AI perceived more negatively than climate science or science in general

Annenberg Public Policy Center of the University of Pennsylvania

Facebook

XLinkedInWeChatBlueskyMessageWhatsAppEmail

ChatGPT was released to the public in late 2022, and the promise and perils of artificial intelligence (AI) have loomed large in the public consciousness ever since. Because perceptions of a new technology like AI can help shape how the technology is developed and used, it is important to understand what Americans think about AI – how positively or negatively they regard the technology, and what hopes and concerns they have about it.

In a new paper, researchers affiliated with the Annenberg Public Policy Center (APPC) of the University of Pennsylvania explore public perceptions of AI science and scientists, comparing those to perceptions of science and scientists in general, and perceptions of climate science and scientists in particular.

The researchers surveyed an empaneled national probability sample of U.S. adults about how they perceived these different scientific domains in terms of each of the “Factors Assessing Science’s Self-Presentation” (FASS) – a rubric that includes credibility, prudence, unbiasedness, self-correction, and benefit.

They found that people perceived AI scientists more negatively than climate scientists or scientists in general, and that this negativity is driven by concern about AI scientists’ prudence – specifically, the perception that AI science is causing unintended consequences. The researchers also examined whether these negative perceptions might be a result of AI being so new and unknown, but found that public perceptions of AI science and scientists did not significantly improve from 2024-2025, even as AI became a more common presence in everyday life.

Perceptions of science are often influenced by political dynamics: Climate science has long suffered from partisan politicization and, after the Covid-19 pandemic, Republicans’ confidence in medical scientists and scientists in general declined. But the researchers found that perceptions of AI are less polarized than perceptions of science and climate science. “Our research suggests that AI has not been politicized in the U.S., at least not yet,” says lead author Dror Walter, an associate professor of digital communication at Georgia State University and an APPC distinguished research fellow.

Walter says that “identifying negative perceptions can help guide messaging about new science,” and that “the public unease about AI’s potential to create unintended consequences invites transparent, well-communicated ongoing assessment of the effectiveness of self or governmental regulation of AI.”

“Public Perceptions of AI Science and Scientists Relatively More Negative but Less Politicized Than General and Climate Science” was published in PNAS Nexus on June 17, 2025, and co-authored by APPC distinguished research fellows Dror Walter, associate professor of digital communication at Georgia State University, and Yotam Ophir, associate professor of communication at the University of Buffalo, State University of New York; Patrick E. Jamieson, director of APPC’s Annenberg Health and Risk Communication Institute; and Kathleen Hall Jamieson, director of the Annenberg Public Policy Center.

---

Journal

PNAS Nexus

DOI

10.1093/pnasnexus/pgaf163 

Method of Research

Survey

Subject of Research

People

Article Title

Public Perceptions of AI Science and Scientists Relatively More Negative but Less Politicized Than General and Climate Science

Article Publication Date

17-Jun-2025

Negative perception of scientists working on AI

PNAS Nexus

Facebook

XLinkedInWeChatBlueskyMessageWhatsAppEmail

A public survey indicates that Americans have negative opinions of scientists who work on AI. Dror Walter and colleagues collected opinions about scientists from thousands of US adults via the Annenberg Science and Public Health survey and compared the perceived credibility, prudence, unbiasedness, self-correction, and benefit of scientists working on AI with those of scientists in general and climate scientists in particular. Previous work has established that high scores on these dimensions predict support for science funding and science-consistent beliefs. Respondents’ perceptions of scientists working on AI were the most negative of the three, a result driven by low scores on the “prudence” dimension, specifically the perception that AI science is causing unintended consequences. Political leanings and media consumption habits did not predict opinions about scientists working on AI to the same degree that these factors predict opinions about climate scientists, suggesting the field has not been politicized—at least not yet. Perceptions of scientists working on AI were negative in both 2024 and 2025 surveys and did not improve over time. The authors interpret the persistent perceptions of scientists working on AI as an indicator that the negativity is unlikely to be solely due to a moral panic prompted by the novelty of AI. According to the authors, the evident unease with AI science suggests that the public would welcome transparent information about the effectiveness of self or governmental regulation of the emerging technology. 

