(Press-News.org) Have you ever tried to convince a conspiracy theorist that the moon landing wasn’t staged? You likely didn’t succeed, but ChatGPT might have better luck, according to research by MIT Sloan School of Management professor David Rand and American University professor of psychology Thomas Costello, who conducted the research during his postdoctoral position at MIT Sloan.
In a new paper “Durably reducing conspiracy beliefs through dialogues with AI” published in Science, the researchers show that large language models can effectively reduce individuals’ beliefs in conspiracy theories — and that these reductions last for at least 2 months — a finding that offers new insights into the psychological mechanisms behind the phenomenon as well as potential tools to fight the spread of conspiracies.
Going down the rabbit hole
Conspiracy theories — beliefs that certain events are the result of secret plots by influential actors — have long been a subject of fascination and concern. Their persistence in the face of counter-evidence has led to the conclusion that they fulfill deep-seated psychological needs, rendering them impervious to facts and logic. According to this conventional wisdom, once someone “falls down the rabbit hole,” it’s virtually impossible to pull them back out.
But for Rand, Costello, and their co-author professor Gordon Pennycook from Cornell University, who have conducted extensive research on the spread and uptake of misinformation, that conclusion didn’t ring true. Instead, they suspected a simpler explanation was at play.
“We wondered if it was possible that people simply hadn’t been exposed to compelling evidence disproving their theories,” Rand explained. “Conspiracy theories come in many varieties — the specifics of the theory and the arguments used to support it differ from believer to believer. So if you are trying to disprove the conspiracy but haven’t heard these particular arguments, you won’t be prepared to rebut them.”
Effectively debunking conspiracy theories, in other words, would require two things: personalized arguments and access to vast quantities of information — both now readily available through generative AI.
Conspiracy conversations with GPT4
To test their theory, Costello, Pennycook, and Rand harnessed the power of GPT-4 Turbo, OpenAI’s most advanced large language model, to engage over 2,000 conspiracy believers in personalized, evidence-based dialogues.
The study employed a unique methodology that allowed for deep engagement with participants' individual beliefs. Participants were first asked to identify and describe a conspiracy theory they believed in using their own words, along with the evidence supporting their belief.
GPT-4 Turbo then used this information to generate a personalized summary of the participant's belief and initiate a dialogue. The AI was instructed to persuade users that their beliefs were untrue, adapting its strategy based on each participant’s unique arguments and evidence.
These conversations, lasting an average of 8.4 minutes, allowed the AI to directly address and refute the specific evidence supporting each individual’s conspiratorial beliefs, an approach that was impossible to test at scale prior to the technology’s development.
A significant — and durable — effect
The results of the intervention were striking. On average, the AI conversations reduced the average participant's belief in their chosen conspiracy theory by about 20%, and about 1 in 4 participants — all of whom believed the conspiracy beforehand — disavowed the conspiracy after the conversation. This impact proved durable, with the effect remaining undiminished even two months post-conversation.
The AI conversation’s effectiveness was not limited to specific types of conspiracy theories. It successfully challenged beliefs across a wide spectrum, including conspiracies that potentially hold strong political and social salience, like those involving COVID-19 and fraud during the 2020 U.S. presidential election.
While the intervention was less successful among participants who reported that the conspiracy was central to their worldview, it did still have an impact, with little variance across demographic groups.
Notably, the impact of the AI dialogues extended beyond mere changes in belief. Participants also demonstrated shifts in their behavioral intentions related to conspiracy theories. They reported being more likely to unfollow people espousing conspiracy theories online, and more willing to engage in conversations challenging those conspiratorial beliefs.
The opportunities and dangers of AI
Costello, Pennycook, and Rand are careful to point to the need for continued responsible AI deployment since the technology could potentially be used to convince users to believe in conspiracies as well as to abandon them.
Nevertheless, the potential for positive applications of AI to reduce belief in conspiracies is significant. For example, AI tools could be integrated into search engines to offer accurate information to users searching for conspiracy-related terms.
“This research indicates that evidence matters much more than we thought it did — so long as it is actually related to people’s beliefs,” Pennycook said. “This has implications far beyond just conspiracy theories: Any number of beliefs based on poor evidence could, in theory, be undermined using this approach.”
Beyond the specific findings of the study, its methodology also highlights the ways in which large language models could revolutionize social science research, said Costello, who noted that the researchers used GPT-4 Turbo to not only conduct conversations but also to screen respondents and analyze data.
