Paulo Granata stands wearing a white blazer and a blue shirt in the Fischer Rare Book Library at the University of Toronto. — *What does it mean to create and learn in an AI-shaped world? SRI Faculty Affiliate* ***Paolo Granata*** *reflects on his new book* Generative Knowledge (2026), outlining a framework for understanding AI as a co-creative partner in research, education, and intellectual life.

As artificial intelligence becomes embedded across research, education, and cultural life, questions about knowledge—how it is created, validated, and shared—have taken on new urgency. In Generative Knowledge: Think, Learn, Create with AI (Wiley, 2026), SRI Faculty Affiliate Paolo Granata offers a conceptual framework for understanding AI not simply as a tool for efficiency or automation, but as an epistemic technology that reshapes intellectual work itself.

An associate professor of book and media studies at the University of Toronto’s St. Michael’s College, Granata is a cross-disciplinary scholar and the founder of the Media Ethics Lab and chair of the Toronto School Initiative. Drawing on media theory, epistemology, semiotics, and the philosophy of technology in his new book, Granata reframes learning, research, and creativity as generative, socially embedded processes—ones that unfold through iteration, collaboration, and tool-mediated inquiry.

In this conversation with the Schwartz Reisman Institute, Granata reflects on the core arguments of the book, the pedagogical and institutional implications of AI, and why cultivating what he calls “epistemic wellness” may be essential for navigating the emerging knowledge economy.

The following conversation has been lightly edited for length and clarity.

Schwartz Reisman Institute: Generative Knowledge reframes AI not as a tool but as a co-creative partner. How does this shift change how we create and share knowledge?

Paolo Granata: I believe we should consider AI not just as a tool, but as an interface for knowledge—one that will eventually change the ways we think. That means asking what the implications are of this new cognitive environment for how we create and produce knowledge. In the book, I focus on the idea of generative knowledge. While “generative” may sound like a buzzword today, it has a longer intellectual history. From my background in semiotics, for instance, generativity has long been central to how meaning and knowledge are understood.

The core issue is that we are now dealing with an unprecedented volume of information. In this context, the fundamental question becomes: how do we create new knowledge? This is not only a technical problem, but a cultural and epistemological one, relevant across disciplines and sectors.

The book adopts a constructionist approach, grounded in the idea that knowledge is an act of creation. Historically, knowledge was often understood as observing or interpreting reality. Today, what we know is increasingly what we make. Advances in knowledge are tied to advances in engineering, design, and production, even in intellectual and philosophical domains. What matters is not only the ability to evaluate reality, but also to create something new. This is why generativity is the key concept of the book, and why AI provides such a powerful case study for rethinking knowledge creation.

SRI: The book proposes six foundational principles—iteration, instrumentality, sociality, inquiry, learnability, and creativity—for understanding AI. How did this framework take shape?

Granata: The process was largely backward. I began by reflecting on the implications of AI for our knowledge economy, particularly in education and creativity. From there, the first principle—the iterative principle—emerged. It captures a simple but crucial idea: it takes knowledge to create new knowledge. Every insight builds on pre-existing understanding.

This has become especially relevant with AI. These tools can appear to generate knowledge autonomously, but without expertise guiding their use, they tend to produce merely plausible content rather than genuine insight. The real value of AI emerges when it is used by knowledgeable minds capable of directing, evaluating, and refining its outputs. I refer to this as epistemic competence. In this sense, AI does not replace expertise; it amplifies the importance of it.

The second principle, instrumentality, comes from the history of technology and media. We think through tools—especially epistemic technologies that extend our cognitive capacities, such as writing, printing, and computing. Human intellectual evolution has always depended on such tools, which allow us to offload tasks and open new possibilities for thought. AI belongs to this long trajectory of cognitive augmentation.

The cover of Generative Knowledge is shown, featuring a colourful splash of powder on a white background — Paolo Granata’s ***Generative Knowledge: Think, Learn, Create with AI*** (Wiley, 2026) explores how AI acts as a new “epistemic technology.”

The third principle is sociality, or collective intelligence. Knowledge is never purely individual; it is always socially produced. We know what we know because we participate in communities of exchange—scientific, cultural, and intellectual. Mechanisms like peer review, collaboration, and shared standards form an invisible college behind us to support knowledge creation. These first three principles reflect what the history of technology and knowledge already tells us: to create new knowledge, we need expertise, intellectual tools, and other minds.