---

Journal

PNAS Nexus

Article Title

Public perceptions of AI science and scientists relatively more negative but less politicized than general and climate science

Article Publication Date

17-Jun-2025

Using AI to find persuasive public health messages and automate real-time campaigns

PNAS Nexus

LinkedIn
MessageWhatsAppEmail

AI can help public health agencies in the quest to end HIV. The United States is pursuing an initiative to end the HIV epidemic by 2030. To achieve this goal, public health agencies and organizations must remind the public about how best to avoid transmitting and acquiring the virus. Public health campaigns are costly, their effectiveness is seldom systematically assessed, and no systematic methods have been developed to build health campaigns in real-time. Dolores Albarracin and colleagues collected public health messages about HIV prevention and testing from US federal agencies, non-profit organizations, and HIV/STI researchers posting on social media. AI was then used to classify those that were actionable, relevant to men who have sex with men, and effective. An online experiment with men who have sex with men, and a field experiment with public health agencies showed that the classification model was successful in picking persuasive public heath messages. Specifically, posts selected by the AI classifier were six times more likely to be selected for reposting by government and community agencies in US counties than general posts about HIV prevention—and the target audience expressed greater interest in sharing AI-selected posts online. According to the authors, community-based organizations can save time and money by using AI to select publicly available public health messages to repost, allowing the organizations to share messages about prevention and testing more often.

---

Journal

PNAS Nexus

Article Title

Living health-promotion campaigns for communities in the United States: Decentralized content extraction and sharing through AI

Article Publication Date

17-Jun-2025

 

Teaching robots to build without blueprints

Robotic swarms build like bees in simulations

University of Pennsylvania School of Engineering and Applied Science

Facebook

XLinkedInWeChatBlueskyMessageWhatsAppEmail

 

video: 

Just like bees, the individual robots in this simulation have no master plan. Simply by responding to local cues, they are able to construct honeycomb-like structures. The method could presage a new era for manufacturing, inspired by nature.

view more 

Credit: Jordan Raney, Mark Yim

Bees, ants and termites don’t need blueprints. They may have queens, but none of these species breed architects or construction managers. Each insect worker, or drone, simply responds to cues like warmth or the presence or absence of building material. Unlike human manufacturing, the grand design emerges simply from the collective action of the drones — no central planning required.

Now, researchers at Penn Engineering have developed mathematical rules that allow virtual swarms of tiny robots to do the same. In computer simulations, the robots built honeycomb-like structures without ever following — or even being able to comprehend — a plan.

“Though what we have done is just a first step, it is a new strategy that could ultimately lead to a new paradigm in manufacturing,” says Jordan Raney, Associate Professor in Mechanical Engineering and Applied Mechanics (MEAM), and the co-senior author of a new paper in Science Advances. “Even 3D printers work step by step, resulting in what we call a brittle process. One simple mistake, like a clogged nozzle, ruins the entire process.”

Manufacturing using the team’s new strategy could prove more robust — no hive stops construction because a single bee makes a mistake — and adaptable, allowing for the construction of complex structures onsite rather than in a factory. “We’ve just scratched the surface,” says Raney. “We’re used to tools that execute a plan. Here, we’re asking: how does order emerge without one?”

 

A New Paradigm for Building

 

From stone tools to space stations, human engineering has relied on planning: imagine the result, then design and build it in steps. Even 3D printing follows the same logic, slicing a model into thousands of precise instructions for the printer to follow.

“What’s so different about our approach is that it sidesteps that entire paradigm,” says Mark Yim, Asa Whitney Professor in MEAM, Ruzena Bajcsy Director of the General Robotics, Automation, Sensing and Perception (GRASP) Lab and the paper’s other co-senior author. “There’s no pre-written script, no centralized plan. Each robot just reacts to its surroundings.” 

