What Survives When AI Comes for Design

Several Scandinavian educators in Michael Moore’s documentary Where to Invade Next (2015) stress that “School is about finding your happiness” and a math teacher even says “I teach them to be happy”. As an educator and happiness expert who tries to share various happiness tools to the world, I must give my hats off the Finland’s education system which relies on autonomy, independent and critical thinking, social interaction, well-being and play, part of my 8Ps happiness framework.

Then, I meet Lefteris Heretakis, European educator, podcaster and progressive thinker who does not shy away from new or old philosophies or ways of thinking as long as they work and are acted upon! And guess what, he confirms that play (8Ps), curiosity (7Cs), creativity (7Cs) and balance of flexibility and grounded solid basis that can always be challenged (13Cs) are some of the keys to elevated education he proposes and daily puts in action with his growing network of peers.

There is a question that design educators have been quietly avoiding for years. Now, with the arrival of generative AI tools that can produce polished visual work in seconds, they can no longer afford to. The question is a simple one: what, exactly, are we here for?

Lefteris has been sitting with this question for a long time. A practitioner and educator with years of experience teaching design across different countries and cultures, he founded the New Art School not as a reaction to AI, but as a response to a set of problems he had been watching accumulate long before the technology arrived. When we spoke on the AWE podcast recently ( https://youtube.com/live/rKRlqvWaUAc?feature=share ), his central argument was both urgent and clarifying: AI is not creating a crisis in design education. It is revealing one that was already there.

"We are confusing output with formation," he said. "Schools are measuring the wrong thing. They're optimising for the measurement." I also know that schools everywhere (except Finland and a few) face pressures to show student enrolment, grades on standardized tests that measure a structure that does not serve learning at its core.

Lefteris’critique is familiar in its bones but striking in its precision. Design schools, particularly those absorbed into universities over recent decades, have increasingly organised themselves around the production of portfolios, the achievement of measurable outcomes, and the appearance of competence. A student who can produce something polished, something that looks the part, is deemed to have succeeded. I have sat in design school interview process and thankfully found it extended to context and relational aspects which took almost a full day of interview but that is the rarity as we all know.

The AI problem is simple: the machine now does polished work faster, cheaper, and without complaining. If the portfolio was the proof of education, and the machine can generate portfolios, then what exactly are students paying for?

Heretakis is not interested in dystopian framing. He speaks instead of selection pressure. The institutions that survive, he argues, will be the ones capable of answering an honest question: what do you offer that a subscription to a generative AI tool cannot? For schools that have quietly reduced design education to a course in software, the answer is uncomfortable. For those willing to go deeper, the question is in fact an invitation.

And can AI be happy? Are students happy? Perhaps the creativity (7Cs) that most curriculum stifles is still the key. Passion (8Ps), craft (13Cs), contribution (13Cs) and purpose (8Ps).

The problems Heretakis identifies did not begin with AI. They began, he suggests, considerably earlier. The Prussian education system of the 19th century, designed in the aftermath of Napoleon's defeat to produce soldiers who followed orders rather than thought for themselves, established an architecture of schooling that has proved remarkably durable. Its priorities were compliance, measurability, and the suppression of individual judgement. Its legacy, for better or worse, is the classroom most of us recognise: the teacher at the front, the student at the bench, the grade as the verdict on a person's worth.

John Taylor Gatto, the American educator who taught in schools serving both the wealthiest and the poorest students in New York, documented this inheritance with unusual clarity (13Cs). He noticed that in schools for wealthy children, homework might be to arrange a meeting with the mayor. In schools for poor children, it was to memorise multiplication tables. The discrepancy pointed to something deeper than inequality of resource. It pointed to inequality of expectation about what a young person's mind was for.

Gatto's conclusion, arrived at after decades in the system, was that conventional schooling was not accidentally producing passive, uncurious, compliance-oriented graduates. It was doing so by design.

The question Heretakis adds to this is: what happens when design education inherits the same logic? What happens when the art school, historically a space for exactly the kind of unruly, exploratory, process-driven thinking that Prussian education was designed to eliminate, finds itself absorbed into a university structure optimised for metrics, audits, and demonstrable outcomes?

