Researchers from Cornell University have conducted a study about the impact of artificial intelligence (AI) on online human conversations.
With the coronavirus forcing millions of people to practice social distancing, there has been a corresponding increase in online communications, be it for work or social life. As such, Cornell University believes that AI systems that supply “smart” reply suggestions in a text could help keep conversations from veering off track.
The study found that when humans are having difficult conversations they should trust AI systems. When things go wrong, AI can reportedly act as a so-called moral crumple zone when a human conversation becomes hostile. Like a car’s crumple zone, which deforms to absorb the impact of a crash, AI can buffer online conversations, detecting when things are going downhill and suggesting conflict resolution strategies to save the conversation.
The study explored the various ways that AI is altering how humans interact. Choosing an AI prompted reply may save typing time, but it can also alter the course of a conversation. The study sheds some light on how people perceive and interact with computer messages and researchers believe that AI could improve human communication with subtle guidance and reminders.
The study was published in Computers in Human Behavior.
