Columbia University Test-Drives Sway, an AI Debate Tool

Is there a role for Artificial Intelligence in facilitating conversations on delicate topics such as abortion, racism, immigration, or the Israel-Palestine conflict? Columbia University is exploring this possibility. The institution has recently initiated trials of Sway, an AI-driven debate tool that is currently being refined. The brainchild of two researchers at Carnegie Mellon University, Sway introduces students with differing perspectives to engage in one-to-one conversations about contentious matters. The tool’s official site describes it as a medium for enhancing interactions.

Sway has already drawn participation from approximately 3,000 students across more than 30 institutions of higher education. News about the possible collaboration with Sway emerges after over two years of growing disputes at Columbia involving students, university management, and federal entities. For some years now, Columbia has been the epicenter of numerous controversies, ranging from expulsion of pro-Palestine student activists to police interventions and federal government mandates.

Sway is currently under trial by members of Columbia’s Teachers College, as a part of their quest to incorporate it into their conflict resolution courses and community integration efforts. There are other groups within Columbia that have expressed interest in utilizing Sway in forthcoming semesters. An ‘AI Guide’ feature of Sway provokes thought by posing challenging questions, with the purpose of enhancing students’ reasoning skills.

Besides encouraging critical thinking, the tool also possess an ability to propose alterations to language it deems offensive. A sample debate scenario portrayed in the introductory video of Sway involves the discussion on whether the US ‘should prioritize Palestinian rights and cease arms delivery to Israel.’ Columbia University, as part of its plan for significant changes, is backing the use of this AI debating tool.

The changes come in the wake of a $200 million settlement designed to counter antisemitism on its premises — an agreement that restores Columbia’s eligibility for receiving up to $1.3 billion in federal funds. To meet the terms of this settlement, the university has an obligation to produce vast amounts of data, implement ‘firm rules to curb disruptive demonstrations’, and ‘enhance regulation of students from abroad.’

Columbia’s management has expressed optimism that deploying the Sway AI tool will foster productive discussions within the campus community. However, the approach of using AI to address societal issues doesn’t escape from criticism. Some detractors argue that this method oversimplifies problems, dismissing the historical and political intricacies that underline these matters.

There is also a concern that such a tool pays too much attention on promoting a harmonious atmosphere among participants instead of providing concrete solutions to the issues at hand. One skeptic even confronted Columbia, insinuating that the university is merely seeking quick-fix solutions. However, the driving concept behind Sway is not to influence students’ beliefs, but to inspire openness towards considering opposing viewpoints, thereby reducing hostility within the discourse.

Despite the backlash, Sway stands by the value they offer to the student community. They affirm that their aim is not to shift viewpoints but to promote a propensity amongst the students to be receptive to contrasting arguments, while keeping animosity to a minimum.