Four years ago we wrote an article predicting the disrupting effect of large language models in the fields of chemical education and research. Here we review and grade our past predictions, give our perspective on some of the progress that has made in the intervening years, and finally give some forecasts of what might be coming next.