We are in the midst of a major shift in the way we think about complex systems like brains and computers. Until recently, the dominant models in cognitive science and artificial intelligence were based on the symbol processing. Researchers in a wide spectrum of disciplines are now exploring a new paradigm for computation based on distributed rather than localized data structures, and information flow in dynamical systems rather than through logical operation.
The impact of this new approach to computation has influenced research directions in many areas of science. In neuroscience, modeling techniques are changing the way that we think about how populations of neurons represent information. In cognitive science, parallel distributed processing models, pioneered at UCSD, have revolutionized concepts of memory, attention and language. In computer engineering, VLSI chips and optical devices are being created that process information in parallel, heralding a new generation of computers that have vast computational power dedicated to problems in vision, speech recognition, and motor control.
One of the most remarkable aspects of these developments is the range of scientific disciplines that are involved. In addition to neuroscientists, cognitive scientists, and computer scientists, some of the leading contributors have been physicists, mathematicians, electrical engineers, linguists, economists and philosophers. Understanding how brains function, at the neural and cognitive levels, is one of the greatest remaining scientific mysteries.
INC's research efforts are rapidly progressing in the direction of understanding the modes of functioning of nervous systems through direct observation, experimental investigation and modeling of neural structures; uncovering cognitive principles through psychological experimentation and parallel distributed processing models; applying neural computation to the solution of major technological and scientific problems; and ultimately building a new generation of massively-parallel computers based on the principles of neural computation.
The central premise of the INC is that these diverse research efforts cannot be adequately achieved independently; instead significant progress will come through the joint efforts of researchers in the relevant disciplines, including neuroscience, philosophy, psychology, cognitive science, physics, mathematics, economics, electrical and computer engineering, computer science and engineering, and linguistics. INC's unique blend of these disciplines is what makes this Institute a leading international center in neural computation.
