Skip to content
Search

Latest Stories

Top Stories

How artificial intelligence can be used to reduce polarization

Rozado is an associate professor of computer science at Te Pūkenga - The New Zealand Institute of Skills and Technology. He is also a faculty fellow at Heterodox Academy's Center for Academic Pluralism. McIntosh is the author of “Developmental Politics” (Paragon House 2020) and coauthor of “Conscious Leadership” (Penguin 2020). He is cofounder and lead philosopher at the Institute for Cultural Evolution.

Amid countless reports of how social media is exacerbating political polarization, many commentators worry that artificial intelligence will come to have a similarly corrosive effect on American culture. In response to these concerns, an innovative new tool has been developed to leverage AI technology to reduce polarization: Meet DepolarizingGPT, a political chatbot designed to tackle polarization head on.

Unlike other AI models, DepolarizingGPT is focused specifically on political issues. It provides three responses to every prompt: one from a left-wing perspective, one from a right-wing perspective and a third response from a depolarizing or “integrating” viewpoint. We created this three-answer model to ameliorate political and cultural polarization by demonstrating a developmental approach to politics, one that synthesizes responsible perspectives from across the political spectrum.


The idea is to combine three models — LeftwingGPT, RightwingGPT and DepolarizingGPT — into one single system. Users are exposed to three perspectives simultaneously, moving beyond the echo chambers that often reinforce entrenched biases. Existing AIs, such as OpenAI's ChatGPT, claim to be unbiased, but this claim has been shown to be false. So rather than denying its bias, which always exists, DepolarizingGPT's three-answer model offers responsible perspectives from left, right and integrated positions. The goal is to foster a more diverse, nuanced understanding of differing political views, and to reduce the tendency to vilify the other side.

Sign up for The Fulcrum newsletter

DepolarizingGPT's left-wing responses have been fine-tuned (within the fair-use provisions of copyright law) by using content from left-leaning publications such as The Atlantic, The New Yorker and The New Republic, and from numerous left-wing writers such as Bill McKibben and Joseph Stiglitz. The model's right-wing responses have been fine-tuned with content from publications such as National Review, The American Conservative and City Journal, as well as from numerous right-leaning writers such as Roger Scruton and Thomas Sowell. And the model's depolarizing responses have been fine-tuned with content from the inclusive political philosophy of the Institute for Cultural Evolution.

The model's depolarizing answers attempt to transcend centrism and avoid simply splitting the difference between left and right. When at their best, these depolarizing responses demonstrate a kind of "higher ground" that goes beyond the familiar left-right political spectrum. Admittedly, however, some of the model's depolarizing responses inevitably fall short of this goal.

This project stems from David Rozado’s academic research, which revealed the inherent left-wing political bias of ChatGPT. To address this issue, Rozado created an experimental AI model with an opposite kind of right-wing bias. His work attracted attention from The New York Times, Wired and Fox News. The intent in demonstrating the political bias of supposedly neutral AIs was to help prevent artificial intelligence from becoming just another front in the culture war.

After reading about Rozado's work, Steve McIntosh proposed that the two team up to create an AI model that could actually help reduce political polarization. Since cofounding the Institute for Cultural Evolution in 2013, McIntosh has been working to overcome hyperpolarization by showing how America can grow into a better version of itself. His institute offers a platform of "win-win-win" policy proposals, which integrate the values of all three major American worldviews: progressive, modernist and traditional. And this same method of integrating values used to build the institute's policy platform is now programmed into DepolarizingGPT's three-answer political chatbot.

Within conventional politics, people are often faced with win-lose propositions. But by focusing on the bedrock values that most people already share, it becomes possible to discover something closer to a win-win-win solution, even if such a solution does not completely satisfy all parties. This win-win-win strategy aims to accommodate the concerns of all sides, not just to get its way, but to make authentic progress through cultural evolution.

By synthesizing values from across the political spectrum, artificial intelligence promises to help American society grow out of its currently dysfunctional political condition.

Read More

Robotic hand holding a ballot
Alfieri/Getty Images

What happens when voters cede their ballots to AI agents?

Frazier is an assistant professor at the Crump College of Law at St. Thomas University. Starting this summer, he will serve as a Tarbell fellow.

With the supposed goal of diversifying the electorate and achieving more representative results, State Y introduces “VoteGPT.” This artificial intelligence agent studies your social media profiles, your tax returns and your streaming accounts to develop a “CivicU.” This artificial clone would use that information to serve as your democratic proxy.

Keep ReadingShow less
Sen. Ron Johnson in front of a chart

Sen. Ron Johnson claims President Biden has allowed 1,700 terrorists to enter the country. That total refers to encounters (people who were stopped)

Tom Williams/CQ-Roll Call, Inc via Getty Images

Has President Joe Biden ‘let in’ nearly 1,700 people with links to terrorism?

This fact brief was originally published by Wisconsin Watch. Read the original here. Fact briefs are published by newsrooms in the Gigafact network, and republished by The Fulcrum. Visit Gigafact to learn more.

Has President Joe Biden ‘let in’ nearly 1,700 people with links to terrorism?

No.

Border agents have encountered individuals on the federal terrorist watchlist nearly 1,700 times since President Joe Biden took office — that means those people were stopped while trying to enter the U.S.

Keep ReadingShow less
Social media app icons
hapabapa/Getty Images

Urban planning can counter social media’s impact on young people

Dr. Jones is a grassroot urban planner, architectural designer, and public policy advocate. She was recently a public voice fellow through The OpEd Project.

Despite the breathtaking beauty of our world, many young people remain oblivious to it, ensnared by the all-consuming grip of social media. A recent Yale Medicine report revealed the rising negative impact social media has on teens, as this digital entrapment rewires their brains and leads to alarming mental and physical health struggles. Tragically, they are deprived of authentic life experiences, having grown up in a reality where speculation overshadows genuine interactions.

For the sake of our society’s future, we must urgently curb social media’s dominance and promote real-world exploration through urban planning that ensures accessible, enriching environments for all economic levels to safeguard the mental and physical health of the young.

Keep ReadingShow less
podcast mic in the middle of a red and blue America
Topdesigner/Getty Images

Fellowship brings Gen Z voices into democracy and podcasting

Spinelle is the founder of The Democracy Group podcast network and the communications lead for the McCourtney Institute for Democracy at Penn State.

According to Edison Research, nearly half of Gen Z are monthly podcast listeners. But their voices are largely absent from podcasts about democracy, civic engagement and civil discourse. The Democracy Group’s podcast fellowship, which recently completed its third cohort, aims to change that.

Keep ReadingShow less
Microchip labeled "AI"
Eugene Mymrin/Getty Images

Preparing for an inevitable AI emergency

Frazier is an assistant professor at the Crump College of Law at St. Thomas University. Starting this summer, he will serve as a Tarbell fellow.

Artificial intelligence is advancing at a speed and in ways that were unanticipated even by the foremost AI experts. Just a few decades ago, AI was largely theoretical, existing primarily in the realms of science fiction and academic research. Today, AI permeates nearly every aspect of our lives, from the algorithms that curate social media feeds to the autonomous systems that drive cars. This rapid advancement, while promising in many respects, also heralds a new era of uncertainty and potential peril.

Keep ReadingShow less