The article deals with a study that shows how large artificial intelligence language models can influence political recommendations based on available sources, and highlights the risks that arise from delegating thinking to technology, especially in the context of elections in Japan. (translated)

