Understanding intelligence is one of the great scientific challenges of our time. But, in spite of extensive research efforts spanning many scientific disciplines, our understanding of intelligence remains fragmented and incomplete.
Science of Intelligence will fundamentally advance our understanding of intelligence, integrate and unify theories, concepts, and insights from existing intelligence-related disciplines, catalyze progress in these disciplines, and - perhaps most importantly - advance our ability to construct intelligent technological artifacts for applications of societal importance.
Science of Intelligence will identify the principles of intelligence. A key methodological strategy towards this goal is the synthetic approach to intelligence research. This approach requires that each insight, method, concept, and theory must demonstrate its merit by contributing to the intelligent behavior of a synthetic artifact, such as a robot or a computer program. These artifacts represent a shared “language” of all involved disciplines, enabling the validation, combination, transfer, and extension of research results. Insights about intelligence that are consistent across disciplines will help identify the underlying principles of intelligence. Furthermore, the synthetic approach facilitates the transfer of these principles into technologies and applications.
Science of Intelligence will implement scientific, structural, and educational measures to establish an integrated, multi-disciplinary research program on intelligence. It will unite intelligence researchers from the Berlin area in a unique environment for research and education. The integrated graduate school will train a new generation of intelligence researchers. The appeal of such an important scientific challenge and the unique combination of disciplines will also be leveraged to equalize the gender distribution within the cluster, especially in the engineering disciplines.
Today, intelligence research is fragmented and distributed across many disciplines, each producing theories and empirical findings about narrow aspects of intelligence. These theories and findings remain disconnected and sometimes inconsistent. Their synthesis into a unified understanding of intelligence represents a formidable scientific challenge.
Science of Intelligence (SCIoI) marks a turning point in the history of intelligence research. SCIoI develops a principled understanding of intelligence by bringing together synthetic and analytical disciplines (see figure above). The convergence of these disciplines is catalyzed by the construction of synthetic artifacts (robotic systems, simulations, computational models, etc.). These artifacts, situated along the “spokes” of the above figure, bridge the gap between synthetic and analytical disciplines. They translate insights into a shared synthetic language. This enables the validation, combination, transfer, and extension of research results across all disciplines, leading to the identification of principles of intelligence.
SCIoI is framed by a sociological and philosophical introspective research effort. The goal of this effort is to identify disciplinary differences and to propose methods to overcome them.
The resulting understanding of intelligence will be constructive: every insight about intelligence will be demonstrated and validated in a technological artifact. As Richard Feynman (1918-1988) put it: “What I cannot create, I do not understand.” This constructive understanding will translate into practical technologies, employable in a broad range of industries.
Read the white paper.