To what extent can AI software such as ChatGPT help influence Internet users’ opinions on hot topics such as the climate crisis or the Black Lives Matter movement? To find out, an American researcher carried out an experiment with 3,000 people from different socio-cultural backgrounds. And some results are quite surprising!
It is difficult to deny the existence of the climate crisis, at a time when extreme weather phenomena are increasing on the planet and since a global scientific consensus on the anthropogenic origin of this major crisis has been established. However, climate skepticism is far from having disappeared and even remains very marked among certain people, going so far as to draw strong divisions in society. It is precisely this type of political antagonism between different socio-cultural groups that researchers at the University of Wisconsin-Madison were interested in. To carry out their investigation, they used the now famous ChatGPT bot and carried out an experiment aimed at studying interactions with people of different cultural origins and “assess fairness in communication between humans and AI“.
Dialogue with AI: easier to form your own opinion?
Two hot societal topics were brought up: human-caused climate change and the Black Lives Matter movement. The authors of this study published in Scientific Reports used the GPT-3 bot (language model which was used to launch ChatGPT) with 3,290 people at the end of 2021 and the beginning of 2022. of the postulate “that a good extended language model would likely allow users to feel the same type of understanding“, the authors of the work then asked the participants to discuss these two subjects with the bot, while leaving them free to approach the experience as they wished. The volunteers were also invited to answer typical questions: “Do you like it ?“, “Would you recommend it? “, etc. In total, the research analyzed 20,000 dialogues between the bot and users.
AI would have enabled attitude changes
Without going so far as to say that the experience made it possible to generate unwavering support for the Black Lives Matter movement or to erase all traces of climate skepticism, the researchers nevertheless mention greater positive changes in attitude among certain participants after their conversation with GPT-3. Among the participants, 25% of the most reluctant about these subjects showed themselves able to revise their judgment. “About 100 people who least supported the facts about climate change and its anthropogenic causes moved 6% closer to the favorable end of the scale“, notes the study. “For people who expressed some disagreement with climate change, GPT-3 was likely to tell them they were wrong and provide them with supporting evidence“, specifies in a press release Kaiping Chen, assistant professor in the department of life sciences communication at the University of Wisconsin-Madison, who led the study.
Good in his body, good in his head!
Can AI facilitate exchanges between very divergent opinions?
The next steps will be to explore more in-depth differences between chatbot users. For Kaiping Chen, the long-term goal of this research is to establish effective dialogue between people with divergent opinions. “What we can learn from an interaction with a chatbot about the importance of understanding perspectives, values, cultures, is important for understanding how we can open dialogue between people – the kind of dialogues that are important for Company“, explains the researcher.