Torras.ai

Welcome to the Torras AI Podcast, where Jordi explores the intersection of artificial intelligence, technology, cybersecurity, and business innovation through insightful conversations with leading voices in the field. In this episode, Jordi speaks with Carlos, President of Asociación Educación Abierta and co-author of the book IA y Educación. Una relación con costuras, to discuss how AI is reshaping the landscape of education, the risks of overpromising technology's role, and the importance of preserving the human essence of learning in an increasingly automated world.

Jordi: Welcome to the Torras AI Podcast, the space where we speak with leaders and experts about artificial intelligence, technology, cybersecurity, and business innovation. Today I’m pleased to speak with Carlos, President of the Asociación Educación Abierta, consultant in educational innovation, and co-author of the book “IA y Educación. Una relación con costuras”. Carlos has spent years analyzing the impact of technology on education with a critical and reflective lens. He encourages us to approach AI in education with depth, context, and purpose. Carlos, it's great to have you with us.

Carlos: Thank you, Jordi. I’m happy to be here and share this space with you.

Career Path and Background

Jordi: Before we dive into the topic, tell us a bit about yourself. How did your professional journey unfold to become one of the most influential voices in the conversation about education and technology?

Carlos: My career has been, like many, somewhat unplanned. I started studying physics, specifically theoretical physics, but simultaneously pursued geography and history. That already showed my curiosity across fields. Initially, I worked in tourism, but about 25 years ago, I entered the education sector. I worked with the regional government in Madrid and have since been involved with universities, business schools, and independent projects. Over time, my focus has moved from practical work to more reflective analysis. I feel very comfortable where I am now.

On Past Promises of Educational Technology

Jordi: In your article “IA y Educación: ¿esta vez sí funcionará?”, you talk about how AI repeats promises of earlier educational technologies that failed. What should we do differently this time to avoid repeating the same mistakes?

Carlos: First of all, let’s make fewer promises. Education is deeply affected by socioeconomic and cultural factors. Every new technology is seen as a possible solution to complex challenges, but history shows us repeated cycles of overpromising and underdelivery. Whether it was eliminating bureaucracy or personalizing learning, many technologies promised a revolution and fell short. With AI, the same two promises reappear: personalized learning and time savings for teachers. While AI can assist in specific tasks, expecting it to transform education entirely is unrealistic. We need to approach it more modestly and critically.

Jordi: That reminds me of tech trends like object-oriented programming or Agile. Each came with grand promises, but the actual transformations were far less dramatic.

Carlos: Exactly. And overpromising followed by underuse is a recurring pattern in educational technology. Technologies that aim to solve deeply human and cultural challenges often ignore the complexity of implementation. Every solution introduces new challenges. That’s something we must acknowledge from the start.

Efficiency vs. Human Values in Education

Jordi: You’ve warned that AI reinforces a logic of efficiency and productivity that already dominates education. How can we push back against that trend and use AI to promote critical thinking, creativity, and collaboration?

Carlos: Technology is inherently geared toward efficiency. Over the last few decades, we’ve seen a technocratic rationalization of education—standardizing processes, setting learning objectives, measuring performance through indicators like PISA. While this helps manage complexity, it risks eroding the essence of education, which is inherently uncertain. Learning is unpredictable and social. Reducing it to measurable outputs diminishes the richness of the educational experience.

Jordi: It’s the difference between managing education like a business with KPIs and embracing its organic, sometimes messy, nature.

Carlos: Exactly. Too much control can stifle curiosity and exploration. Education should help students navigate uncertainty—not eliminate it. AI could either support this vision or reinforce a reductionist, overly structured model.

Personalization: Promise or Trap?

Jordi: You’ve also questioned the promise of personalized learning through AI. What risks do you see, especially in under-resourced school systems?

Carlos: Personalization has long been a marketing mantra. The risk is that it becomes just a more efficient path to the same standard outcomes—everyone reaching the same goals, just through slightly different routes. That’s not real personalization. True education should open possibilities, not narrow them. Another risk is individualization—isolating learners in algorithmic learning paths, which undermines the social aspect of education. In affluent contexts, students may get rich, interactive learning. In low-resource settings, we risk offering automated, one-size-fits-all instruction, which could deepen educational inequality.

Jordi: That’s a real concern. Those with more means always access better education, and now AI might further widen that gap.

Carlos: Exactly. Think about tech executives sending their kids to Waldorf schools—low-tech environments that emphasize creativity and interaction. Meanwhile, in less wealthy regions, governments might turn to low-cost automated systems that prioritize efficiency over richness and depth. That’s a dual-speed education system.

“Seams” in the Relationship Between AI and Education

Jordi: Let’s talk about your book IA y Educación. Una relación con costuras. What do you mean by “seams” in this relationship?

Carlos: The metaphor came naturally. Tech strives for seamless experiences—frictionless, invisible, smooth. And that’s great in many ways. But learning requires friction. It’s the struggle to understand, the effort to express, the discomfort of confronting the unknown. If AI eliminates all the seams, it might remove what's necessary for deep learning. There’s also another meaning: education reveals the seams of the world—it helps us see what’s hidden, question power structures, uncover what’s beneath the surface. That’s crucial in a time when AI can obscure or simplify complex realities.

Jordi: That makes a lot of sense. Even the idea of seamlessness is often an illusion. Nothing is truly seamless—there’s always work being done under the hood.

Carlos: Absolutely. In tech, engineers work hard to hide that complexity. But in education, making everything “invisible” can be dangerous. If AI becomes a black box, educators lose agency. They need to understand how algorithms work, not just use them. Otherwise, we risk reinforcing superficial learning and undermining the very purpose of education.

Back to Basics: Understanding AI and Its Foundations

Jordi: I recently told someone on TV that we shouldn’t teach kids “AI” per se—but rather the underlying concepts: vectors, matrices, trigonometry. AI isn’t magic; it’s math.

Carlos: Totally agree. Kids need to understand computational thinking, not just use AI tools blindly. The danger is that AI may reinforce algorithmic thinking—step-by-step processes without understanding. That’s not the goal. Education should focus on deep understanding, whether it’s math or history. For example, applying the quadratic formula is one thing; understanding how it was developed over centuries is quite another.

Jordi: I remember that formula being taught as a given, with no explanation of its history or derivation. But those concepts were the product of centuries of thought and progress.

Carlos: Exactly. Fortunately, modern pedagogy is moving toward that deeper understanding. Parents might be confused when their kids learn math differently, but it’s often because they’re being taught to grasp the “why,” not just the “how.” AI might reinforce old teaching methods we’ve been trying to evolve away from. We have to be cautious.

Final Thoughts and How to Connect

Jordi: Carlos, thank you so much for this insightful conversation. I know you’re busy, so just one last question: how can people connect with you or continue learning from your work?

Carlos: I’m easy to find online. People can reach me on LinkedIn, on X (formerly Twitter), or through my blog. Just searching my full name, Carlos Magro Mazo, will lead you to my work. I try to stay accessible and open to conversation.

Jordi: Thank you again, Carlos. It’s always a pleasure to talk with you. And to our audience, don’t miss our next episode—we’ve got another special guest coming soon. See you next time on the Torras AI Podcast!