"When the art schools lost their autonomy," he says, "they had to comply to certain structures that suit other disciplines, but really do not suit art and design education."

The result, as he sees it, is a generation of educators who have never faced a real client, never navigated the complexity of bringing something into existence in a commercial environment, and who are teaching not from inside an experience but around it. The transmission breaks. The student receives a description of swimming rather than the thing itself.

Embodied education is an essential part of Dewey’s education philosophy as well as indigenous relational education which I both studied (and embodied) with pleasure.

There is, running beneath all of this, a question about roots. Heretakis returns to it repeatedly, and it is one of the most useful images in the conversation. Strong roots, flexible branches. A designer educated at depth, in the history of culture, architecture, music, philosophy and proportion, has something to stand on when the ground shifts. A designer educated in the features of a particular software package, or in the visual language of a particular moment, has nothing to stand on at all when the software updates or the moment passes.

The problem, he observes, is that students educated in a shallow system have no way of knowing what they have missed. They do not have anything to compare it to. They receive what the school offers and assume it is sufficient, until they encounter the world and discover that it is not.

This is compounded by the temporal pressure under which contemporary design education operates. Heretakis is direct on the point: to properly train a designer takes something closer to seven years than two or three. Architecture has maintained this understanding; it is almost alone in doing so. Most design degrees compress the process into a period that produces, at best, a promising beginning. The graduates who understand this are the ones who treat their degree as the start of a longer education rather than its conclusion. The others, as Heretakis puts it, have about ten years before they need to fundamentally re-evaluate who they are and what they are doing.

This ten year cycle is something, he notes, that very few educators tell their students.

Friedrich Fröbel, the German educator who developed the concept of the kindergarten in the 19th century, understood something that much of contemporary schooling has forgotten: that play (8Ps) is not the opposite of learning, it is one of its primary forms. Rudolf Steiner, working in the early 20th century, built on this insight to argue that education must address the whole human being. His formulation of head, heart and hand, of thinking, feeling and willing, described a kind of formation that no single measurable output can capture.

I also had a recent podcast discussing play and play therapy ( https://youtube.com/live/7wFdYs2Ntz0?feature=share )

The Waldorf principle that children up to the age of seven should be immersed in art, music, story, movement and practical making before being introduced to abstract academic subjects is not sentimentality. It reflects a developmental understanding of how human minds grow into their capacity for complex thought. Introduce abstraction too early, Steiner argued, and you do not accelerate learning. You produce, instead, a lasting experience of inadequacy. The child who cannot yet grasp what they are being asked to grasp concludes not that the material is premature, but that they are not intelligent. The wound is one that many carry throughout their lives, into the design studio and out of it. It starts to design their life.

Heretakis is emphatic about this. The problems he encounters in higher education design students are not primarily problems of talent or even of effort. They are problems of early formation. We both agreed that students who were never given space to be creative conclude that they are not creative. Students who were taught that every question has one correct answer, held by the teacher, are unable to inhabit the space of design, art or creative thinking, which proceeds on the assumption that there are many possible answers, not only one as the standard education often suggests, overtly or covertly, and that the task is to develop a methodology for navigating among them instead of choosing one.

In life, too we can live in uncertainty and ambiguity to find high happiness after the initial discomfort. Happiness and creativity can be taught and practiced.

"That," he says of this incapacity to be in the unknown, "is a successful application of schooling. That's exactly what schooling is designed to do."

The neuroscience here is not incidental. The default mode network, associated with daydreaming, internal narrative and the kind of associative thinking that connects apparently unrelated ideas, works in productive tension with the central executive network, associated with focused, directed thought. A third system, the salience network, mediates between them. Research published by Beaty and colleagues (2016) has shown that creative individuals demonstrate greater connectivity between all three. This is very important as these networks normally run in opposition or antagonistic relationships, but creativity is the sparks that makes them collaborate (13Cs)! Yes, creativity makes your brain collaborate!

You cannot be maximally creative by staying only in the focused mode, only in the daydreaming mode, or only in the mode of detecting what matters. You need the movement between them.

