Masataka Watanabe
Masataka Watanabe trained in theoretical and experimental neuroscience, with research stays at Caltech and the Max Planck Institute for Biological Cybernetics before joining the University of Tokyo, where he now leads a lab studying the neural mechanisms of consciousness. His 2022 book, From Biological to Artificial Consciousness, an English expansion of a 2017 Japanese original, sets out two linked proposals, a neural algorithm theory of what consciousness is and a concrete experimental test for whether a machine has it.
The theory holds that consciousness is not a general byproduct of complex information processing but the product of a specific kind of neural algorithm, one Watanabe associates with chaotic, temporally fluctuating activity that propagates globally through a network rather than staying confined to a local circuit. On this view, replicating consciousness in an artificial system means replicating that particular dynamical signature, not simply matching a brain’s input-output behavior. The claim puts him at odds with purely functionalist accounts that treat any sufficiently capable computation as a candidate, and closer to the view that consciousness depends on specific causal or dynamical properties, a position that runs alongside Integrated Information Theory’s causal-structure requirement even though Watanabe’s algorithm is not phi.
Watanabe’s proposed test, developed from Roger Sperry’s split-brain research, asks whether a person could ever directly experience a machine’s output as part of their own subjective field. His proposal is to connect a candidate artificial neural system to one hemisphere of a human brain through a brain-machine interface modeled on the corpus callosum, then ask whether the person reports a unified stream of experience spanning both the biological and artificial sides. If they do, Watanabe argues, that is the only kind of evidence that could settle the question, because every external, third-person test is vulnerable to the philosophical zombie problem, a system that passes every behavioral check while having no experience at all. The same proposal extends toward what he calls seamless mind-uploading, gradually rehoming a living person’s conscious experience into a neural network while they remain alive, a route intended to avoid the philosophical problems that come with copying a mind only after death.
Known for. The neural algorithm theory of consciousness and a split-brain test for machine consciousness using a brain-machine interface
Coverage on this site
Every researcher covered on this site is indexed in the directory of consciousness researchers, grouped by the position their published work defends. The positions themselves are set out in the index of consciousness theories.