PISA - School Is a Dead Horse

Lesedauer 7 Minuten

Prologue: PISA (Program for International Student Assessment), or more specifically the “PISA studies,” have been conducted annually by the OECD since 2000. The focus is on the competency areas of reading, mathematics, and science. Each year, supplementary competencies such as creative thinking (PISA 2022) or foreign languages (PISA 2025) are assessed.

This year (2026), however, the PISA results for Germany are causing quite a stir, as they are alarming. Everyone is writing about PISA—how we should respond, where the problem lies, and what exactly needs to be done, by whom, when, and where. Back in 2012, in my first book, *Liquid Work*, I had already criticized the state of education at the time and even presented an alternative concept. Last year, I published my AI-based discussion on intergenerational equity in education. Thanks to the PISA results, both topics are more relevant than ever.

In my next article, I want to address a topic that hardly any of the reactions to PISA described above take into account— technological singularity.

This article is dedicated to my grandson Beute.

I was born in 1966 and just turned 60. My wife’s grandchild—my “step-grandchild”—was born in December 2024. By the time he’s as old as I am today, it will be around the year 2084. What will the world look like then? The honest answer is: We don’t have the faintest idea.

The past 60 years alone have fundamentally changed the world we live in. In 1966, there was no Internet, no smartphones, and no personal computers; the moon landing was still to come; genetic engineering and digitalization played no role in everyday life; and my future dream car rolled off the assembly line.

Today we’re discussing artificial intelligence, gene editing, quantum computers, and technologies whose societal impacts we can barely fathom even as they’re being developed. Meanwhile, what has changed in the way school works, aside from the fact that Germany is steadily slipping in the PISA rankings?

When we think about the “future,” we often make a mental mistake: We view the future as if progress would continue at roughly the same pace we’ve seen in the past. But that’s exactly what doesn’t happen. Technological development is generally not linear. Advances enable further advances; tools improve the tools that, in turn, are used to develop the next generation of technologies. This raises a question for me, now that I’m 60, that seems somewhat abstract at first:

If we view the technological changes of the past 60 years as a single entity—how much progress could take place in the next 60 years?

If we assume, as a highly simplified conceptual model (e.g., based on Moore’s Law or similar curves), that technological progress doubles roughly every ten years, then six doublings already result in a factor of 64.

But of course, “progress” is not a physical unit that can be measured precisely. Artificial intelligence, biotechnology, energy supply, medicine, communication, and social change do not develop at the same rate or independently of one another. The calculation is therefore not a valid forecast, but a valid thought experiment.

When does 60 years become one month?

So let’s turn the question around. In an exponential growth model, when could approximately as much change occur in a single month as between 1965 and 2025?

If we make the simplified assumption that the number doubles every ten years, we end up roughly at the end of this century. In the model used here, that point would be around 2099, which means the past 60 years would have been compressed into a single month.

If not only does the rate of development accelerate, but doubling times also become shorter, this point in time shifts significantly forward. This is referred to as “superexponentiality.”

Ray Kurzweil links this idea to his theory of a technological singularity: a point at which technological change becomes so rapid that humans can barely keep up with it using their current ways of thinking and adapting. He assumes that the time it takes for technology to double will itself be cut in half. If so, we will reach that point much sooner.

If, for example, the doubling time is halved every 20 years: • Then this tipping point may already lie between 2045 and 2070.

2045— 60 years of development that I have followed will then unfold in less than a week.

2045— My “prey” grandson will be 20 by then. He won’t even be able to begin to anticipate his own future.

2045— The point at which technological progress is so rapid that we humans can no longer keep up with it.

What does this mean for PISA and education?

Is school a dead horse?

Our educational system dates back to a time when knowledge remained relatively stable for long periods and was continually expanded. Curricula could be developed, books written, content taught, and exams administered years later based on that knowledge. Today, however, we are already experiencing something different.

Children and adolescents navigate digital interfaces with ease, using smartphones, apps, and, increasingly, AI systems. At the same time, we observe that skills such as sustained reading, the ability to comprehend complex texts with concentration, or independent research do not automatically develop alongside this technical proficiency. Furthermore, skills are not simply disappearing but are changing, and new skills are emerging—regardless of whether they are important for our species or not. For example, knowing how to use a smartphone is by no means the same as having digital literacy.

