Charlotte Blease at Dr Bot:
The fourth estate voices mounting unease over chatbot-induced harms, including for people with mental health conditions. Despite this growing angst, we still know little about who – or what – on balance, better helps patients. There is still limited research on this issue. In this week’s Substack I examine some of the findings on humans versus chatbots when it comes to therapy.
We’ve seen this paper before but it is worth returning to. Alastair Howcroft and colleagues reviewed 15 studies comparing the empathy of AI chatbots with that of healthcare professionals. In 13 studies, AI responses were rated as significantly more empathic. When the researchers pooled the results, ChatGPT-3.5 and GPT-4 had a substantial advantage – roughly equivalent to scoring two points higher on a ten-point empathy scale.
Across multiple studies, AI-generated healthcare responses consistently sounded more caring and understanding than those written by human professionals. Of course, the fact that these were mainly text-based responses is a limiting factor.
More here.
Enjoying the content on 3QD? Help keep us going by donating now.