethical Considerations
This vision is not without challenges. Concerns about equity and access must be addressed, ensuring that decentralized education systems don’t simply reproduce existing societal divides. Privacy concerns are paramount, particularly when combining AI’s data hunger with the permanence of blockchain records. And questions of appropriate governance—who sets the rules in decentralized education systems—require careful consideration.
Educational approaches using Web3 and AI must embrace principles of consent, transparency, and justice. Students should understand how their data is used, have agency in deciding what is recorded on immutable ledgers, and participate in setting rules for technology use. Educational institutions must develop literacy not just in how to use these technologies but in critically examining their implications.
Moreover, we must acknowledge that technology alone cannot solve education’s challenges. The human relationships at education’s heart—between teachers and students, among peers, and within communities—remain essential. Technology should enhance rather than replace these connections.
Conclusion: Toward a Democratic Educational Future
The convergence of AI and Web3 technologies offers an opportunity to reimagine education as a fundamentally democratic enterprise, where students actively create knowledge, take ownership of their learning, and participate in governance of educational systems. This approach aligns education with broader democratic values of agency, participation, and community self-determination.
For policymakers, this suggests several priorities:
Invest in digital infrastructure that supports decentralized models, ensuring equitable access across communities.
Develop regulatory frameworks that protect student privacy and data rights while enabling innovation.
Support teacher professional development that empowers educators to facilitate rather than direct learning in technology-enhanced environments.
Encourage experimentation with decentralized education models, gathering evidence about what works and for whom.
Engage communities in educational governance, particularly in decisions about technology adoption and use.
The most crucial task for education in the age of AI is not teaching students to compete with machines but fostering the distinctly human capacities that make us more than computational beings: creativity, empathy, ethical reasoning, and collaborative problem-solving. By combining AI’s power with Web3’s potential for decentralization, we can create education systems that nurture these capacities while empowering students as active participants in knowledge creation.
As we navigate this technological transition, our north star must be a vision of education that serves not merely economic efficiency but human flourishing and democratic community. The choices we make now will shape not just how we learn, but what kind of society we become.
Transparency Statement: In the development of this piece, I utilized Claude.ai, an AI assistant by Anthropic. This approach was taken deliberately, as it aligns with the essay’s themes of critically engaging with AI tools while maintaining human agency. The ideas, experiences, and educational philosophy expressed are my own, drawn from my work in education and the Lern2ern initiative, with Claude.ai assisting in organizing and articulating these concepts. This collaborative process exemplifies how AI can serve as a tool that enhances rather than replaces human thinking—precisely the relationship between humans and AI that I advocate for in education.
Innovation in Education 