New 'moonshot challenge' at Harvard aims to make giant leap in AI

The grant aims to help scientists figure out why mammalian brains are so good at learning, and then design better computers accordingly. Harvard researchers will record activity in the brain's visual cortex in "unprecedented detail," map its connections and then reverse-engineer the data to inspire better computer algorithms for learning.

“This is a moonshot challenge, akin to the Human Genome Project in scope,” said project leader David Cox, assistant professor of molecular and cellular biology and computer science at Harvard.

The grant was awarded by The Intelligence Advanced Research Projects Activity (IARPA), part of the U.S. government's Office of the Director of National Intelligence. 

Simply recording the activity of so many neurons and mapping their connections has "enormous" scientific value, Cox said, but that's only part of the project. "As we figure out the fundamental principles governing how the brain learns, it's not hard to imagine that we’ll eventually be able to design computer systems that can match, or even outperform, humans.”

Applications for such systems could include detecting network invasions, reading MRI images and driving cars.

To make it happen, researchers will first train rats to recognize various objects on a computer screen. Cox’s team will record the activity of the rats' visual neurons using laser microscopes built for the purpose with collaborators at Rockefeller University.

Next, the rats' brains will be studied physically using what Harvard says is the world’s first multi-beam scanning electron microscope in the university's Center for Brain Science.

The resulting petabyte or so of data will be analyzed to reconstruct cell boundaries, synapses and connections, and visualize them in three dimensions. Eventually, the researchers aim to build better algorithms for learning and pattern recognition.

“This project is not only pushing the boundaries of brain science, it is also pushing the boundaries of what is possible in computer science,” said Hanspeter Pfister, the An Wang Professor of Computer Science at Harvard. “We will reconstruct neural circuits at an unprecedented scale from petabytes of structural and functional data. This requires us to make new advances in data management, high-performance computing, computer vision and network analysis.”