Page 163 - Izobraževanje v dobi generativne umetne inteligence

Generative AI and the Croatian Educational System

sector each year. This conference provides a platform for teachers to share in-
novative methods and experiences with integrating Generative AI and other
advanced technologies into their teaching practices.
The exchange of best practices, supported by initiatives like the CARNET
conference, is crucial for addressing the challenges posed by Generative AI.
Whether these systems function as black-box, grey-box, or transparent mod-
els (Adadi & Berrada, 2018), the emphasis remains on equipping educators
with the tools and knowledge to critically evaluate AI-generated outputs.
This empowers teachers to integrate GEN-AI in ways that enhance learning
without undermining the human elements of education, such as creativity,
empathy, and individualised support.
As Generative AI technologies continue to evolve, Croatia’s proactive ap-
proach demonstrates the value of shared policies, ethical guidelines, and
community collaboration. By fostering these principles and supporting ed-
ucation through pilot projects, programming curricula, and best practice ex-
changes, Croatia positions itself as a leader in integrating Generative AI into
education. Ultimately, this collective commitment ensures that Generative AI
serves as a valuable partner in education, enhancing rather than replacing
the uniquely human aspects of teaching and learning.

Pedagogical Approaches for Teaching With, About and For GEN-AI
The integration of Generative AI into education is not merely a technologi-
cal advancement but a fundamental shift that challenges traditional notions
of teaching and learning. To understand its impact, it is essential to connect
its use with established educational theories. The works of Jean Piaget, Lev
Vygotsky, Jerome Bruner, and Seymour Papert offer critical frameworks for
examining how students learn and how tools like Generative AI can either
support or hinder the learning process.
Jean Piaget’s constructivist theory (Waite-Stupiansky, 2022) emphasises
the active role of students in constructing their own knowledge through
experiences and interactions with their environment. Generative AI, while
capable of providing answers to questions, risks undermining this active pro-
cess if used solely as a source of direct answers without encouraging deep-
er exploration or critical thinking. Unlike a teacher or a peer, who can adapt
their responses to the learner’s developmental stage and cognitive needs,
Generative AI lacks the nuanced understanding required to scaffold learning
effectively in line with Piaget’s principles without prior user profiling.
Lev Vygotsky’s sociocultural theory, particularly his concept of the Zone
of Proximal Development (ZPD) (Doolittle, 1995; Verenikina, 2003), highlights

163