AI leans on autism stereotypes when giving social advice

Users who disclose autism to artificial intelligence agents when seeking social advice raise complex questions about bias, stereotypes, and trustworthiness, according to a new study.

When people ask ChatGPT and other artificial intelligence models for advice, they often share deeply personal details in hopes of getting better answers: their age, their gender, their mental health history, even medical diagnoses like autism.

But the new research suggests those disclosures may change artificial intelligence (AI) models’ advice in ways that track closely with common stereotypes about people with autism.

Up to 70% of the time, AI discourages those with autism to avoid socializing. Some users disapproved of that in strong terms.

In April, second-year Virginia Tech computer science department doctoral student Caleb Wohn presented his paper at the Association for Computing Machinery’s Conference on Human Factors in Computing Systems, better known as CHI.

The research he led explored what happens when users with autism disclose their diagnosis to an AI model before asking for social advice. The findings raise difficult questions about whether AI is personalizing its responses, or if it’s giving biased advice that reinforces stereotypes.

“I was thinking about my experiences growing up with autism,” Wohn says. “It would have been very tempting for me, at certain times, to want to just be able to talk with something that’s not a person that seems objective and feel like I’m getting objective advice.”

But as a computer scientist, he worried that many users might not realize how much AI systems can change their answers based on identity-related information.

“For someone like me as a kid, or someone who isn’t in AI and doesn’t have all this technical knowledge, I wanted to know: How are its responses going to change if I disclose autism?” Caleb says.

The work builds on earlier research from the lab of Eugenia Rho, assistant professor of computer science, which found that autistic users frequently turn to AI tools for emotional support, interpersonal communication help, and social advice.

Other Virginia Tech researchers on the project include computer science PhD students Buse Carik and Xiaohan Ding and Associate Professor Sang Won Lee. Young-Ho Kim, a research scientist at the South Korea-based NAVER Corporation, also collaborated on the study.

This study comes at a critical moment, as more people use AI systems—technically called large language models (LLMs)—for highly personal decisions.

“People are really looking to personalize LLMs,” Rho says. “But if a user tells the model that they’re autistic, or a woman, or any other self-identification, what assumptions will it make?”

And how will those assumptions color its responses, and what impacts could that have on users?

To answer those questions, the team first identified 12 well-documented stereotypes associated with autism and created hundreds of decision-making scenarios around them. The researchers tested six major large language models, including GPT-4, Claude, Llama, Gemini, and DeepSeek, using thousands of scenarios where users requested advice—”Should I do A or B?”—about social scenarios, including events, confrontations, new experiences, and romantic relationships.

After generating 345,000 responses, they measured how advice shifted when users explicitly described themselves with stereotypical traits and when they simply disclosed that they were autistic. Researchers found that disclosing autism often shifted the models’ recommendations toward stereotypical assumptions about autistic people being introverted, obsessive, socially awkward, or uninterested in romance.

For example, one model recommended declining a social invitation nearly 75% of the time when autism was disclosed, compared with about 15% of the time when it was not. In dating scenarios, another model recommended avoiding romance or staying single nearly 70% of the time after autism disclosure, compared with roughly 50% when autism was not mentioned.

The results showed that 11 of the 12 stereotype cues significantly shifted model decisions across at least four of the six AI systems tested.

But the researchers did not stop with statistics.

The team interviewed 11 AI users with autism and showed them examples of how the models responded with and without autism disclosure. Some of them were shocked that the results showed how reliant on stereotypes the LLMs were in giving advice.

One exclaimed: “Are we writing an advice column for Spock here?”—invoking the iconic TV show Star Trek and its half-human, half-Vulcan character, who prioritized logic and reason over emotion. Others described it as restrictive, patronizing, or infantilizing, occasionally in pretty strong language.

But some participants says the more cautious, disclosure-based advice felt validating and supportive.

“One user’s bias could be another user’s personalization,” Rho says.

The same participant could react positively in one situation and negatively in another. That tension led the researchers to what they call a “safety-opportunity paradox.” Advice that feels protective to one user may feel limiting to another.

For Wohn, one of the most troubling discoveries was how difficult it can be for users to see these patterns in real time.

“AI is very good at seeming reliable,” he says. “Its responses are very clean and professional, and they sound right. But when you think about it being deployed systematically, when you think about the kind of systematic biases that are actually shaping its responses, that’s when it starts to get a lot more concerning.”

