AI Can Save the Public Sphere (www.persuasion.community)

The emergence of large language models (LLMs) like ChatGPT and Claude is reshaping the landscape of public discourse, potentially reversing the trend of populism seen in social media. Unlike social media, which has democratized voice access but often amplifies sensational and divisive content, LLMs tend to focus on delivering expert, evidence-based information. This shift towards "technocratization" suggests that LLMs could steer public opinion closer to expert consensus, improving the overall quality of political discourse and counteracting misinformation. Research indicates that engaging with LLMs can enhance belief in accurate information and diminish the influence of conspiracy theories. This development is significant for the AI/ML community as it illustrates the potential for LLMs to serve as credible intermediaries in complex information environments, fostering a more informed public. Their ability to systematically address nuanced questions without the biases or frustrations that human experts might exhibit further enhances their persuasive power. As LLMs become tools for fact-checking and content creation across digital platforms, they may significantly influence societal beliefs and political discussions, promoting a more rational and balanced dialogue in democratic societies. This evolution underscores the contrasting roles of technology—where social media has fragmented voices, LLMs might unify them around informed perspectives.