notesthoughts

Anti-Elitism in the Age of AI

Why do people cheer when AI shakes the authority of experts? I've been thinking about the hope of democratizing knowledge, resentment toward intellectual elites, and where that authority might go next.

Recently, OpenAI released more than 700 AI-generated drafts of mathematics papers, causing a stir in the mathematics community. Some mathematicians expressed alarm at the possibility that AI could solve problems they had spent years studying. Others raised concerns about what this means for research and how the results should be verified.

What I found more interesting, though, was how people reacted to it, rather than the technical achievement itself. Instead of sympathizing with the unease mathematicians felt, some seemed to take satisfaction in watching their authority shaken.

Fields built on specialized knowledge, including software development, mathematics, and science, have long had high barriers to entry. Those barriers helped ensure expertise, but they also gave experts a particular kind of social authority. As AI begins to perform intellectual work that once required advanced expertise, the hierarchies built on that expertise are starting to shift.

This episode made me think about how people’s excitement does not necessarily come from having gained greater access to knowledge. They may understand mathematics no better than before and have played no part in a new discovery, yet still feel a kind of liberation at the sight of experts losing their authority. This seems to express both a hope for the democratization of knowledge and resentment toward established intellectual elites.

But there is a paradox here. The weakening of established experts’ authority does not mean that the power of knowledge will be shared by everyone. That authority could instead move to the small number of companies that develop and control AI. We may question the judgment of human experts while accepting AI’s results without adequately understanding or verifying them ourselves.

Perhaps AI is changing what expertise means and where authority resides, rather than eliminating expertise. And the anti-elite sentiment emerging in that process reveals how much we have respected knowledge itself, or how sensitive we have been to the social status granted to those who possess it.

Are people excited because AI is surpassing the limits of human intellect, or because people once considered more capable than themselves are no longer so special?