The machine wrote it better. People still wanted the human

Here is a result that sounds like a contradiction and isn’t. A team led by Penn State researchers, publishing in Communications Psychology, found that people rated empathetic responses written by AI as higher in quality than ones written by humans. Asked to choose, they still preferred to receive empathy from a human, even when it was worse¹.

Doctoral student Joshua Wenger calls it the “AI empathy choice paradox”. The machine’s replies made people feel more heard, cared for and validated, but people “still prefer to receive empathy — even lower quality empathy — from humans”¹. In one study, AI replies also beat responses written earlier by trained crisis responders on empathic quality most of the time¹.

The interesting part is effort. Empathy, the researchers argue, signals something because it costs the person who gives it. When I spend my emotional resources on you, I am telling you that you matter¹. For a machine, empathy costs nothing. Oddly, participants still rated the AI replies as more effortful, and replies that looked more effortful were rated as more empathetic¹. As human effort rose, the gap between human and AI empathy narrowed¹. People who use AI often were not noticeably more willing to choose AI empathy¹.

The researchers do not say the machine replaces the friend. Associate professor C. Daryl Cameron says people may accept less polished empathy “if it’s coming from someone I know really well”¹. Wenger suggests that AI’s best use may be when no human is available, not as a rival when one is¹. Cameron’s own example is modest: when you are too tired to write another condolence message, a model can help you start¹.

That is the sensible way to read it. Use the tool for the draft if you need to. Then send the message yourself, in your own name, and maybe a little less polished. The effort is part of what the other person receives.