Truth-Seeking in a Polarized World

AI and evidence-based thinking can be allies in the journey toward understanding.

Posted July 29, 2025 | Reviewed by Margaret Foley

For years, I have been fascinated by how injustice shapes people’s lives, particularly the lives of polarized groups locked in historical conflicts. In my research at Harvard University, I brought together, for the first time, descendants of Holocaust survivors and descendants of Nazis to examine their interactions.

We used a statistical approach called sequential analysis to analyze their conversations, measuring the probability of one person saying something and the likelihood of how the other would respond. What we found was striking: Both groups felt like victims, deeply entrenched in their own histories of pain and justification.

We observed similar patterns when descendants of enslaved people and slaveholders came together. Each side was strongly identified with its own suffering or a sense of inherited guilt .

The Struggle to Hear the Other Side

This realization, that each group in a conflict is so rooted in its own narrative that it struggles to hear the other side, is not limited to historical atrocities. The same dynamic plays out in marriages, where couples argue past each other; in political discussions, where opponents treat each other as enemies; and even in classrooms.

As a psychology professor who has spent three decades teaching and researching polarizing issues, I’ve noticed a shift in classroom dynamics. Discussions have become increasingly tense, more like battlegrounds where students defend fixed positions rather than explore ideas.

The prevailing model is debate: It's designed for winning, not for truth-seeking, deep understanding, or viewpoint diversity. Whether in polarized groups, relationships, classrooms, or public discourse, people often become entrenched in their views and unable to truly hear the other side. Many become overly confident in their own rightness, making it difficult to genuinely consider alternative perspectives.

The Dunning-Kruger Effect in Polarized Discussions

This overconfidence aligns with the Dunning-Kruger effect , a cognitive bias in which individuals with limited knowledge overestimate their competence. Lacking the self-awareness to recognize gaps in their understanding, they mistakenly believe they are more informed than they are.

In polarized discussions, this tendency fuels rigid thinking, making debate a tool for self-confirmation rather than a path to deeper understanding.

Introducing the Science of Diversity Method

These experiences led me to develop what I call the Science of Diversity Method. When I brought together descendants of Holocaust survivors and Nazis, and descendants of enslaved people and slaveholders, I wasn’t seeking a reconciliation meeting. I was studying their interactions, and the data revealed a consistent pattern: Each side was quick to defend its position rather than ask questions.

This insight became the foundation of the Science of Diversity Method, which encourages people to shift from defending beliefs to examining their thinking through inquiry and evidence. While the term “diversity” has become controversial in some circles, I use it simply to mean differences , in viewpoints.

As a professor of advanced research methods, I’ve spent years training students to design studies, analyze data, and think scientifically using evidence-based reasoning. Over time, I began to see that the same tools we use to examine research questions could be applied to address polarizing social issues.

Scientific thinking involves asking questions, forming hypotheses, gathering data, and being open to challenging our own assumptions. Unlike debate, which aims to defend a position, scientific thinking tests ideas and encourages intellectual humility.

The Role of Limitations in Scientific Thinking

In fact, every peer-reviewed article includes a section on limitations so other researchers can replicate the work, address those limitations, and get closer to the truth.

Imagine if, during a debate, someone said, “The limitation of my argument is…” —it would be surprising. Yet in science, acknowledging limitations is essential to truth-seeking and advancing deeper understanding.

An Evidence-Based Approach to Polarizing Issues

Take a polarizing issue like social media and mental health. One side argues that social media is harming young people’s well-being; the other sees it as reflecting existing social dynamics.

An evidence-based, scientific approach avoids taking sides. It begins by clarifying definitions (e.g., What counts as a mental health decline? Over what time frame? How is it measured? ), gathering data, testing alternative explanations, and staying open to whatever results emerge, even if they contradict prior beliefs.

How Artificial Intelligence Can Help

Now, we have a powerful new tool to enhance this process: artificial intelligence . AI can process vast amounts of data quickly and, when used responsibly, can help identify patterns and insights that might be difficult for humans to see.

Humans are influenced by beliefs, group identities, and subconscious motives, even when we try to be objective. AI, by contrast, can help us see patterns and contradictions we might otherwise miss.

For example, AI can analyze millions of social media posts to detect mood shifts, loneliness , or anxiety . It can uncover correlations across datasets from user engagement to mental health surveys and identify subgroups for whom social media may have different effects.

This moves us beyond simplistic “good” or “bad” research conclusions toward a more nuanced, deeper evidence-based understanding of social media’s impact.

AI as a Partner in Evidence-Based Thinking

Still, AI is not enough. Without scientific, evidence-based thinking, its outputs risk being misused or misunderstood. Some fear AI will replace human thinking, but I see it as a tool that augments our minds by revealing patterns that might otherwise go unnoticed. This only works if we approach AI’s data with skepticism, treating its findings as hypotheses to explore, not as ammunition to defend our beliefs.

Fostering Humility and Curiosity

By combining AI with evidence-based thinking, we can move beyond our own narrow perspectives and analyze complex, polarizing social data at an unprecedented scale.

In my classroom, I ask students not, “ Who’s right? ” but “ What evidence might shift your mind? ” and “ What are your limitations? ” These questions foster humility and curiosity, qualities that support more constructive dialogue in our polarized world.

AI and evidence-based thinking are not magic fixes, but they offer a path forward. Addressing polarizing social issues requires moving beyond debate toward truth-seeking, considering viewpoint diversity, and adopting a mindset focused not on winning, but on learning and gaining deeper insights.

Weissmark, M. S. (2004). Justice matters: Legacies of the holocaust and World War II . Oxford University Press, USA.

Weissmark, M. S. (2020). The science of diversity . Oxford University Press, USA.

Weissmark, M.S. (2025). Seeing the other side: Shifting perceptions . Amazon KDP.

Share this post Facebook Bluesky Linkedin Email

There was a problem adding your email address. Please try again.

This article is part of the Bringwise Psychology Journal — daily insights on human behavior, mental health, and personal growth.