Updated: August 14, 2025
AI as "wisdom aggregators" makes a powerful tool for consensus-making and making the signaling of public information stronger
Lowering the cost of public signal aggregation may create repercussions for the information balance, making some markets less efficient, e.g., prediction markets
The long-term consequences of the signal aggregation mechanics of AI may be adverse for the diversity in the public information space and exacerbate the creation of echo-chambers of even smaller size than they are at the current moment
Artificial intelligence can be understood not only as a computational tool but also as a social opinion aggregator. By drawing on the vast repository of human-generated content available online, AI produces outputs that tend toward the statistical center of collective expression. In this sense, its role is less about originality and more about offering a synthesized signal that reflects the “average” of available opinions.
This perspective resonates with recent discussions in Harvard Business Review on AI’s potential for collective decision-making. If AI functions as an aggregator that regresses toward the mean, it can provide individuals with a valuable reference point: how their own views deviate from the collective norm. Yet aggregation alone does not equate to decision-making. Consensus requires social interaction, negotiation, and the reconciliation of differences—functions that belong to human deliberation rather than machine computation.
The parallel to Surowiecki’s wisdom of the crowd is striking. Both crowd wisdom and AI aggregation rely on the averaging of dispersed information into a collective signal. Prediction markets operationalize this principle by transforming individual beliefs into tradable forecasts. However, AI now encroaches on this territory by algorithmically generating an “aggregated signal” from human knowledge at scale. This raises an important question: which source of collective intelligence will prove more valuable—human-driven prediction markets or machine-processed aggregations of human input?
The dominance of AI as a mean signal generator may alter the balance of market efficiency. If average signals become ubiquitous and easily accessible, the relative value of private, non-average information increases. Informed agents holding unique insights could capture outsized returns, while the rising cost of producing or accessing such private signals may exceed the incentives for broader information sharing. This dynamic could concentrate informational advantages and undermine the participatory spirit of collective intelligence systems.
The consequences are far-reaching. First, AI developers and the owners of proprietary data inputs gain disproportionate power, as they control both the aggregation mechanism and access to underlying signals. Second, markets such as prediction platforms may be destabilized if participants defer excessively to AI-generated averages rather than contributing their independent judgments. Third, and perhaps more surprisingly, an overreliance on AI-mediated averages could erode the diversity of perspectives that underpins both crowd wisdom and democratic deliberation. Instead of fostering more informed decisions, AI’s “regression to the mean” risks creating an echo chamber of statistical centrality—precisely the opposite of the vibrant, plural knowledge ecosystems it draws from.
Grossman, S. J., & Stiglitz, J. E. (1980). On the impossibility of informationally efficient markets. The American Economic Review, 70(3), 393–408.
Harvard Business Review. (2025, August). How AI can help tackle collective decision-making. Harvard Business Review. https://hbr.org/2025/08/how-ai-can-help-tackle-collective-decision-making
Surowiecki, J. (2004). The wisdom of crowds: Why the many are smarter than the few and how collective wisdom shapes business, economies, societies, and nations. Doubleday. Colorado Mountain College
YouTube. (2025, August). A Cheeky Pint with Robinhood CEO Vlad Tenev [Video]. YouTube. https://youtu.be/_F8SfqaYeq4?si=Je1Sj97Wzzshn_ra
Please cite this article as:
Petryk, M. (2025, August 14). When Machines Compete with the Wisdom of Crowds. MariiaPetryk.com. https://www.mariiapetryk.com/blog/post-19