The Human Element: Why Top Forecasters Still Have an Edge (For Now)

In the growing debate over AI versus human intelligence, a recent forecasting competition offered a nuanced verdict: AI is catching up fast, but the best humans still have a critical edge. While British AI ManticAI scored an impressive eighth place, the top spots in the Metaculus Cup were still occupied by elite human “superforecasters.”
This result is supported by research from Philip Tetlock, the co-author of “Superforecasting,” whose work has found that expert humans, on average, still outperform the top-performing bots. The key question is why. The answer seems to lie in the uniquely human ability to handle ambiguity, sparse data, and complex logical chains.
According to Warren Hatch, CEO of a company co-founded by Tetlock, humans retain the edge in “categories with sparse data that require more judgment.” An AI, which thrives on vast datasets, can struggle when information is limited or contradictory. A human expert can draw on years of experience and intuition to make an educated guess, a skill machines have yet to master.
Furthermore, humans excel at spotting logical inconsistencies. As Metaculus CEO Deger Turan noted, AIs can sometimes produce forecasts that are not internally coherent, a flaw that a human analyst is more likely to catch. This ability to perform a final “sanity check” is an invaluable part of the forecasting process.
However, the “for now” is a crucial qualifier. AI is improving at an exponential rate. The consensus is not that humans will reign supreme forever, but that the most powerful forecasting will come from a human-AI partnership. This hybrid model will leverage the strengths of both, combining human judgment with the AI’s tireless data analysis to achieve what neither could alone.