Because no single robot needs to understand the big picture, construction can continue even if some robots fail or go off course. And since all robots operate simultaneously, rather than waiting their turn, the process could one day be faster — and more robust to individual failures.

 

Planning Behavior, Not Buildings

 

While inspired by nature, the researchers didn’t try to precisely mimic how bees, ants or other natural builders behave. Unlike artificial intelligence researchers, who often look to the brain for clues about how to design learning algorithms, this team wasn’t trying to copy biology. 

Instead, they focused on the deeper principle that nature uses: simple behaviors, repeated many times in parallel, can add up to create something complex and useful.

“What we wanted was a system where structure emerges from behavior,” says Raney. “Not because the robots know what they’re building, but because they’re following the right set of local rules.”

The hard part was figuring out what those rules should be. “There are countless ways you could program a robot to react to its surroundings,” says Yim. “We had to narrow it down to something simple, but still useful.” 

 

Finding the Right Rules

 

In the end, the team focused on a handful of basic questions: What should a robot do when it bumps into something another robot built? Should it turn left or right, and by how much? How far should each robot go before stopping?

This resulted in a dozen variables — like the robots’ speed and the angle at which they turn left or right — that the researchers played with over the course of many simulations. “By simulating the robots’ activity,” says Raney, “we could focus on fine-tuning which rules mattered the most.” 

Ultimately, the amount of disorder in the system played a crucial role in the final structure. “The more we varied parameters like the turning angle, the more variation we got in the final structure,” says Yim. 

As prior work by Penn Engineers has found, adding the right amount of disorder to lattices like honeycombs can actually enhance their toughness. “We essentially found a lever that lets you vary the geometry of the final outcome, which can affect its resistance to cracking,” adds Raney.

 

Building the Swarm in Reality

 

While the team created prototypes, actually building a swarm of robots is still a step away. First, they plan to update their simulation to better reflect how tiny robots might work in the real world. 

“In our early models, we imagined the robots laying down material in straight lines, like a mini 3D printer,” says Yim. “But that may not be the most practical method. A better approach might be to use electrochemistry, where the robots grow metal structures around themselves.”

Making that happen will require progress in building tiny robots that can move, sense and interact with materials, but the team believes the concept itself represents perhaps their most important advance. 

“We hope this gets people thinking in new ways about how things can be built,” says Raney. “Nature doesn’t start with a master plan, it starts with lots of small actions that come together into something bigger. Now we’re learning how to do that, too.”

This study was conducted at the University of Pennsylvania School of Engineering and Applied Science and supported by the National Science Foundation (2036881).

Additional coauthors include co-first authors Jiakun Lu and Xiaoheng Zhu, as well as Walker Gosrich, all of Penn Engineering.

Jordan Raney Explains Robotic Swarm [VIDEO] | 

---

Journal

Science Advances

DOI

10.1126/sciadv.adu8260 

Method of Research

Experimental study

Subject of Research

Not applicable

Article Title

Design of nondeterministic architected structures via bioinspired distributed agents

 

Why resisting social pressure is harder than you think

Researchers find most don’t think tendency for obedience applies to them

Ohio State University

Facebook

XLinkedInWeChatBlueskyMessageWhatsAppEmail

COLUMBUS, Ohio – Whether you have a rebellious personality or not, most people imagine they are better at overcoming pressure to violate their own principles than they really are, finds a new study. 

Researchers found that most individuals think they would be more likely than the average person to disobey an immoral or unlawful order from an authority figure. 

This phenomenon, called the “better-than-average effect,” reveals that people are fairly resistant to internalizing beliefs that may harm their self-perceptions. In extreme cases, ignoring how everyone is subject to social pressure could leave a person vulnerable to the desires of malicious actors. 

“Social pressures are way more powerful and impactful than we give them credit for,” said Philip Mazzocco, lead author of the study and an associate professor of psychology at The Ohio State University. “If you fall under the sway of these pressures, you could end up engaging in behavior inconsistent with your values and morals.”

