A group can solve what individuals cannot, if structured to aggregate knowledge rather than suppress it.

Individuals have knowledge gaps. They reason from partial information, carry blind spots, and miscalculate under uncertainty. Yet properly structured groups routinely outperform even the best individual experts on prediction and problem-solving tasks. This is not mystical crowd appeal but a measurable phenomenon rooted in how distributed knowledge combines.

Philosopher and theorist Pierre Lévy coined the term “collective intelligence” to describe a situation where nobody knows everything but everyone knows something, and what any given member knows is accessible to others on request. He argued that networked culture creates new power structures stemming from people’s ability to pool knowledge, collaborate in research, debate interpretations, and collectively refine their understanding. This is intelligence not as computation but as networked access and synthesis.

James Surowiecki’s “The Wisdom of Crowds” (2004) grounded this in empirical study. Surowiecki showed that when groups are diverse, independent, decentralized, and use a mechanism to aggregate individual judgments, their collective decision is more accurate than expert judgment alone. The four conditions matter: diversity ensures different perspectives; independence prevents groupthink; decentralization keeps local knowledge from concentrating in one person; aggregation ensures no single voice dominates the result.

Later empirical work by Woolley and others identified a general collective intelligence factor (“c”) that predicts a group’s performance on diverse tasks, distinct from the intelligence of individual members. Factors that raise collective intelligence include higher average member theory-of-mind (ability to model others’ thinking), more equal communication patterns, and higher proportion of women (correlated with social sensitivity).

Collective intelligence is neither a wish nor a guarantee. It requires structure. Hierarchy, censorship, and concentrated authority degrade it. So does false consensus or homogeneity masquerading as agreement. When the architecture enables genuine synthesis of distributed knowledge, groups solve what isolation cannot.