AI amplifies this effect dramatically. Anyone who simply asks ChatGPT a question receives an answer within seconds. It’s convenient and incredibly powerful. At the same time, there’s a risk that we’ll lose a skill that has been an essential part of learning until now: figuring out for oneself where knowledge comes from. Discovering and validating—skills that were not taught even in traditional classrooms over the past decades—are essentially deteriorating further, because even the path by which knowledge is acquired no longer matters. This is reflected particularly clearly in the “PISA” results.

So which skills should we absolutely not outsource to AI, because otherwise we’ll eventually lose the ability to evaluate what it’s giving us?

PISA – Which skills are (once again) becoming relevant?

Ability to: be patient, conduct research, verify information, etc.

For me, research is clearly a key part of it. But research alone won’t be enough. The ability to verify information will be at least as important. We are heading toward a time when it will likely become increasingly difficult to distinguish between truth, manipulation, deliberately created misinformation, and content produced to look deceptively real. Texts, images, voices, and videos can already be artificially generated today in such a way that they are no longer recognizable. What will therefore be crucial is a skill that, to me, can be described in a simple sequence:

Research → Analysis → Validation → Formulation

First, I need to be able to find information (not just one piece!) on my own. Then I need to understand what I’ve found, recognize connections, and weigh different statements against one another. Afterward, I need to be able to verify whether the information actually holds up. Only then can I form my own conclusion, which I can also phrase in a way that ensures it is understood.

If we hand over the research to AI and then let it handle the analysis, evaluation, and formulation as well, we skip precisely the mental process through which judgment is formed. We may end up with a perfectly formulated answer without being able to determine for ourselves whether it is correct.

The Revival of Intuition

For me, there is one more skill that has hardly been given any attention in our current educational system: intuition.

Anyone who spends a significant amount of time conducting their own research, analyzing, and validating information does more than just accumulate knowledge. They gain experience. With this experience comes an increasingly keen sense of when something doesn’t add up, when an argument has a flaw, or when a piece of information sounds convincing but is nonetheless inconsistent. For me, therefore, intuition is not a substitute for research or critical thinking; rather, it arises precisely from the critical integration of perception, experience, and knowledge.

In schools and vocational training, this form of perception has so far been largely subordinated to rational thinking. The focus was on absorbing knowledge, reproducing it accurately, and, based on that, providing a clear explanation of how one arrived at a conclusion—like painting by numbers. This is also reflected in the structure of the PISA assessment. There was hardly any room for one’s own intuition about whether something made sense.

In a future marked by ever-increasing volumes of information and ever-more-complex interrelationships, it is precisely this connection that could prove decisive: a trained mind capable of researching, analyzing, and validating, and an equally trained intuition that senses when something is amiss despite all its apparent plausibility.

Humans themselves could become the bottleneck

When I wrote my first book, *Liquid Work – Work 3.0*, in 2012, I explored, among other things, the Kondratieff cycles.

Even back then, I believed that after the great cycle of information technology, the decisive bottleneck would be neither data nor energy nor traditional healthcare. The bottleneck could become human beings themselves—or, more precisely, their psyche.

In theory, technologies can develop at an ever-faster pace. However, the human nervous system, our perception, our social structures, and our ability to psychologically adapt to change follow different timeframes and developmental scales—see evolution. “Habituation” is often mistakenly viewed as a kind of “high-speed evolution,” which can lead to dire consequences: burnout, depression, and even suicide. Habituation is nothing more than the body’s emergency response—an attempt to cope with change. It’s like an emergency generator that kicks in automatically. But just as such a generator isn’t designed for long-term use, neither is habituation. We can get used to noise, sleep deprivation, or even an abusive partner… but the harm doesn’t go away; it just becomes less noticeable.

PISA—or rather: “Quo vadis, SCHOOL?”

So how much sense does school still make, as we know it? Are we still beating a dead horse? I’ve already come up with an idea of my own, but the purpose of this article isn’t to present that solution—that’s not what this is about at all. I want to show that we’re currently feeling the strong gust of wind left behind by the future as it swept past the education system years ago. In order to even begin to raise children like my grandson Beute to become people capable of surviving in the next 20 years, we must let go of everything we currently associate with school, because the clock is ticking, and the survival of an entire species is at stake.

We can’t possibly give a child born today the right instruction manual for the year 2045. What we can give them is the ability to think for themselves, to feel, to regulate their own responses, to research, to examine, to perceive, and to arrive at their own assessment—to stand by it, but to question it again and again. It somehow reminds me of what science was actually supposed to be… but is no longer.

No matter what our planet looks like in 20 years—highly technologized or back in tune with nature, or both intelligently integrated—my great-grandchild should be able to cope with it.

For me, intergenerational justice begins right there, no matter what PISA says.