Talking to AI chatbots helped climate change sceptics trust the science more, a study found

Whether you love it or hate it, there's no denying that the AI revolution is well upon us. And while the ongoing flurry of controversial deep fakes continues to divide the public against the revolutionary tech, in another surprising turn of events, AI chatbots have emerged as a rather unexpected ally for another supremely pressing issue: climate change.

A new study has revealed a surprising phenomenon: interacting with AI chatbots like GPT-3, a precursor to OpenAI's ChatGPT platform, can subtly influence people's views on contentious issues like climate change, even if the conversation leaves them feeling dissatisfied.

The research focused on how GPT-3 handled discussions about climate change with users holding diverse viewpoints. While most participants reported similar levels of satisfaction with the chatbot experience, regardless of their background, a noteworthy pattern emerged among those initially sceptical of the scientific consensus on climate change.

Despite feeling less happy with their GPT-3 interactions compared to others, these climate sceptics showed a measurable shift in their attitudes towards climate change. Those initially least supportive of the facts surrounding human-caused climate change moved 6% closer to the supportive side of the scale after their chat. And all this happened within the average conversation span of just eight back-and-forths!

This unexpected outcome suggests that GPT-3, despite its limitations in catering to diverse viewpoints, may be inadvertently stimulating critical thinking and prompting users to re-evaluate their pre-existing beliefs. The researchers attribute this to the chatbot's varied response styles. When faced with climate change scepticism, GPT-3 offered more justification and evidence to counter the user's arguments.

Lead author Professor Kaiping Chen emphasises that the goal is not necessarily to make users happy, but to encourage learning and understanding, even if it involves challenging existing perspectives. In addition to climate change, a similar trend was also observed for those critical of the Black Lives Matter movement in the West.

“I won't say they began to entirely acknowledge human-caused climate change or suddenly they support Black Lives Matter, but when we repeated our survey questions about those topics after their very short conversations, there was a significant change: more positive attitudes toward the majority opinions on climate change or BLM,” Chen remarks.

With barely any time left to realise our climate targets, it is critical that the public acts in tandem with our institutions to help mitigate the looming climate crisis. While studies have shown that most Indians believe in the phenomenon, climate change still hangs as a major point of political controversy in the West, with many Americans still reluctant to believe in the validity of the science.

Thus, the study's findings hold significant implications for the future of AI-powered dialogue. By fostering critical thinking and prompting users to engage with evidence, chatbots could potentially play a role in bridging the divide on sensitive topics like climate change.

The findings of this research have been published in Scientific Reports.