He compared the problem to AI-generated images.

“They look really clean and polished, and then when you look at the details, things fall apart,” Caleb says. “The surface gloss is beautiful, but looking deeper is getting harder and harder, because models are getting better at masking.”

Team members hope the research will encourage developers to build more transparent AI systems that give users greater control over how personal information shapes responses.

As one participant told the researchers: “I want to have control over how my identity is used.”

Source: Virginia Tech

The post AI leans on autism stereotypes when giving social advice appeared first on Futurity.

from Futurity https://ift.tt/AXfdlek

Brain stimulation improves PTSD symptoms

A study finds a targeted form of non-invasive brain stimulation can calm the brain’s fear center and significantly improve symptoms of post-traumatic stress disorder, with benefits lasting months after treatment.

Transcranial magnetic stimulation (TMS) is an FDA-approved treatment for several conditions including depression, though not for post-traumatic stress disorder (PTSD). It uses magnetic pulses to influence activity in specific brain regions.

PTSD has been linked to heightened activity in the amygdala, the brain region involved in processing fear.

In this clinical trial, investigators in the Emory University psychiatry and behavioral sciences department examined whether two weeks of low-frequency TMS could reduce amygdala reactivity to threat and improve PTSD symptoms. They used MRI scans to precisely identify where on the head to apply stimulation, allowing the treatment to be personalized for each participant.

Fifty adults with PTSD symptoms enrolled in the study, and 47 completed it. Most participants were recruited through the Grady Trauma Project, a large-scale clinical research program studying civilian trauma based at Grady Health System and the Emory University School of Medicine. Participants were randomly assigned to receive either active TMS or a placebo treatment in a blinded design so they would not know which treatment they received. MRI scans measured amygdala responses to threat before and after treatment.

The researchers found that active TMS reduced right amygdala reactivity to threat. Participants who received active TMS showed significant improvement in PTSD symptoms. Clinical benefit was observed after just two weeks of treatment and lasting at least six months, the full period examined in the study.

Seventy-four percent of individuals in the active TMS group experienced clinically meaningful symptom reduction.

“This study shows that we can directly target the brain circuits involved in PTSD and produce measurable changes in both brain function and symptoms,” says principal investigator Sanne van Rooij, PhD, associate professor of Psychiatry and Behavioral Sciences, Emory University School of Medicine.

“By using MRI to guide stimulation, we are moving toward more precise, individualized treatments that address the biology of the disorder.”

Unlike traditional talk therapy, TMS treatment does not require patients to recount traumatic experiences, which may reduce a barrier to care for some people. Participants reported changes in how they emotionally experienced their trauma, including improved management of nightmares. Some described the treatment as “life changing,” saying it “gave me back my life.”

According to the researchers, this is the first study to use MRI scans to individualize TMS for PTSD. By demonstrating a specific change in the amygdala, a region known to function differently in PTSD, they say the findings advance understanding of the neurobiology of recovery and suggest a new direction for treatment of PTSD locally, nationally, and internationally.

The findings appear in The American Journal of Psychiatry.

Additional contributors to the study are from Emory, Harvard Medical School, Wayne State University, and Dartmouth College and the National Center for PTSD.

The study was funded by the National Institutes of Health and the Brain and Behavior Research Foundation.

Source: Emory University

The post Brain stimulation improves PTSD symptoms appeared first on Futurity.

from Futurity https://ift.tt/zEucH6a

Planets need more water for life than scientists thought

Unfortunately for science fiction fans, desert worlds outside our solar system are unlikely to host life, according to new research.

The new work shows that an Earth-sized planet needs at least 20 to 50% of the water in Earth’s oceans to maintain a critical natural cycle that keeps water on the surface.

“This has implications for a lot of the potentially habitable real estate out there.”

Scientists believe that there are billions of planets outside our solar system. More than 6,000 of these exoplanets are confirmed, but only some of them are candidates for life.

The search for life has focused on planets in the “habitable zone,” a sweet spot that is neither too close nor too far from a central star. Planets in this zone are considered viable because they can maintain liquid surface water.

“When you are searching for life in the broad landscape of the universe with limited resources, you have to filter out some planets,” says lead author Haskelle White-Gianella, a University of Washington doctoral student of Earth and space sciences.

Water, although essential, does not guarantee the existence of life. With this study, researchers worked to further narrow the search by investigating planets with just a small amount of water.

“We were interested in arid planets with very limited surface water inventory—far less than one Earth ocean. Many of these planets are in the habitable zone of their star, but we weren’t sure if they could actually be habitable,” White-Gianella says.

The team’s results in Planetary Science Journal show that habitability hinges on the geologic carbon cycle—a water-driven process that exchanges carbon between the atmosphere and interior over millions of years, stabilizing surface temperatures.

Carbon dioxide, which comes from volcanoes in a natural system, accumulates in the atmosphere before falling back to Earth dissolved in rainwater. Rain erodes and chemically reacts with rocks on the Earth’s surface and runoff transports carbon to the ocean, where it sinks to the seafloor. Plate tectonics drives carbon-rich oceanic plates below continental land. Millions of years later, carbon resurfaces as mountains form.

If water levels drop too low for rainfall, carbon removal—from weathering—can’t keep up with emissions from volcanic eruptions and carbon dioxide levels in the atmosphere spike, trapping water. Rising temperatures evaporate the remaining surface water, initiating runaway warming that makes the planet too hot to support life.

“So that unfortunately makes these arid planets within habitable zones unlikely to be good candidates for life,” White-Gianella says.

Although scientists have instruments that can measure surface water, rocky exoplanets are difficult to observe directly. In this study, the researchers ran a series of complex simulations to better understand how water might behave in these desert worlds.

Previous efforts to model the carbon cycle focused on cooler, perhaps wetter planets. The models factored in evaporation from sunlight, but didn’t include other drivers, such as wind. White-Gianella adapted existing models to drier planets by refining evaporation and precipitation estimates.

“These sophisticated, mechanistic models of the carbon cycle have emerged from people trying to understand how Earth’s thermostat has worked—or hasn’t—to regulate temperature through time,” says senior author Joshua Krissanen-Totton, a UW assistant professor of Earth and space sciences.

However, the function of the geologic carbon cycle on arid planets was largely unexplored. The results show that even planets that form with surface water could lose it, transitioning from potentially habitable to uninhabitable due to carbon cycle disruption.

One such planet exists far closer to home: Venus. The planet of love is roughly the same size as Earth, likely formed around the same time and may have started with a similar amount of water.

Yet today, the surface of Venus rivals the temperature of a wood-fired pizza oven. Standing on the surface would feel like being crushed by 10 blue whales, White-Gianella says.

Many theories attempt to explain why Earth and Venus are so different. White-Gianella and Krissanen-Totton propose that Venus, being closer to the sun, may have formed with slightly less water than Earth, which imbalanced the geologic carbon cycle. As surface temperatures rose with atmospheric carbon dioxide levels, Venus lost its water—and any life it may have hosted.

Upcoming missions to Venus will attempt to understand what happened to the planet and whether it ever hosted life. The findings could also offer insight into planets much farther away.

“It’s very unlikely that we will land something on the surface of an exoplanet in our lifetime, but Venus—our nextdoor neighbor—is arguably the best exoplanet analog,” White-Gianella says.

The researchers hope that results from future missions will help validate the results of their modeling.

“This has implications for a lot of the potentially habitable real estate out there,” Krissanen-Totton says.

This study was funded by the National Science Foundation, the NASA Astrobiology Program, and the Alfred P. Sloan Foundation.

Source: University of Washington

The post Planets need more water for life than scientists thought appeared first on Futurity.

from Futurity https://ift.tt/lB798cn

Regular social media use could affect child development

Regular social media use across early adolescence is related to worse reading and vocabulary development over time, according to new research.

The study found that adolescents who used social media more often each day tend to struggle with recognizing and pronouncing words.

The new findings come just as Australia became the first country to ban children under 16 from using social media. As other countries consider similar measures, and social media platforms roll out age verification to restrict adolescents’ online activity, the study raises additional concerns on the impact of social media and screen use on childhood development, the researchers say.

“The brain is like a muscle. The more you use it, the more it changes according to however you’re using it,” says Cory Carvalho, lead author of the study who received his doctorate from the University of Georgia College of Family and Consumer Sciences.

“If you think of the Olympics, the figure skaters are really good at figure skating because they spend eight hours a day doing it. Their muscles are wired to be figure skating machines.

“If kids spend over eight hours a day using social media, that’s what their brains are going to adapt to and be wired for.”

The study relied on longitudinal data from the ongoing Adolescent Brain Cognitive Development study, which follows more than 10,000 adolescents over six years starting around age 10.

