Finding Our Voice within AI

As an educator who has spent decades in classrooms and now builds educational technology, I find myself at a crossroads that feels both daunting and exhilarating. The arrival of generative AI in our schools has created what creativity researcher, Ron Beghetto, calls a "critical inflection point" between two possible futures for education (OECD, 2026).

On one path, we risk creating what Beghetto terms "digital puppets": students who mindlessly repeat AI-generated content without critical thought or personal input. On the other, we have the opportunity to cultivate what Tara Westover describes as education as "a process of self-discovery, of developing a sense of yourself and what you think" (Gates & Westover, 2018).

Performance Without Substance

I've watched students engage in what I've come to think of as an "AI ventriloquist act." They present their ideas like ventriloquist's dummies. Their mouth moving, but their voice, their words, their ideas all come from the AI hidden behind the curtain.

From the outside, it looks like the student is "speaking" (writing the essay, solving the problem, answering the question). But in reality, they're just animating a script they didn't create and may not even understand. The performance might get a passing grade, but the student never learns how to find their own voice.

Learning coach, Justin Sung, observed this in his analysis of AI learning tools. They create an "illusion of learning" where students measure progress by how quickly they get through content rather than by actual understanding and application (Sung, 2024). Students become what Sung calls "passengers in their learning" rather than active drivers.

A Constructionist Alternative

At Sage.Education, we're building a different path, one grounded in Seymour Papert's constructionist pedagogy. One where learning occurs through the construction of tangible artifacts (Papert, 1980). Rather than asking students to consume AI outputs, we invite them to build their own AI tools. This "maker mindset" directly counters the cognitive offloading that research shows weakens critical thinking and deep learning (Jose, 2025).

When students build an AI assistant to help analyze documents or create a tool to visualize concepts, they engage in what Beghetto calls "structured uncertainty." They have clear objectives but must find their own paths to meet them. This process requires them to understand problems deeply, generate ideas, prepare for action, and plan their approach—the core components of creative problem-solving identified by Isaksen et al. (2011).

Making the Process Visible

The crucial element missing from most educational AI tools is what Sung identified as the fatal flaw: they give students the product, like a mind map, summary, or cue cards, without requiring the process of organizing and connecting knowledge. As Sung notes, "it is the process of trying to organize it that creates the learning. It is not the end product that matters" (Sung, 2024).

This is why educator oversight isn't just about monitoring for misuse; it's about pedagogical visibility.

At Sage.Education, we've built tools that let educators see not just what students produce, but how they get there. Educators can view conversation maps, track iteration patterns, observe model switching, and witness the creative process unfold. This allows teachers to provide what Sung calls "targeted guidance", asking the right questions at the right time to push students toward deeper understanding.

From Fast AI to Slow AI

Beghetto's distinction between "Fast AI" and "Slow AI" resonates deeply with our approach (OECD, 2026). Fast AI is the one-off question-and-answer approach that leads to dependency. Slow AI is the collaborative, iterative process where AI serves as what Beghetto calls "a partner in possibility thinking." Our platform encourages this slower approach: students "earn the response" through their own questions and curiosities, much like Sung recommends when he advises students to "pay for a response with a question or a curiosity or a gap in your knowledge that you've identified" (Sung, 2024).

Developing the Skills That Matter

UNESCO's AI Competency Framework for Students (2024) emphasizes creating with AI and ethical co-creation as core competencies. The World Economic Forum's Future of Jobs Report (2025) identifies analytical thinking, creative problem-solving, and AI literacy as essential skills for future readiness. These aren't skills developed through passive consumption or ventriloquist acts. They're cultivated through the messy, iterative, sometimes frustrating process of creation.

When students build their own AI tools, they develop what Beghetto describes as comfort with uncertainty. They learn that creativity requires both "originality and appropriateness"(OECD, 2026). They experience firsthand the cognitive challenge of connecting many pieces of information and develop strategies for navigating this complexity (Sung, 2024).

A New Educational Imperative

We stand at what Beghetto warns could become "a crisis of meaning in education" (OECD, 2026). If education becomes about delivering inert content for students to reproduce, machines will ultimately do that better. But education has never been just about content delivery; it's been about developing minds, fostering creativity, and cultivating human potential.

The path forward requires tools that don't just give answers but create spaces for exploration. It requires educators who can see not just the final product but the creative process. And it requires students who understand that, as Westover discovered through her own educational journey, the real learning happens in the struggle to make sense of things for yourself.

At Sage.Education, we're committed to building this future. One where students move from being digital puppets to becoming architects of their own learning, where AI serves not as a ventriloquist but as a collaborator in the creative process, and where education remains fundamentally human even as it embraces the most advanced technologies.

Isabelle Plante is Co-Founder and CEO of Sage.Education, an open-source, private-by-design AI platform that empowers students to become active creators rather than passive AI consumers. A veteran educator and professional musician, she holds an MA in Curriculum Instruction & Development and serves as President of the Quebec Music Educators' Association.

References: