To what extent can a synthetic system be rational?
A brand new MIT course, 6.S044/24.S00 (AI and Rationality) doesn’t try to answer this query. Instead, it challenges students to look at this and other philosophical issues through the lens of AI research. For the subsequent generation of scientists, concepts of rationality and agency could prove to be an integral a part of AI decision-making, especially as they’re influenced by how humans understand their very own cognitive limitations and their limited, subjective views of what’s rational and what shouldn’t be.
This inquiry is rooted in a deep relationship between computer science and philosophy, which have long worked together to formalize what it means to form rational beliefs, learn from experience, and make rational decisions in pursuit of 1's goals.
“You can imagine that computer science and philosophy are pretty far apart, but they’ve at all times overlapped. The technical parts of philosophy really overlap with AI, especially early AI,” says course instructor Leslie Kaelbling, Panasonic Professor of Computer Science and Engineering at MIT, recalling Alan Turing, who was each a pc scientist and a philosopher. Kaelbling himself has a bachelor's degree in philosophy from Stanford University and points out that computer science was not available as a serious on the time.
Brian Hedden, a professor within the Department of Linguistics and Philosophy and holds a joint appointment within the MIT Schwarzman College of Computing with the Department of Electrical Engineering and Computer Science (EECS), who co-teaches the category with Kaelbling, notes that the 2 disciplines are more aligned than one may think, adding that the “differences are in emphasis and perspective.”
Tools for further theoretical consideringG
First offered in the autumn of 2025, Kaelbling and Hedden developed AI and Rationality as a part of the Commonalities for computer science education, an overarching initiative of the MIT Schwarzman College of Computing that brings together multiple departments to develop and teach latest courses and launch latest programs that bridge computer science with other disciplines.
With over two dozen students enrolled, AI and Rationality is considered one of two Common Ground courses with a philosophical foundation, the opposite being 6.C40/24.C40 (Ethics of Computing).
While “Ethics of Computing” examines concerns concerning the societal impact of rapidly advancing technology, “AI and Rationality” examines the controversial definition of rationality through several components: the character of rational agency, the concept of a totally autonomous and intelligent agent, and the attribution of beliefs and desires to those systems.
Since the implementation of AI is amazingly comprehensive and every use case raises different problems, Kaelbling and Hedden brainstormed topics that would enable fruitful discussion and debate between the 2 perspectives of computer science and philosophy.
“When I work with students studying machine learning or robotics, it will be important that they take a step back and examine the assumptions they’re making,” says Kaelbling. “Thinking about things from a philosophical perspective helps people draw conclusions and higher understand learn how to place their work within the actual context.”
Both instructors emphasize that this shouldn’t be a course that gives concrete answers to the query of what it means to construct a rational agent.
Hedden says, “I view the course as constructing their foundation. We're not giving them a doctrine that they’ll learn, memorize after which apply. We're equipping them with tools to think critically about things as they pursue their chosen profession, whether in research, industry or government.”
The rapid progress of AI also presents science with latest challenges. Kaelbling believes it's an unattainable task to predict what students might want to know in five years. “What we’d like to do is give them the tools at the next level — the habits of mind, the ways of considering — that can help them approach the things that we actually can't predict straight away,” she says.
Mixing disciplines and difficult assumptions
So far, the course has attracted students from a big selection of disciplines – from those deeply knowledgeable about computer science to others serious about exploring how AI intersects with their very own areas of study.
Throughout the semester, students grappled with different definitions of rationality and the way they pushed back against assumptions of their fields.
What surprised her concerning the course, says Amanda Paredes Rioboo, a senior at EECS: “We're form of taught that mathematics and logic are this gold standard or truth. This course showed us loads of examples of individuals acting in contradiction to those mathematical and logical frameworks. We opened an entire can of worms to ask, is it people who find themselves irrational? Is it the machine learning systems that we designed which are irrational? Is it mathematics and logic itself?”
Junior Okoroafor, a graduate student within the Department of Brain and Cognitive Sciences, appreciated the challenges of the category and the way in which the definition of a rational agent can change depending on the discipline. “Presenting what each field understands by rationality in a proper framework makes clear which assumptions should be shared between the fields and that are different.”
The course's co-teaching and collaborative structure, as with all Common Ground efforts, gave students and college the chance to listen to diverse perspectives in real time.
For Paredes Rioboo, that is her third Common Ground course. She says: “I actually just like the interdisciplinary aspect. They at all times felt like a pleasant mixture of theory and practice as they need to work across disciplines.”
According to Okoroafor, Kaelbling and Hedden showed an obvious synergy between areas and said it felt like they were engaging and learning alongside the category. How computer science and philosophy may be used to tell one another allowed him to grasp their similarities and invaluable perspectives on overlapping topics.
He adds: “Philosophy may surprise you.”