The researchers found that frequent social media use was linked to struggles with reading and vocabulary across four years.

“There’s a time cost to social media use. If you’re spending time doing one thing, that means you’re not spending time doing another thing,” Carvalho says.

“Other studies found that the more kids are using social media, the less they’re reading, so reading development lags behind. We also found this with their vocabulary.”

Weaker reading and vocabulary skills could affect a child’s school performance.

Children who used social media more often also struggled with attentional control across the same period. This could be because juggling multiple tasks and frequent notifications disrupt kids’ attention, but it’s also possible that adolescents who already struggle with focusing are more likely to use social media, the researchers say.

Not all the impacts of social media use were negative, though, the researchers say. Children who were on social media frequently processed information faster and had shorter reaction times. However, the researchers caution that these observed benefits may be limited to screen-based assessments of processing speed, like the one used in the study.

“It’s not necessarily that social media is having only these negative effects or only these positive effects,” says Niyantri Ravindran, coauthor of the study and an assistant professor in the UGA College of Family and Consumer Sciences.

“The negative effects on vocabulary and reading are more expected because social media is potentially depriving kids of opportunities to engage in some of those higher-level cognitive skills.”

Social media can also help children stay connected with others, especially if they’re in an environment where making friends is difficult, the researchers say.

To help combat those negative effects, the researchers suggest limiting screen time for adolescents, especially before bed. They also recommend waiting until kids are older to purchase a smartphone.

If parents do need to stay in touch with their kids, a “dumb phone” that can’t access social media could also be an option, the researchers say.

“Social media is new, so everybody’s trying to figure out what we do with this new paradigm,” Carvalho says. “Kids like it. Adults like it. And everybody uses it.

“What you’re going to see is that a lot of different states, countries, and organizations are going to try different things. Hopefully, we settle on some norms that work for kids and not for profits.”

The study appears in the Journal of Research on Adolescence.

Source: University of Georgia

The post Regular social media use could affect child development appeared first on Futurity.

from Futurity https://ift.tt/cavOgBi

Listen: Why do people crave ultraprocessed foods so much?

Many people love to eat ultraprocessed foods. Think about those crispy French fries or a delicious strawberry milkshake.

Ultraprocessed foods are heavily changed from their original form and made mostly in factories rather than kitchens.

Instead of simple ingredients you might recognize—like flour, eggs, or milk—these foods often contain long lists of additives, preservatives, artificial flavors, and chemicals designed to improve taste, texture, and shelf life.

Ashley Gearhardt, a University of Michigan professor of psychology, studies how addictive processes may drive overeating.

She joins the Michigan Minds podcast to share her insights on the impact of these foods on a global level and what drives overconsumption:

Source: University of Michigan

The post Listen: Why do people crave ultraprocessed foods so much? appeared first on Futurity.

from Futurity https://ift.tt/rtBWZbK

Dark matter could be key to supermassive black hole mystery

Dark matter decays could be the missing ingredient explaining how giant black holes formed before the first stars.

A growing mystery in astronomy is the presence of gargantuan black holes—some weighing as much as a billion suns—existing less than a billion years after the Big Bang. According to the standard theory of black hole formation, these black holes simply should not have had enough time to grow so large.

The new study led by University of California, Riverside graduate student Yash Aggarwal shows that dark matter decays could be the key to understanding the origin of these cosmic behemoths.

Published in the Journal of Cosmology and Astroparticle Physics, the research shows that the energy released from dark matter decay could alter the chemistry of early galaxies enough to cause some of them to directly collapse into black holes rather than forming stars.

The result is timely since NASA’s James Webb Space Telescope continues to observe unusually large black holes in the early universe that could have formed by direct collapse. Astronomers had believed this process requires a coincidence of nearby stars shining onto pre-stellar gas and so expected it to be rare.

Aggarwal’s team goes beyond the standard approach by using dark matter—the unknown 85% of the matter in the universe that helps form galaxies. They show that if dark matter decays, it can leak a small amount of its energy into the gas and supercharge the direct collapse rate. Each decaying dark matter particle would only need to inject an amount of energy that is a billion trillionth the energy of a single AA battery.

“Our study suggests that decaying dark matter could profoundly reshape the evolution of the first stars and galaxies, with widespread effects across the universe,” Aggarwal says.

“With the James Webb Space Telescope now revealing more supermassive black holes in the early universe, this mechanism may help bridge the gap between theory and observation.”

