Recent advancements in artificial intelligence (AI) may have the potential to improve public discourse, which has been significantly impacted by social media's fragmentation of reality. Various AI applications, including ChatGPT, Claude, and Microsoft Copilot, are now accessible on smartphones, showcasing a range of capabilities that could help unify perspectives.
Historically, technological developments have diminished the authority of experts and diversified public debates. In the mid-20th century, the landscape was dominated by a few television networks like ABC, NBC, and CBS, which limited news coverage and often relied on a narrow set of official sources. This created a media environment fostering trust in mainstream institutions but also facilitated the government’s manipulation of information.
Despite the risks associated with AI, such as potentially worsening public discourse, there is evidence that large language models (LLMs) can promote a shared understanding of reality. They have demonstrated success in dissuading users from adopting false beliefs, presenting a contrasting incentive structure compared to social media companies.