The research began as a class project designed by Mazzocco’s students, who based it on the Milgram experiment, a 1960s-era study that aimed to understand the link between obedience and authority. In these experiments, participants were asked by an authority figure to deliver what they believed to be painful and, at times, lethal electric shocks to another person. 

The Milgram experiment suggested that people would obey authority even in conflict with their own beliefs. 

The new study by Mazzocco and colleagues was recently published in the journal Current Psychology.

In this work, Mazzocco’s team had more than 400 adults read first-person accounts of the shock study before asking them to predict their responses and those of the average person. Prior to these predictions, half were told that 65% of those in the original study exhibited “complete obedience,” whereas the other half were given no additional information about the results. 

Participants were asked at what voltage level, if any, they thought they would disobey and end the study. They could choose on a dial which ranged from 1 (which meant quitting the study after the first shock was delivered) to 31 (exhibiting complete obedience throughout the experiment). On average, participants thought they themselves would quit the study around dial 7. In contrast, participants surmised that the average person would not stop the study until approximately dial 12. 

Those who were informed of the results of the Milgram study – that 65% of the original participants continued all the way to the final voltage level – did predict that the average person would administer significantly higher voltages than did those who were not told. But those who were informed did not think they themselves would deliver significantly higher shocks than did those who were not told. This was another indication of the better-than-average effect.

These results were almost identical to obedience levels reported by previous studies and correctly fit the team’s theory that most would underpredict their likely obedience in a classic Milgram scenario. This suggests that in the absence of real compliance pressures, even fully imagining yourself in a situation can still lead a person to underestimate its influence on them. 

“Just reading about a situation is not sufficient, as doing so doesn’t really internalize the point that we're all really susceptible to these pressures,” said Mazzocco. The study also likens the perceived difference between predicted and actual obedience to watching a horror movie play out from the safety of home versus the certainty of actually being pursued by someone. 

Notably, while 65.2% of participants had not heard of the Milgram experiment before, researchers found that prior knowledge of the Milgram experiment didn’t alter how participants viewed the likeliness of their participation. Finally, personality and value tests were given to participants to determine what role personal characteristics might play in a real-life situation. 

One significant predictor of actual obedience in a Milgram-like scenario was conscientiousness – the personality trait of being responsible and having a tendency to adhere to rules and norms. Those who exhibited this trait were more likely to want to appease the experimenter. 

But singular personality traits aside, not every human can be the exception to the rule, said Mazzocco. “Studies like these are relevant to society because if we all assume we’re so resistant to obedience, we are not going to immunize ourselves against authority figures who want to take advantage of us,” he said.

Such immunization techniques include learning to avoid situations where intense social pressures exist or having a strategy to deal with or escape a potentially negative encounter. 

Still, Mazzocco admits removing oneself isn’t always possible, and recommends cultivating curiosity to help a person keep true to their values. 

Other Ohio State co-authors include Katie Reitler, Lauren Little, John Korte, Monicka Ridgill and Xamina Stalnaker. 

#

Contact: Philip J. Mazzocco, Mazzocco.6@osu.edu

Written by: Tatyana Woodall, Woodall.52@osu.edu

---

Journal

Current Psychology

DOI

10.1007/s12144-025-07962-1#Abs1 

Article Title

Milgram shock-study imaginal replication: how far do you think you would go?

 

Living near harmful algal blooms reduces life expectancy with ALS

Researchers say avoiding these toxic blooms may reduce risk

Michigan Medicine - University of Michigan

Facebook

X
BlueskyMessageWhatsAppEmail

Living close to cyanobacterial harmful algal blooms — which are present nationwide but are more common in coastal and Great Lake states — heightens the rate of dying from amyotrophic lateral sclerosis, or ALS, a study suggests.

These blooms occur when cyanobacteria, also called blue-green algae, grows dense and out of control, producing toxic agents that can poison people and the environment. 

Researchers at Michigan Medicine surveyed participants with ALS who were seen at the University of Michigan Pranger ALS Clinic.

Investigators measured the duration and extent of each participant’s exposure to cyanobacterial harmful algal blooms by compiling their residential and health histories, as well as satellite data from the Cyanobacteria Assessment Network.

