Blaise Agüera y Arcas and James Manyika argue that advances in artificial intelligence are prompting a fundamental reconsideration of what intelligence actually is. The authors identify five interconnected paradigm shifts: natural computing exists throughout the physical world, neural computing architectures can improve AI efficiency, intelligence fundamentally involves predictive modeling, general intelligence may already exist in current AI systems, and intelligence operates through collective cooperation across scales. They contend that intelligence does not necessarily require biologically based computation, suggesting humanity’s understanding of intelligence itself requires reconceptualization as AI capabilities expand.