“Psychology research used to depend on graduate students interviewing or conducting interventions on other students, which was inherently limiting,” Costello said. “Then, we moved to online survey and interview platforms that gave us scale but took away the nuance. Using artificial intelligence allows us to have both.”
These findings fundamentally challenge the notion that conspiracy believers are beyond the reach of reason. Instead, they suggest that many are open to changing their views when presented with compelling and personalized counter-evidence.
“Before we had access to AI, conspiracy research was largely observation and correlational, which led to theories about conspiracies filling psychological needs,” said Costello. “Our explanation is more mundane — much of the time, people just didn’t have the right information.”
END
Can AI talk us out of conspiracy theories?
New MIT Sloan research shows that conversations with large language models can successfully reduce belief in conspiracy theories
2024-09-12
ELSE PRESS RELEASES FROM THIS DATE:
‘Even the deepest of rabbit holes may have an exit’
2024-09-12
Embargoed: Not for Release Until 2:00 pm U.S. Eastern Time Thursday, 12 September 2024
‘Even the deepest of rabbit holes may have an exit’
Pathbreaking psychology study reveals conversations with AI models can reduce conspiracy theory beliefs
(WASHINGTON, D.C.) Sept. 12, 2024 – ‘They’re so far down the rabbit hole of conspiracy theories that they’re lost for good’ is common thinking when it comes to conspiracy theorists. This generally accepted notion is now crumbling.
In a pathbreaking research study, a team of researchers from American University, ...
An exit for even the deepest rabbit holes: Personalized conversations with chatbot reduce belief in conspiracy theories
2024-09-12
Personalized conversations with a trained artificial intelligence (AI) chatbot can reduce belief in conspiracy theories – even in the most obdurate individuals – according to a new study. The findings, which challenge the idea that such beliefs are impervious to change, point to a new tool for combating misinformation. “It has become almost a truism that people ‘down the rabbit hole’ of conspiracy belief are almost impossible to reach,” write the authors. “In contrast to this pessimistic view, we [show] that a relatively brief conversation with a generative AI model can produce ...
How is open access transforming science communication?
2024-09-12
In a Policy Forum, Mark McCabe and Frank Mueller-Langer explore how new open access (OA) mandates and agreements are changing how scientists share their work. They outline key contemporary unknowns in the open access landscape, as well as avenues for continued research. Since 2003, many national governments and international organizations have supported the Berlin Declaration on Open Access (OA) to Knowledge. More recently, some governments and organizations have introduced mandates to ensure open access ...
US food waste bans fail to reduce landfill waste, except in Massachusetts
2024-09-12
State-level bans on commercial organic waste disposal have largely failed to reduce landfilled waste across the U.S., with one state standing out as the lone success, according to a new study. Massachusetts alone achieved a significant reduction in landfilled waste when it implemented food waste bans. The findings underscore the importance of well-designed and enforced policies, with Massachusetts offering a potential model for effective waste management. “Our study shows that food waste bans are far from guaranteed to be successful,” ...
Greenland landslide-induced tsunami produced global seismic signal that lasted 9 days
2024-09-12
In 2023, a massive rockslide in East Greenland, driven by glacial melt, triggered a towering tsunami and a rare global seismic signal that resonated for nine days, according to a new study. The study provides insights into how climate change-induced events like glacial thinning can lead to significant geophysical phenomena with impacts extending throughout the Earth system. Due to climate change, steep slopes are increasingly vulnerable to landslides. In Arctic regions – which are undergoing the most rapid warming globally ...
Climate change-triggered landslide unleashes a 650-foot mega-tsunami
2024-09-12
In September 2023, scientists around the world detected a mysterious seismic signal that lasted for nine straight days. An international team of scientists, including seismologists Alice Gabriel and Carl Ebeling of UC San Diego’s Scripps Institution of Oceanography came together to solve the mystery.
A new study published today in Science provides the stunning solution: In an East Greenland fjord, a mountaintop collapsed into the sea and triggered a mega-tsunami about 200 meters (650 feet) tall. The giant ...
New study reveals food waste bans ineffective in reducing landfill waste, except in Massachusetts
2024-09-12
Of the first five U.S. states to implement food waste bans, only Massachusetts was successful at diverting waste away from landfills and incinerators, according to a new study from the University of California Rady School of Management.
