¶ Demand 5: AI Education
Those who do not understand what an algorithm is cannot judge what an algorithm does to them.
¶ Core Demand
AI literacy -- critical thinking about algorithms, data protection and AI-generated content -- must be mandatory from secondary school. ETH Zurich and EPFL are to become national competence centres for AI education.
¶ The Education Gap
¶ The Generation of Digital Illiterates
Swiss pupils use AI tools daily -- ChatGPT for essays, Midjourney for images, GitHub Copilot for code. But they do not understand how these systems work. They cannot assess whether an AI-generated answer is correct, why an algorithm displays certain content, or which of their data is being used in the process [1].
This is like teaching a generation to drive without explaining that cars need brakes.
¶ PISA and Switzerland
The 2022 PISA study shows: Swiss 15-year-olds perform well in mathematics and natural sciences, but digital competence -- especially the critical evaluation of digital information -- remains below average [2]. In a world where AI generates billions of texts, images and videos every day, the ability to distinguish fact from fiction is a basic skill like reading and arithmetic.
¶ What AI Education Must Include
¶ Lower Secondary Level (from age 12)
- How does an algorithm work? Basic concepts: input, processing, output. Sorting algorithms as an introduction.
- What are training data? Why AI systems inherit the biases of their data.
- Fake or real? Recognising AI-generated texts, images and deepfakes.
- My data belongs to me: Data protection, consent, right to erasure.
¶ Upper Secondary / Gymnasium (from age 16)
- Understanding machine learning: Supervised, unsupervised, reinforcement learning -- without a mathematics degree, but with comprehension.
- Ethics of AI: Bias, fairness, transparency. Case studies: facial recognition and racism, credit scoring and discrimination.
- AI and society: Labour market, democracy, surveillance. Debate format.
- Practical: Training simple AI models of one's own (e.g. with Google's Teachable Machine).
¶ Universities
- ETH Zurich and EPFL as national competence centres for AI education: teaching materials, continuing education for teachers, research on AI didactics.
- Compulsory module "AI and Society" in all degree programmes -- not just in computer science.
¶ Role Models
| Country | Measure | Since |
|---|---|---|
| Finland | "Elements of AI" -- free online course, target: 1% of the population | 2018 |
| Estonia | Programming from year 1, AI strategy in the curriculum | 2019 |
| Singapore | AI for Everyone / AI for Industry -- national continuing education | 2019 |
Switzerland introduced the Lehrplan 21 competence area "Media and Computer Science" [3]. This is a start, but it covers AI-specific topics only marginally.
¶ ETH AI Center and EPFL AI Center
The ETH AI Center, opened in October 2020 in Zurich-Oerlikon, already brings together AI research across departmental boundaries [4]. The EPFL AI Center does the same in Romandy. Both institutions could assume a leading role in AI education -- not only in research but also in teacher training and the development of teaching materials for all school levels.
¶ Sources
[1] OECD, Education at a Glance 2024: Digital Skills and AI Literacy.
[2] OECD, PISA 2022 Results, Vol. I: Switzerland country report.
[3] D-EDK, Lehrplan 21, Competence area "Media and Computer Science".