Campus News in Brief

Carnegie Mellon endeavors to reverse engineer the brain for artificial intelligence while also seeking to figure out why people fail to claim their tax credit.

CMU endeavors to reverse-engineer the brain for research on artificial intelligence

Researchers at Carnegie Mellon have embarked on a five-year, $12 million research project to reverse-engineer the brain in order to better understand neural circuitry and learning methods. They hope to eventually apply their findings to the field of artificial intelligence. The project is led by Tai Sing Lee, a professor in the computer science department at Carnegie Mellon and the Center for the Neural Basis of Cognition (CNBC), a joint program between Carnegie Mellon University and the University of Pittsburgh.

The project is funded by the Intelligence Advanced Research Projects Activity (IARPA) through its Machine Intelligence from Cortical Networks (MICrONS) research program. MICrONS is part of President Barack Obama’s BRAIN Initiative which aims to revolutionize our understanding of the human brain.
“MICrONS is similar in design and scope to the Human Genome Project, which first sequenced and mapped all human genes,” Lee stated. “Its impact will likely be long-lasting and promises to be a game changer in neuroscience and artificial intelligence.”

In particular, researchers hope to discover the mechanisms the brain’s visual systems use to process information. This could potentially revolutionize machine learning algorithms and computer vision, while also improving the performance of neural networks, which are computational models for artificial intelligence inspired by the nervous systems in animals. Neural networks are used today in cases where computers can learn to recognize faces, understand speech and handwriting, and make decisions for self-driving cars.

Researchers also plan to collaborate with MICrONS teams at other universities to evaluate computational and learning models of the brain, and hope to eventually build better computer algorithms for learning and pattern recognition. “The hope is that this knowledge will lead to the development of a new generation of machine learning algorithms that will allow AI machines to learn without supervision and from a few examples, which are hallmarks of human intelligence,” Lee said.

CMU seeks to understand why millions of Americans fail to claim their tax credit

Around seven million Americans every year, 25 percent of those eligible, fail to claim the Earned Income Tax Credit (EITC), the primary channel of government support for the working poor. These credits are worth, on average, a month of income. Saurabh Bhargava, an assistant professor of economics at Carnegie Mellon, led a field experiment with the Internal Revenue Service (IRS) in order to find out why so many people don’t sign up for valuable government benefits and to identify ways to improve participation.

The study suggested that the reasons why people don’t claim credit is not due to stigma or time required to apply, as predicted by traditional economic theory, but rather the result of “psychological frictions” like low awareness, confusion regarding eligibility, and inattention to program information.
The study estimated that a redesign and expansion of the way notices are distributed could increase the number of claimants by several hundred thousand. Since the study’s completion, the IRS has circulated redesigned notification forms.

“This study demonstrates the value and need for rigorous evidence-based approaches to public policy. It also suggests that, in an increasingly complicated world, simplicity and a commonsense recognition of how individuals make decisions can play a critical role in the ultimate success of policies like the EITC,” Bhargava said.