Teaching writing to children using artificial intelligence with GPT-3

Abstract

Teaching writing is a fundamental process in the development of language and social emotional skills. The advent of the Generative Pre-Train Transformer 3 (GPT-3) has sparked interest in how this new tool derived from artificial intelligence can be integrated into the teaching of writing, but there are also concerns about its impact on the creativity and individuality of Gen Z and Alpha. According to sociologist Helga Nowotny, we should adopt a techno-ethical approach to artificial intelligence as an answer to adapting to a digital society. GPT-3 can generate text from input, providing an opportunity to develop a dialogue between the child and the machine. It can also be used as a tool to support learning, problem solving and creative writing activities. Through the integration of the TPACK theoretical framework, a narrative action research methodology was applied to collect texts derived from writing between children and artificial intelligences, offering the possibility of opening pedagogy to the world of advanced language models.