Robots and the Liberal Arts

Last week, I read an article in the Chronicle of Higher Education with the provocative title “How Robots Will Save Liberal Education”. The author, Eboo Patel, is a Rhodes scholar, trained in sociology, and he served on Obama’s advisory council on faith-based neighborhood partnerships. The essay begins with a vignette of the author’s mother trying to persuade him as an undergraduate to major in business for job security reasons. She viewed his intended sociology major as being a luxury, but perhaps not very useful. Twenty years later, it’s clear that Patel has put his training to good use, in an entrepreneurial way no less.

Patel argues that the “hallmarks of a liberal education – building an ethical foundation that values the well-being of others, strengthening the mental muscles that allow you to acquire new knowledge quickly, and developing the skills to apply it effectively in rapidly shifting contexts – are not luxuries but necessities for preparing professionals for the coming transformation of knowledge work to relationship work.” Certainly these are all good outcomes of higher education, but it isn’t clear that they are provided only by a liberal education, and that business and engineering majors, would lack these skills without the appropriate classes, or in particular pedagogical method.

He speculates “that the 15-student seminar discussing Plato’s relevance to contemporary situations [could turn out] to be better preparation for the jobs of the future than working through problem sets alone for a science or engineering class.” He posits that the seminar requires attentiveness to diverse viewpoints, working on a synthesis of multiple viewpoints, constructing and communicating an argument, and iterate through this process to make stronger arguments. “All of the above happens in the space of a few minutes in an actual room with actual people. The problem set can be done in a split second by a computer.”

But careful thinking, iterative processing, and synthesizing information from multiple sources isn’t limited to the philosophy seminar (or other liberal arts courses in the humanities and social sciences). This is what scientists and engineers are trained to do in their education; they might even get more rigorous training. Now as a scientist in a liberal arts setting, I think our science majors potentially get the best of both worlds – while the paucity of science requirements for non-science majors does them a major disservice. While I understand that Patel may highlight a stereotypical distinction to make a point, his pedagogic argument is confused, perhaps because he has less teaching experience in the college classroom. Yes, having students discuss and make supporting arguments is good, but only if they have done the reading and thought about it on their own to some extent and wrestled with it. Only then is the dialectic in the classroom enhancing. And the purpose of a problem set isn’t to solve the problem with a known answer. It requires the students to wrestle with, think about, and provides them with the foundation to then do something else more complicated. It’s a crucial part of the learning process. My students also work in groups and we do have class-wide discussions, but part of the “alone” time working on the problem is key to learning – I’m sure this is true broadly across disciplines and not just in the sciences.

Where I do agree with Patel is that teaching and learning is best in a relational context. They should not be divorced to learning in a disembodied content. This is true not just for human learning, but also from observations of animal learning. (I recently finished Frank De Waal’s Are We Smart Enough to Know How Smart Animals are?) The social aspects may be even more important in the ape culture where technology is more limited, and the relationships are nuanced and complex. Technology allows us to rely less on human relationships to some extent, although it can also enhance relationships across space and time that would have been more difficult to do without writing, movable type, the computer and the internet.

Patel concludes that “robots may well perform medical operations and process our financial transactions in the not-too-distant future, but they are unlikely to replace pastors in pulpits, teachers in classrooms, nurses in hospitals, or coaches on the basketball court.” I’m not so sure, particularly with regard to the notion of place. I think there will still be pastors, teachers, nurses and coaches. But they may not exist in physical or localized pulpits, classrooms, hospitals or gyms. I agree with Patel that “people need interaction with other people to become better people”, and that liberal education, broadly construed, helps. But what if that’s not the goal of the people? The ideal society envisioned by Plato’s philosopher kings may not be shared by its democratic populace. Robots, part of the technological advance, raise particular questions about relationship and identity. Are we increasingly choosing to be Alone Together? And if so, perhaps liberal education is truly in jeopardy.