GPT-3 informs and disinforms us better

Compared to humans, artificial intelligence (AI) language models produce accurate tweets that are easier to understand and tweets containing disinformation that are harder to detect, according to a recent study. While these results can be leveraged to …