Artificial intelligence has long been part of the everyday lives of children and young people – for example, in the form of chatbots, learning aids or digital assistants. Many of these systems communicate in a way that mimics human characteristics. Younger users in particular therefore tend to attribute feelings, intentions or even a consciousness of their own to them.

The EU Kids Online 2020 study shows that children often interpret digital offerings in social terms and may struggle to understand how they actually work. Even though the focus is not exclusively on AI, one thing is clear: the younger the users are, the more likely they are to ‘humanise’ technologies. This presents parents and teachers with the task of contextualising this perception and imparting a basic understanding of digital systems.

For despite its outward appearance, artificial intelligence is not a ‘thinking’ system. It is based on mathematical processes that analyse large amounts of data, recognise patterns and make predictions on this basis. Models such as neural networks process inputs via weightings and probabilities – they react to data without actually understanding the content or having intentions of their own.

This is precisely where the fischertechnik STEM Coding Ultimate AI school set comes in. It enables pupils to understand how AI works in a practical way, rather than just discussing it theoretically. Pupils can train a simple neural network themselves and observe how its behaviour changes in response to data.

The system combines physical models with digital components: sensors collect data, a controller processes it, and actuators visibly implement the results. The individual steps – from data collection through training to application – remain transparent and traceable. AI is thus not experienced as a ‘black box’, but as a structured process.

The kit offers several advantages for teaching: abstract concepts from computer science and mathematics become tangible, and key skills such as algorithmic thinking, problem-solving and critical thinking are fostered. At the same time, pupils learn to recognise the limitations of AI and to contextualise its results.

Another important aspect is local processing: the models are trained on a computer or tablet and then transferred to the system where they are executed – without a cloud connection. This ensures that data handling remains traceable and controllable – an issue that is also becoming increasingly important in an educational context.

Overall, the system helps teachers make AI understandable and tangible. Pupils learn not only how such systems work, but also how to evaluate them realistically – as powerful tools, not as human counterparts.