Climate change in Poland – the assessment of the conversation with ChatGPT

Abstract

Abstract ChatGPT, along with its applications, possibilities, limitations and future development, is currently one of the most often discussed topics worldwide. One of the issues raised in those discussions is its ethically questionable role in science and education. The goal of this paper is to assess the accuracy and correctness of the responses given by ChatGPT, using climate change in Poland as an example. Eight questions related to this topic were posed to ChatGPT, and each answer was subsequently verified and assigned a grade on a scale of 0–10. The overall grade obtained was 3.8, indicating that only 30–40% of the information provided by ChatGPT was accurate. This poor result can be attributed to fake references, inaccurate data, overgeneralizations and simplification. Nevertheless, with proper training and development, ChatGPT has tremendous potential to serve as a valuable tool for ethically sound applications in the field of science.