Meta has recently partnered with Stanford's Deliberative Democracy Lab and the Behavioral Insights Team (BIT) to conduct a Community Forum on Generative AI. The forum, which included over 1,500 participants from Brazil, Germany, Spain, and the United States, aimed to gather public feedback on the principles that should guide the development and use of generative AI chatbots.
Community Forums bring together representative groups of people from around the world to discuss and debate important issues with each other and experts in the field. The recent Community Forum on Generative AI, designed by Stanford's Deliberative Democracy Lab and BIT, focused on two overarching questions:
1. What types of principles should be reflected in generative AI chatbots?
2. How should these chatbots interact with people and provide guidance and advice?
The Community Forum on Generative AI revealed several key takeaways, including:
1. A majority of participants from each country believe AI has had a positive impact, and this view increased through deliberations.
2. Participants support AI chatbots using past conversations to improve responses, as long as people are informed.
3. After deliberating, a majority of participants agreed that AI chatbots can be human-like, provided that people are informed.
Notably, participants' perspectives changed as a result of the deliberation process, sometimes by as much as 20.5% on certain questions, often moving toward consensus on pressing topics.
Meta's Community Forum on Generative AI, conducted in partnership with Stanford's Deliberative Democracy Lab and BIT, showcases an inclusive and informed approach to AI governance. Collaboration, debate and transparency are of utmost importance in shaping what we expect of an AI infused future.