Conversational #AI and equity through assessing GPT-3’s communication with diverse social groups on contentious topics

source: https://www.nature.com/articles/s41598-024-51969-w

Our algorithm auditing #study focuses on two topics that we let our participants converse with the GPT-3 #chatbot: climate change and Black Lives Matter (BLM). The first topic represents a classic controversial science issue across the globe. Research has shown how public perceptions toward it vary across populations. Particularly, there exists a minority group that holds skeptical and denial attitudes toward the reality and human-induced nature of climate change, and this group is often challenging to persuade. By studying the topic of climate change, we have an excellent opportunity to examine how GPT-3 responds to this group compared to the opinion majority. Additionally, we can investigate whether social learning and attitudinal change occur post-chat. The second topic represents a highly-charged social issue that has received extensive media coverage and public attention over the past decade. We acknowledge that the word dialogue encompasses rich meanings especially when it comes to human’s dialogues with and via technologies. This paper uses the word #dialogue in the most broad sense and do not evaluate the genuineness, vulnerability and listening qualities of the dialogue.

#science #research #society #BlackLivesMatter #climate #technology #chatgpt

2