The remaining principles—inquiry, learnability, and creativity—extend this foundation into practical applications for thinking, learning, and creating with AI.

SRI: As you were writing the book, did your thinking about creativity shift in any significant way?

Granata: One of my goals was to address a persistent misconception about creativity. When people hear the term, they often immediately think of artistic or aesthetic production. Creativity becomes narrowly associated with art, design, or cultural expression.

My aim was to shift the focus from artistic creativity to intellectual creativity. A philosopher or a computer scientist can be creative in the same fundamental sense: by creating the conditions for new knowledge. Research and learning are creative acts. Traditionally, creativity is defined in terms of novelty, originality, and value—but many things beyond artworks meet these criteria.

While it was not my initial intention to focus so explicitly on defining creativity, I realized during the writing process that this reframing was necessary, especially given how generative AI has made creativity itself a contested and consequential concept. Creativity as an intellectual practice cannot be taken for granted, because the opposite assumption—that creativity belongs primarily to the arts—is still widespread.

SRI: Many educators worry that AI diminishes student agency. How can a generative mindset support inquiry and critical thinking in learning environments?

Granata: What we see right now is a strong polarization. Some claim that AI systems exhibit genuine intelligence or emergent properties, while others dismiss them as nothing more than statistics and stochastic mimicry. Similar divides appear in education. Many are very skeptical about embracing the adoption of AI in education while others are enthusiastic, and there is still a big stigma on using it.

I experienced this directly when teaching a course on AI literacy last term. When I asked my students whether they used AI daily, no one raised their hand. When I pressed them, they admitted they were using it—but felt they could not say so openly. This stigma is deeply counterproductive. It prevents meaningful discussion about how AI is actually shaping learning practices. At the end of term, after sustained discussion about what AI literacy truly means and why it matters, I asked the question again and every hand went up.

Empirical research reflects this polarization. Some studies suggest AI use weakens critical thinking; others suggest it can enhance it. I think the outcome depends on the pedagogy we develop for the use of AI in education. With a lack of pedagogy and without a proper understanding of this technology, the implications will be very negative. But with a fully integrated pedagogy and an awareness of what AI means in terms of enhancement of our intellectual life, the outcomes will be different. We need to foster more awareness on how critically embracing AI in education can have positive implications on critical thinking and intellectual development, while being aware of the potential risks. And so, it's about learning how to learn—that is the responsibility we have as educators.

SRI: Which concept from the book do you think will have the most lasting influence, and why was it important to introduce new vocabulary now?

Granata: We need new language to think differently. The concept I emphasize is “epistemic wellness.” In recent years, concerns about digital media have often been framed through metaphors of pollution—misinformation, disinformation, and toxic online environments.

With generative AI, what I fear is not disinformation, but dis-understanding. AI changes not just what we know, but how we know. Epistemic wellness refers to our capacity to maintain healthy conditions for how we come to know things—conditions that support reflection, judgment, and intellectual growth.

If we don't take care of our epistemic wellness while speeding up our efficiency by using AI to automate tasks, instead of having more time to think and reflect, we are just going to get busier than ever. My hope is for a world in which AI helps us reclaim time for thinking, creativity, and meaningful inquiry. Epistemic wellness may sound like something new, but it's one of the many virtues in philosophy, aimed at fostering ways to help us advance knowledge instead of just surviving in an overwhelming society.

SRI: Looking ahead, how might generative knowledge reshape the role of public institutions in an AI-mediated world?

Granata: Public institutions have a crucial role to play. One priority is supporting open and publicly developed AI systems as alternatives to dominant proprietary models. Access to AI should be treated as a public good, much like access to healthcare or education.

Cultural and academic institutions, in particular, must focus on what they do best: providing context rather than content. Content can now be easily generated and distributed. Context—relationships, collaboration, shared inquiry—remains essential. Universities, libraries, and museums provide a kind of social glue by creating spaces where people can engage with processes of knowledge-making, not just finished products.

This speaks to the fact that AI is highlighting the difference between product and process. The product is easy to generate, and easy to see. We pay too much attention to the final product, but now the process is becoming more important. Where can people understand and experience the process? The answer is in institutions, which are the places where interaction, collaboration, and engagement will reveal how process is important, particularly in the knowledge economy. And so, I see these fundamental shifts from product to process, from content to context, and this is fundamentally the role that institutions should play to thrive in the new knowledge economy that AI is informing.