In the 21st century, a science studying the actions of very large groups of people might be developed. It was described in the fiction of Isaac Asimov, who called it Psychohistory.
Asimov described it with this analogy: in a gas, while the motion of a single molecule is very difficult to predict, the mass action of the gas can be predicted to a high level of accuracy.
Such science might arise from a combination of history, sociology, mathematical statistics, behavioural economics, psychology and other fields. An attempt to create psychohistory was Karl Marx's historical materialism.
If this science is be developed before the Technological singularity, it could help those who have access to it to directly influence course of humanity's future.
Accepting this possibility significantly complicates our ability to predict events before the Singularity, or to influence them. It's like trying to plot a strategy in a game of chess when there is a rule that one of the players becomes able to read the mind of his opponent once there are 20 pieces left on the board.
In Socio-tech... the predictive science of societal behavior by Yonas et al. in the original NBIC report socio-tech was defined as "the predictive — not descriptive — science of the behavior of societies. It is the convergence of information from the life sciences, the behavioral sciences (including psychology and the study of cognition), and the social sciences."
- Socio-tech... the Predictive Science of Societal Behavior, G. Yonas, J. Glicken Turnley. In Converging Technologies for Improving Human Performance (NBIC report)