In Pursuit of Smarter Machines: the Brain as Archetype
Despite the efforts of many researchers, machine capabilities still lag far behind the level and complexity of processing in the biological brain. To address this gap, the United States Government’s Intelligence Advanced Research Projects Activity (IARPA) has invested approximately $100 million in the MICrONS (Machine Intelligence from Cortical Networks) program. The overall goal is to revolutionize machine learning by reverse-engineering the algorithms of the brain, which involves multidisciplinary collaboration of a large group of scientists from multiple institutions. The approach combines neuroscience and data science to understand the cortical computations underlying neural information processing and ultimately to develop novel machine-learning algorithms inspired by the workings of the brain.
As part of a $21 million contract, Chris Xu, Applied and Engineering Physics, is working with collaborators at Baylor College of Medicine, California Institute of Technology, University of Toronto, University of Tuebingen, and Columbia and Rice Universities. Xu’s lab is creating and deploying technologies that have the unprecedented capability of mapping the brain functions of an entire cortical column of a behaving animal. The successful completion of this program will not only advance understandings of the brain but also provide significant opportunities to achieve major breakthroughs in machine learning.