A study of 34 years of online discussions from Usenet to YouTube shows that, when it comes to rude behaviour, people — not platforms — are the root of incivility. Researchers used Google’s artificial-intelligence (AI) ‘toxicity classifier’ to identify “rude, disrespectful or unreasonable” comments. They found that over three decades, longer discussions tend to be more toxic, but heated debates don’t necessarily escalate or drive away participants. “Despite changes in social media and social norms over time, certain human behaviours persist, including toxicity,” says data scientist and co-author Walter Quattrociocchi.
https://www.nature.com/articles/s41586-024-07229-y