This movement is what unstructured play, field trips, music, making, and genuine exploration provide. It is also what a screen-dominated, output-optimised, metrics-driven education systematically prevents. The irony is that the environments most hostile to creativity are producing graduates who will need creativity most, in a world where the machine can handle the rest. Almos

Curiosity, Heretakis and I both agree, is the one thing AI genuinely cannot replicate. Not because AI lacks the ability to generate novel outputs, perhaps even better and more eclectic than most humans as research suggests (Bellemare-Pepin et al., 2024), but because it lacks the intrinsic drive to look for something in the absence of a prompt. Curiosity is generative. It asks questions that were not asked of it. It follows threads that nobody told it to follow. In a learner, it is not merely useful. It is the engine of everything else.

AI can be more “creative” but never curious!

The analogue and the digital present Heretakis with one of his most practical concerns. Students born into a digital world, he observes, often lack the experiential foundation from which digital design draws its grammar. The digital is a simulation of the analogue. To design well in digital space, you need to have experienced analogue space with some depth, to have handled materials, navigated physical proportion, understood how objects behave in the world. Students who have not had this experience are, in his account, genuinely baffled by certain fundamental design challenges, not because they lack intelligence but because they lack the embodied reference points that give design decisions their meaning.

The solution is not to reject digital tools but to insist, deliberately and sometimes forcefully, on analogue experience alongside them. The field trip to the design museum, the act of drawing by hand, the handling of physical objects, these are not nostalgic indulgences. They are the roots without which the branches cannot grow.

Paul Rand, the graphic designer whose work shaped the visual identity of some of the most recognised institutions in the world, put the underlying principle with characteristic precision. Heretakis quotes him: "Self-expression is real only after the means to it have been acquired."

This is not a conservative argument against experimentation. It is an argument for the sequence in which experimentation becomes possible. The student who has not yet acquired the means to self-expression is not expressing themselves freely. They are parroting, in the dark, without knowing it.

What, then, survives when AI comes for design? Heretakis's answer is not defensive. It is, if anything, clarifying. What survives is what always mattered and what too many institutions quietly abandoned in the name of measurability and efficiency.

It is the transmission of genuine practitioner experience, the kind of teaching that comes from inside the work rather than from a description of it. It is the cultivation of process over product, of methodology over output, of the capacity to trace one's own thinking and return to the point where it went wrong. It is the development of cultural awareness deep enough to understand proportion, harmony, the golden section, and the history of visual thought. It is the formation of curiosity, of resilience, of the confidence to produce and fail and produce again.

It is, in the deepest sense, the formation of a person rather than the production of a portfolio.

The institutions that can do this honestly will have no difficulty answering the question that selection pressure is now asking of them. The institutions that cannot, will find considerable restraint, that students will rightly stop attending, suggests Heretakis. I may be slightly more cynical and call it an ideal world. Still, this is a call to action and we are both men of actions, so let’s build this world!

For educators willing to hear it, this is not a crisis. It is an invitation to do the work they always knew mattered.

Heretakis's answer to schools still embedded in the old system draws on Sam Conniff's formulation from Be More Pirate: begin with small acts of resistance. You do not need to bring the structure down to change what happens inside it. You begin with one thing you can adjust, one conversation you can have differently, one assignment that prioritises process over product. Many small changes, made by enough people, create the conditions for a larger change.

And then, crucially, you stop researching and start doing.

"We have the solutions," he says. "We just need to apply them."

This, perhaps, is the most important thing that AI cannot do for us. It cannot apply the solution. It cannot take the risk. It cannot stand in front of a class of students from different cultures and different backgrounds and adjust, in real time, to who is there in the flesh, the context of the world and not only of a chat, database or web search. It cannot transmit, through the quality of its own engagement and attention, what it “feels” like to do the work. AI’s tokens are indeed tokens compared with the richness of life! We are, we participate with curiosity and camaraderie.

That remains, for now, irreducibly human. The question is whether we will use it.

Reference:

Beaty, R. E., Benedek, M., Silvia, P. J., & Schacter, D. L. (2016). Creative Cognition and Brain Network Dynamics. Trends in Cognitive Sciences, 20(2), 87-95.

Bellemare-Pepin, A., Lespinasse, F., Thölke, P., Harel, Y., Mathewson, K., Olson, J. A., ... & Jerbi, K. (2024). Divergent creativity in humans and large language models. arXiv preprint arXiv:2405.13012.