Learning through disagreement

Sway is a chat platform that connects students with differing perspectives into one-on-one conversations and facilitates better discussions between them. Developed with Heterodox Academy and inspired by John Stuart Mill's radical view that engaging with opposing perspectives is an essential tool for improving reasoning and solving complex problems, Sway aims to create online spaces where we can all learn to discuss controversial issues more openly and constructively.

Students practice intellectual humility

Most constructive dialogue programs can only report what participants say about themselves. Sway provides behavioral evidence — for example, each student’s opinion is measured before and after the conversation.

10,000+ MATCHED
PRE/POST OPINION PAIRS

Moderating extreme views

0×

more likely to moderate an extreme view than to harden a moderate one.

In one 30‑min chat

drop in the share of students holding extreme positions.

With polarizing topics

of students who entered at a polar extreme left holding a more moderate view.

Benchmarked against a control group: students who did not converse on Sway showed only 14% of this effect.

Inclusion & fairness

Guide treats all students with respect

Guide, Sway’s AI discussion facilitator, participates in all student chats. After each chat, students rate the statement:

“Guide treated me and my partner with equal respect.”

Only 3% disagree.

N = 3,500+ STUDENTS

87%AGREE

10%NEUTRAL

Guide doesn’t offer opinions.

It joins every conversation to pose challenging questions, surface hidden assumptions, and de‑escalate tense moments — without ever taking sides. Instead of arguing, it pushes students to reason more carefully and engage the strongest version of the opposing view.

600+ politically diverse Americans found no bias.

Partisan reviewers evaluated real transcripts on abortion, healthcare, immigration, and trans athletes. Left and right perceived bias against their own side to the same degree — the signature of neutrality — and correctly flagged bias when it was experimentally introduced.

Depolarizing campus discourse

How student attitudes shift

Students at the extremes moderate. Students with no opinion start forming their views.

10,264 CONVERSATIONS 55.8% OF EXTREME STARTERS MODERATED 4.7× DEPOLARIZATION‑TO‑POLARIZATION RATIO

7Strongly agree

55%

← Strong opinions soften

6Moderately agree

48%

18%

5Somewhat agree

36% ← Moderate positions strengthen

4No idea / opinion

70% the undecided develop a view

3Somewhat disagree

35%

2Moderately disagree

48%

14%

1Strongly disagree

57%

← MOVED TOWARD CENTER

MOVED AWAY FROM CENTER →

Share of students at each starting position who moved toward or away from the center of a 7‑point scale after one conversation. Students at the poles (1 and 7) can only move inward; however, regression to the mean explains only a small fraction of this change.

Between partners

Rational updating, one pair at a time

Many people expect the most common outcome to be both students moving closer to the center — but that’s not what we see in the data. More than twice as often, only one student moves toward the other. This is rational updating in light of evidence, not compromise for the sake of getting along.

N = 5,079 CONVERSATION PAIRS 72.0% ENDED CLOSER TOGETHER 16.9% ENDED FARTHER APART

the signature of rational updating →

5.6%n = 286

11.0%n = 561

11.3%n = 574

22.1%

n = 1,123

49.9%

n = 2,535

Both moved
apart

Neither
moved

One moved away
from partner

Both moved
closer together

Only one moved
toward partner

How student opinions changed after engaging with an opposing viewpoint. In 72.0% of pairs at least one partner moved toward the other; in only 5.6% did both move apart. The ordering is unchanged on the hardest political topics — immigration, Israel–Palestine — with fewer chats but the same shape.

Privacy & safety

Students treat each other with respect

One‑on‑one, students engage with respect and curiosity — nothing like the social‑media caricature of Gen Z.

Moderation at scale

<0%

of student messages flagged as potentially unconstructive.

250,000⁺MESSAGES REVIEWED

Personal attacks and insults are blocked automatically.

Student protections

Encrypted conversations

Instructors and staff cannot read identifiable transcripts or view any individual student’s opinions.

Students control what’s shared

Students choose what goes to instructors; delete your account at any time.

Never used to train commercial AI

Student data stays out of third‑party model training.

Research is opt‑in

Participation in studies is always a student’s own choice.

some of the schools where students have used Sway

Carnegie Mellon U
Columbia University
Community College of Baltimore
George Washington U
Georgia State
Grand Valley State
Harvard University
Indiana University
Madison Area Technical College
Michigan State U
Northeastern Illinois U
Ohio State University
Pennsylvania State U
Pitzer College
Providence College
Purdue University
Rutgers University
Saint Louis University
Stanford University
Swarthmore College
U of Alabama at Birmingham
UNC Chapel Hill
UNC Charlotte
UCLA
U of Alabama
U of Delaware
U of Georgia
U of Pennsylvania
U of Ottawa
U of Queensland
U of West Florida
Virginia Tech

Map of the United States with colored dots marking the schools where students have used Sway

SWAYBETA.AI