The paper, published today in Science, suggests a need to reevaluate current strategies, citing Massachusetts' approach as a benchmark for effective policy implementation.
Between 2014 and 2024, nine U.S. states made it unlawful for commercial waste generators—such as grocery chains—to dispose of their food waste in landfills, expecting a 10–15% waste reduction.
“We ...
New research reveals how El Nino caused the greatest ever mass extinction
2024-09-12
Mega ocean warming El Niño events were key in driving the largest extinction of life on planet Earth some 252 million years ago, according to new research.
The study, published today in Science and co-led by the University of Bristol and China University of Geosciences (Wuhan), has shed new light on why the effects of rapid climate change in the Permian-Triassic warming were so devastating for all forms of life in the sea and on land.
Scientists have long linked this mass extinction to vast volcanic eruptions in what is now Siberia. The resulting carbon dioxide emissions rapidly accelerated climate warming, resulting in widespread stagnation and the collapse ...
Climate-change-triggered landslide caused Earth to vibrate for nine days
2024-09-12
A landslide in a remote part of Greenland caused a mega-tsunami that sloshed back and forth across a fjord for nine days, generating vibrations throughout Earth, according to a new study involving UCL researchers.
The study, published in the journal Science, concluded that this movement of water was the cause of a mysterious, global seismic signal that lasted for nine days and puzzled seismologists in September 2023.
The initial event, not observed by human eye, was the collapse of a 1.2km-high mountain peak into the remote Dickson Fjord beneath, causing a backsplash of water 200 metres in the air, with a wave up to 110 metres high. This ...
Microbe dietary preferences influence the effectiveness of carbon sequestration in the deep ocean
2024-09-12
Woods Hole, Mass. (September 13, 2024) - The movement of carbon dioxide (CO2) from the surface of the ocean, where it is in active contact with the atmosphere, to the deep ocean, where it can be sequestered away for decades, centuries, or longer, depends on a number of seemingly small processes.
One of these key microscale processes is the dietary preferences of bacteria that feed on organic molecules called lipids, according to a journal article, "Microbial dietary preference and interactions affect the export of lipids to the deep ocean," published in Science.
"In ...
LAST 30 PRESS RELEASES:
Arthropods dominate plant litter decomposition in drylands
World-renowned organic chemists attend inaugural science symposium hosted by Rice’s Global Paris Center
The trees of Miami’s future
MIT team takes a major step toward fully 3D-printed active electronics
Accelerated three-year medical school students perform as well as peers in traditional four-year programs
SwRI-developed instruments will study potential habitability of Jupiter’s moon Europa
Proposed scoring system may enhance equity in organ transplantation, increase transplant rates and improve patient survival
Survivors of childhood brain cancer are more likely to be held back in school
Updating offshore turbine designs to reflect storms’ complexity is key
Hospital strain during the COVID-19 pandemic and outcomes in older racial and ethnic minority adults
Scientists unveils key role of “selfish DNA” in early human development
Bonobos may be more vulnerable than previously thought, suggests genetics study
Scripps Research scientists discover chemical probes for previously “undruggable” cancer target
Giant Magellan telescope begins primary mirror support system testing
Experimental cancer drug eliminates bone metastases caused by breast cancer in lab models
Political candidates who fight climate change stand to benefit in election
Stand up to Cancer announces new grants supporting pioneering research in six cancer types
Researchers awarded $1.3M to help military Veterans battling Acute Myeloid Leukemia
New hub for high-energy astrophysics — CTAO Science Data Management Centre opens at DESY in Zeuthen
JMIR publications CEO and Executive Editor Gunther Eysenbach achieves #1 ranking as most cited researcher in Medical Informatics for fifth consecutive year
ERC grant for groundbreaking wearable health tech
NIH announces winners of prize competition to improve postpartum maternal health and health equity through innovative diagnostics
APS and SPR honor Dr. Cynthia F. Bearer with the 2025 Mary Ellen Avery Neonatal Research Award
Election delays and voter trust
US air pollution monitoring network has gaps in coverage, say researchers
Continuous monitoring of fatigue in factory workers
Farmer ants’ wearable bacteria
Political polarization and trust
Study uncovers how silkworm moth's odor detection may improve robotics
New study links obesity to elevated hypertension risk among young middle eastern women
[Press-News.org] Can AI talk us out of conspiracy theories?New MIT Sloan research shows that conversations with large language models can successfully reduce belief in conspiracy theories