AI Literacy

There is increasing talk of, the ability to understand AI in some broader context and to work critically with it. For a relatively long time, the topic of AI has been a particular science discipline or a science fiction genre in literature and film. However, the development of individual sub-projects is gradually showing that working with them is essential for personal use, competitiveness, entertainment, science, security education and creativity. Our aim in this area will not - given the scope and objectives of the text - be to offer any complete presentation of the topic, but rather to build on what we emphasised in the introduction: AI literacy is a critical element in the ability to work with AI in a genuinely creative way.

Ng et al. stress that it is essential that students not only know how to use AI tools but should be able to reflect on ethical issues and approach them responsibly. At the same time, they stress the importance of understanding at least the basic principles underlying these tools. About specific aspects of literacy, they speak of four pillars:

• Knowledge and understanding of tools
• Using and applying tools
• Evaluation of tools and design of new procedures
• Ability to reflect on ethical issues

Two security examples illustrate that knowledge of more profound principles is essential. The first is the call to regulate the TikTok network, which uses artificial intelligence to deliver relevant content. The way it does this is with an algorithm so aggressive that it can lead to depressive feelings or addictive behaviour. At the same time, it collects behavioural data, which, in the hands of Chinese intelligence services, can be an essential source of data on the behaviour of people in Western society and their sensitivity to specific messages. Barbie.me operates as a service where the user uploads personal information and a photograph and receives a picture of themselves in the form of Barbie or Ken. In reality, however, this is a severe security threat, where the user pays for the AI-generated image with well-structured personal data to a completely unknown company that can misuse it individually or as a data set.

Wijer and Hakim show that AI literacy is strongly domain-sensitive. In their case, namely for the health disciplines, it is much more directed towards data handling, visualisation or the ability to find a place for technology in the decision-making process of providing quality treatment. Yang argues that AI literacy should be developed in preschoolers. Dawy et al. discuss that one way to understand how AI works may be to use it in creative writing. Students are encouraged in the creative process and get a clear idea of how to generate outputs or the limits of the chosen tools.