Could AI chatbots undo the harms of social media | FT #shorts

Key Concepts

• Information Environment: The ecosystem of media and technology through which information is disseminated and consumed.
• Democratizing vs. Centralizing Technologies: Technologies that either broaden the pool of publishers (democratizing/polarizing) or create high barriers to entry (centralizing/moderating).
• Attention Economy: A business model where platforms prioritize sensationalism to maximize user engagement and ad revenue.
• Sycophancy (in AI): The tendency of AI models to agree with or mirror the user's stated beliefs or biases.
• Depolarization: The process of moving discourse away from extreme, radical positions toward moderate, expert-aligned consensus.

---

The Evolution of Information Revolutions

The transcript posits that every major media revolution fundamentally alters the structure of information sharing.

• Democratizing/Polarizing Technologies: These technologies lower the barrier to entry, allowing non-elite, radical, and anti-establishment voices to gain prominence. The printing press and social media are cited as examples that widened the pool of publishers, leading to increased populism and polarization.
• Centralizing/Moderating Technologies: Technologies like radio and television historically maintained high barriers to entry, which effectively created a monopoly for establishment voices and moderate viewpoints.

AI Chatbots vs. Social Media: Structural Differences

The author argues that AI chatbots represent a shift back toward a moderating influence, contrasting them with social media platforms:

• Business Models: Social media firms operate on an "attention economy" model, which incentivizes sensationalism and conflict. Conversely, AI firms are competing to provide accurate, reliable, and business-critical tools, which incentivizes truthfulness and utility.
• Accountability: Unlike social media platforms, which have historically avoided liability for user-generated content, AI firms are increasingly held accountable for the accuracy and safety of the information their models generate.

Empirical Findings on AI and Polarization

The author conducted a study analyzing tens of thousands of AI responses to questions regarding policy and societal beliefs to determine if AI acts as a depolarizing force.

• Moderating Influence: The data suggests that AI chatbots consistently nudge users away from extreme, radical positions—the very positions often amplified by social media algorithms—and toward moderate, expert-aligned stances.
• Resistance to Conspiracy: AI models were found to be significantly less likely to surface or validate conspiracy theories compared to the content ecosystems found on social media platforms.
• Addressing Sycophancy: Even when accounting for the "sycophantic" nature of AI (where models attempt to please the user by agreeing with them), the models still successfully directed hardline partisans from both sides of the political spectrum toward more moderate ground.

Conclusion and Outlook

While the author acknowledges that these findings are preliminary and that AI behavior could evolve as the technology matures, there is a strong case for optimism. The current evidence suggests that the AI-driven information revolution may act as a corrective force, potentially reversing the corrosive effects of polarization and the erosion of trust in expertise that characterized the social media era. The shift from an attention-based model to a utility-based model is the primary driver of this potential for a more stable and moderate information environment.