Heidi Fraser-Krauss argues that ChatGPT is not as bad—or as good—as you think
Many column inches have been written about the artificial intelligence tool ChatGPT (the GPT stands, not very snappily, for Generative Pretrained Transformer) since its launch at the end of November. Some of the uses to which it has been put are hugely entertaining, my personal favourite being a train cancellation announcement in the style of the Book of Revelation.
However, there has also been a big focus on the challenge the tool poses to education and research. It is easy to see ChatGPT as a threat. It can produce grammatically correct, well-structured and highly plausible pieces of content, in seconds. This means it has obvious potential to be used to produce essays or outline research proposals and academic papers with minimal input and effort.