On September 23, Professor Viktor Mayer-Schönberger—known as the “Father of Big Data” and a professor of Internet Governance and Regulation at the Oxford Internet Institute, University of Oxford—addressed the 13th Thinkers’ Forum. He delivered a keynote speech titled “AI and Human Decision-Making,” providing in-depth analysis of the interaction logic between artificial intelligence and human decision-making in the context of big data.
At the beginning of the lecture, Schönberger posed three multiple-choice questions about global changes: the trend of extreme poverty, the vaccination rate of one-year-old children, and the global income distribution. He used these questions to test the audience’s cognition of the world—and the results showed that most people gave wrong answers due to information bias. This interaction vividly supported his core view: decision-making is the key “steering wheel” for individual and social development. Small decisions, such as daily action choices, and major ones, such as decisions about academic pursuits, marriage, and consumption, all shape an individual’s development track; group decisions, in turn, have a bearing on social development. The prerequisite for high-quality decisions is sufficient information—insufficient information often leads to inaccurate human decisions. Information is like “glasses,” helping people understand and perceive reality more clearly.
When talking about data utilization, Schönberger revealed a “secret”: over 80% of global data and more than 85% of European data are only collected but never used. This inefficiency stems from human inertia—people tend to make decisions with minimal data or even no data at all, and this creates an opportunity for AI to step in. Schönberger cited cases such as AlphaGo defeating Go champions, AlphaFold predicting protein folding, and Waymo self-driving vehicles operating efficiently to demonstrate AI’s disruptive capabilities in decision-making. At its core, AI is a data-driven model system. Through ultimate utilization of massive data, it overcomes many limitations of human decision-making.
Schönberger emphasized that AI is not omnipotent, and its value depends on application scenarios. “There is no simple, one-size-fits-all answer to whether AI is good or bad. The answer lies in specific contexts—it depends on whether the decision-making challenge you face is suitable for AI modeling.” In the era of big data, humans have access to massive data resources, but they also face the task of optimizing data utilization and decision-making. AI, he noted, is an important tool to help humans unlock the value of data and improve decision-making.
“The future of decision-making is a symphony of humans and machines united by data and insights, moving together toward a more rational and creative destination,” Schönberger concluded. Looking ahead, humans need to build on “information sufficiency,” use AI tools to unleash the potential of data, and at the same time deepen their irreplaceable cognitive advantages. This is not only an inevitable outcome of technological evolution but also a sign of human wisdom in proactively embracing change. Amid the surging tide of AI development, AI tools will surely assist humans in opening up a new realm that combines efficiency and compassion.
