Experts, Crowds, and Algorithms: Who Really Knows Best?

In the age of data and artificial intelligence, the question arises: who makes better decisions, experts or machines? Michael Mauboussin’s Think Twice dives into this debate, showing that while experts offer valuable insights, they are often outperformed by crowds and algorithms in many domains.

Consider a company’s internal prediction market where hundreds of employees wager on future sales outcomes. This market achieved 99.5% accuracy, surpassing expert forecasts. Similarly, Netflix’s Cinematch algorithm revolutionized movie recommendations by analyzing vast data patterns, outperforming human clerks relying on intuition.

Why do crowds excel? Diversity of opinion, independent thinking, and effective aggregation reduce individual errors. Proper incentives motivate truthful input, making collective estimates surprisingly precise. However, crowds are not infallible; homogeneity and groupthink can degrade performance.

Experts shine in complex, strategic, or novel problems where intuition and experience matter. Yet, overreliance on experts can lead to blind spots and overconfidence. The best approach blends human judgment with computational tools, leveraging strengths of both.

Understanding these dynamics empowers leaders and individuals to harness the right decision-making sources at the right times, improving outcomes and innovation.

As technology advances, the interplay between human and machine intelligence will only grow more important. Embracing this synergy is key to thriving in a complex world.

2 , 3