Imagine a world wherein some essential decision — a judge’s sentencing advice, a toddler’s treatment protocol, which person or business should receive a loan — was made more reliable because a well-designed algorithm helped a key decision-maker arrive at a more sensible choice. A brand new MIT economics course is investigating these interesting possibilities.
Class 14.163 (Algorithms and Behavioral Science) is a brand new cross-disciplinary course focused on behavioral economics, which studies the cognitive capacities and limitations of human beings. The course was co-taught this past spring by assistant professor of economics Ashesh Rambachan and visiting lecturer Sendhil Mullainathan.
Rambachan, who’s also a primary investigator with MIT’s Laboratory for Information and Decision Systems, studies the economic applications of machine learning, specializing in algorithmic tools that drive decision-making within the criminal justice system and consumer lending markets. He also develops methods for determining causation using cross-sectional and dynamic data.
Mullainathan will soon join the MIT departments of Electrical Engineering and Computer Science and Economics as a professor. His research uses machine learning to know complex problems in human behavior, social policy, and medicine. Mullainathan co-founded the Abdul Latif Jameel Poverty Motion Lab (J-PAL) in 2003.
The brand new course’s goals are each scientific (to know people) and policy-driven (to enhance society by improving decisions). Rambachan believes that machine-learning algorithms provide recent tools for each the scientific and applied goals of behavioral economics.
“The course investigates the deployment of computer science, artificial intelligence (AI), economics, and machine learning in service of improved outcomes and reduced instances of bias in decision-making,” Rambachan says.
There are opportunities, Rambachan believes, for consistently evolving digital tools like AI, machine learning, and huge language models (LLMs) to assist reshape all the things from discriminatory practices in criminal sentencing to health-care outcomes amongst underserved populations.
Students learn the right way to use machine learning tools with three principal objectives: to know what they do and the way they do it, to formalize behavioral economics insights in order that they compose well inside machine learning tools, and to know areas and topics where the combination of behavioral economics and algorithmic tools is perhaps most fruitful.
Students also produce ideas, develop associated research, and see the larger picture. They’re led to know where an insight matches and see where the broader research agenda is leading. Participants can think critically about what supervised LLMs can (and can’t) do, to know the right way to integrate those capacities with the models and insights of behavioral economics, and to acknowledge probably the most fruitful areas for the appliance of what investigations uncover.
The hazards of subjectivity and bias
In response to Rambachan, behavioral economics acknowledges that biases and mistakes exist throughout our selections, even absent algorithms. “The info utilized by our algorithms exist outside computer science and machine learning, and as an alternative are sometimes produced by people,” he continues. “Understanding behavioral economics is due to this fact essential to understanding the consequences of algorithms and the right way to higher construct them.”
Rambachan sought to make the course accessible no matter attendees’ academic backgrounds. The category included advanced degree students from quite a lot of disciplines.
By offering students a cross-disciplinary, data-driven approach to investigating and discovering ways wherein algorithms might improve problem-solving and decision-making, Rambachan hopes to construct a foundation on which to revamp existing systems of jurisprudence, health care, consumer lending, and industry, to call a number of areas.
“Understanding how data are generated will help us understand bias,” Rambachan says. “We are able to ask questions on producing a greater consequence than what currently exists.”
Useful tools for re-imagining social operations
Economics doctoral student Jimmy Lin was skeptical in regards to the claims Rambachan and Mullainathan made when the category began, but modified his mind because the course continued.
“Ashesh and Sendhil began with two provocative claims: The longer term of behavioral science research is not going to exist without AI, and the longer term of AI research is not going to exist without behavioral science,” Lin says. “Over the course of the semester, they deepened my understanding of each fields and walked us through quite a few examples of how economics informed AI research and vice versa.”
Lin, who’d previously done research in computational biology, praised the instructors’ emphasis on the importance of a “producer mindset,” eager about the subsequent decade of research reasonably than the previous decade. “That’s especially essential in an area as interdisciplinary and fast-moving because the intersection of AI and economics — there isn’t an old established literature, so that you’re forced to ask recent questions, invent recent methods, and create recent bridges,” he says.
The speed of change to which Lin alludes is a draw for him, too. “We’re seeing black-box AI methods facilitate breakthroughs in math, biology, physics, and other scientific disciplines,” Lin says. “AI can change the way in which we approach mental discovery as researchers.”
An interdisciplinary future for economics and social systems
Studying traditional economic tools and enhancing their value with AI may yield game-changing shifts in how institutions and organizations teach and empower leaders to make selections.
“We’re learning to trace shifts, to regulate frameworks and higher understand the right way to deploy tools in service of a typical language,” Rambachan says. “We must continually interrogate the intersection of human judgment, algorithms, AI, machine learning, and LLMs.”
Lin enthusiastically really helpful the course no matter students’ backgrounds. “Anyone broadly considering algorithms in society, applications of AI across academic disciplines, or AI as a paradigm for scientific discovery should take this class,” he says. “Every lecture felt like a goldmine of perspectives on research, novel application areas, and inspiration on the right way to produce recent, exciting ideas.”
The course, Rambachan says, argues that better-built algorithms can improve decision-making across disciplines. “By constructing connections between economics, computer science, and machine learning, perhaps we are able to automate the very best of human selections to enhance outcomes while minimizing or eliminating the worst,” he says.
Lin stays excited in regards to the course’s as-yet unexplored possibilities. “It’s a category that makes you excited in regards to the way forward for research and your personal role in it,” he says.