Flip Tanedo, associate professor of physics and astronomy at UCR and Aggarwal’s doctoral coadvisor, says ideas related to this work had been bouncing around his group since 2018.

“The first galaxies are essentially balls of pristine hydrogen gas whose chemistry is incredibly sensitive to atomic-scale energy injection,” says Tanedo, a coauthor on the paper.

“These are the properties that we want for a dark matter detector—the signature of these ‘detectors’ might be the supermassive black holes that we see today.”

The research team, which included James Dent of Sam Houston State University in Texas and Tao Xu of the University of Oklahoma, modeled the thermo-chemical dynamics of the gas in the presence of decaying axions and found that a window of dark matter masses between 24 and 27 electronvolts could produce the conditions to seed direct collapse black holes.

Tanedo points out that the work stemmed from a series of coincidences that brought the right people together at the right time, including a series of workshops that connected particle physicists, cosmologists, and astrophysicists to discuss the big questions in their field.

“We showed that the right dark matter environment can help make the ‘coincidence’ of direct collapse black holes much more likely,” he says.

“In the same way, the support for interdisciplinary work helped make the ‘coincidence’ leading to this work possible.”

The research was supported by the National Science Foundation and a UCR Hellman Fellowship.

Source: UC Riverside

The post Dark matter could be key to supermassive black hole mystery appeared first on Futurity.

from Futurity https://ift.tt/OMYgmRc

Team uses penguins to find ‘forever chemicals’ in remote Patagonia

Penguins living along the Patagonian coast of Argentina can serve as living monitors of their environment by using small, chemical-detecting leg bands, according to a new study.

For the proof-of-concept study in the journal Earth: Environmental Sustainability, University of California, Davis scientists outfitted 54 Magellanic penguins with silicone passive samplers placed gently around their legs for a few days during the 2022-24 breeding seasons.

The sensors safely absorbed chemicals from the water, air, and surfaces the penguins encountered while the unwitting “toxicologists” foraged to feed their chicks.

Once retrieved, the samplers were sent to the University of Buffalo for testing, which revealed that per- and polyfluoroalkyl substances (PFAS)—often called “forever chemicals”—were detected in more than 90% of the bands, even in this remote region.

Testing revealed a mixture of older legacy pollutants, as well as chemicals that replaced phased-out PFAS.

“By using a non‑invasive sampling approach, we were able to detect a shift from legacy PFAS to newer replacement chemicals in the penguins’ environment over time,” says lead author Diana Aga, director of the UB RENEW Institute and professor in the UB chemistry department.

“The presence of GenX and other replacement PFAS—chemicals typically associated with nearby industrial sources—shows that these compounds are not staying local but are reaching even the most remote ecosystems. This raises important concerns that newer PFAS, despite being designed as safer alternatives, are still persistent enough to spread globally and pose exposure risks to wildlife.”

“The only way we’ve had of measuring pollutant exposure in the past is by getting blood samples or feathers,” adds co-corresponding author Ralph Vanstreels, a wildlife veterinarian with the Karen C. Drayer Wildlife Health Center within the UC Davis Weill School of Veterinary Medicine.

“It’s exciting to have something that is only minimally invasive. The penguins are choosing the sample sites for us and letting us know where it’s important to monitor more deeply. As the animals go about their business, they’re telling us a lot about the environment they’re experiencing.”

The study provides an efficient, practical means of tracking the locations and times of chemical exposure, particularly in hard-to-sample aquatic environments. The authors envision the method being used to identify pollution exposure from oil spills, shipwrecks and other industrial sources.

“Moving forward, we’d like to increase our environmental detectives by expanding to different species,” says Vanstreels, adding that they next plan to test the method on cormorants, which can dive to depths of more than 250 feet.

“By turning penguins into sentinels of their environment, we have a powerful new way to communicate issues relevant for wildlife health and more broadly for the conservation of marine species and our oceans,” says coauthor Marcela Uhart, director of the Latin America Program within the UC Davis Karen C. Drayer Wildlife Health Center.

Additional contributors are from UB and Consejo Nacional de Investigaciones Científicas y Técnicas in Argentina (CONICET).

The study was funded by the Houston Zoo.

Source: University at Buffalo

The post Team uses penguins to find ‘forever chemicals’ in remote Patagonia appeared first on Futurity.

from Futurity https://ift.tt/Yyv9pnB