Many participants lived within three miles of a harmful algal bloom. Living near blooms, especially if swimming, boating or using a local and possibly contaminated water source, was associated with dying of ALS nearly one year sooner.

The results are published in the International Journal of Environmental Research and Public Health.

“Harmful algal blooms are a growing problem across the country, and it is not uncommon for people to live near them,” said senior author Stephen Goutman, M.D., Harriet Hiller Research Professor, director of the Pranger ALS Clinic and associate director of the ALS Center of Excellence at University of Michigan.

“Our study shows a clear link between living in close proximity to cyanobacteria blooms over the course of one’s life and adverse outcomes of ALS.”

People may be exposed to cyanobacteria toxins through ingestion, inhalation and contact with the skin from several activities, such as swimming and fishing.

In the study, people with the most significant exposures both lived near harmful blooms and used a private well as their water source.

Cyanobacteria produce several toxic agents that are linked neurodegenerative diseases, including ALS, Alzheimer’s and Parkinson’s. 

One such toxin, called ß-methylamino-L-alanine, has been detected in brain and cerebral spinal fluid samples of participants with ALS.

“While there is still limited research into the mechanism by which cyanobacteria toxins affect neurodegenerative diseases, our findings suggest that living near or participating in activities in these water bodies may influence the progression of ALS,” Goutman said.

ALS is in part influenced by genetics, but the disease is also largely driven by environmental factors.

The accumulation of exposures to toxins and pollutants, such as pesticides and volatile organic compounds, is known as the ALS exposome.

In the United States, cases of ALS are highest in the Midwest.

Researchers believe that is in part due to pervasive industrial and agricultural productions in the region.

“If exposure to cyanobacteria toxins is a meaningful risk factor for ALS, the large number of inland lakes from to such bacteria in the Midwest may partly explain why the disease incidence is much higher than other parts of the country,” said Stuart Batterman, Ph.D., first author and professor of environmental health sciences at the U-M School of Public Health.

“Given the complexity of ALS, we must maintain an epidemiological approach that encompasses environmental exposures over the life course. These studies provide will continue to provide insight into risk facts associated with disease onset and the effects on survival.”

Additional authors: Md Kamrul Islam, M.S., Dae Gyu Jang, Ph.D., and Eva L. Feldman, M.D., Ph.D., all of University of Michigan.

Funding/disclosures: This study was supported by the National Institute of Neurological Disorders and Stroke (NS127188), and the National Institute of Environmental Health Sciences (ES027221, ES030049, ES017885) of the National Institutes of Health.

It was also supported by the Centers for Disease Control and Prevention (TS000344).

The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the CDC.

Additional support from the Scott L. Pranger ALS Clinic Fund, the NeuroNetwork for Emerging Therapies, the Andrea and Lawrence Wolfe Brain Health Initiative, the Robert and Katherine Jacobs Environmental Health Initiative Fund and the Coleman Therapeutic Discovery Fund.

Paper cited: “Life Course Exposure to Cyanobacteria and Amyotrophic Lateral Sclerosis Survival,” International Journal of Environmental Research and Public Health. DOI: 10.3390/ijerph22050763

---

Journal

International Journal of Environmental Research and Public Health

DOI

10.3390/ijerph22050763 

Method of Research

Computational simulation/modeling

Article Title

Life Course Exposure to Cyanobacteria and Amyotrophic Lateral Sclerosis Survival

COI Statement

This study was supported by the National Institute of Neurological Disorders and Stroke (NS127188), and the National Institute of Environmental Health Sciences (ES027221, ES030049, ES017885) of the National Institutes of Health. It was also supported by the Centers for Disease Control and Prevention (TS000344). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the CDC. Additional support from the Scott L. Pranger ALS Clinic Fund, the NeuroNetwork for Emerging Therapies, the Andrea and Lawrence Wolfe Brain Health Initiative, the Robert and Katherine Jacobs Environmental Health Initiative Fund and the Coleman Therapeutic Discovery Fund.