1 region of the brain links all your senses

A new study shows that the senses stimulate a region of the brain that controls consciousness.

Humans perceive and navigate the world around us with the help of our five senses: sight, hearing, touch, taste, and smell.

The study in the journal NeuroImage sheds new light on how sensory perception works in the brain and may fuel the development of therapies to treat disorders involving attention, arousal, and consciousness.

In the study, a research team led by Yale’s Aya Khalaf focused on the workings of subcortical arousal systems, brain structure networks that play a crucial role in regulating sleep-wake states. Previous studies on patients with disorders of consciousness—such as coma or epilepsy—have confirmed the influence of these systems on states of consciousness.

But prior research has been largely limited to tracking individual senses. For the new study, researchers asked if stimuli from multiple senses share the same subcortical arousal networks. They also looked at how shifts in a subject’s attention might affect these networks.

For the study, researchers analyzed fMRI (functional magnetic resonance imaging) datasets collected from 1,561 healthy adult participants as they performed 11 different tasks using four senses: vision, audition, taste, and touch.

They made two important discoveries: that sensory input does make use of shared subcortical systems and, more surprisingly, that all input—regardless of which sense delivered the signal—stimulates activity in two deep brain regions, the midbrain reticular formation and the central thalamus, when a subject is sharply focused on the senses.

The key to stimulating the critical central brain regions, they found, were the sudden shifts in attention demanded by the tasks.

“We were expecting to find activity on shared networks, but when we saw all the senses light up the same central brain regions while a test subject was focusing, it was really astonishing,” says Khalaf, a postdoctoral associate in neurology at Yale School of Medicine and lead author of the study.

The discovery highlighted how key these central brain regions are in regulating not only disorders of consciousness, but also conditions that affect attention and focus, such as attention deficit hyperactivity disorder. This finding could lead to better targeted medications and brain stimulation techniques for patients.

“This has also given us insights into how things work normally in the brain,” says senior author Hal Blumenfeld, a professor of neurology who is also a professor in neuroscience and neurosurgery and director of the Yale Clinical Neuroscience Imaging Center.

“It’s really a step forward in our understanding of awareness and consciousness.”

Looking across senses, this is the first time researchers have seen a result like this, says Khalaf, who is also part of Blumenfeld’s lab.

“It tells us how important this brain region is and what it could mean in efforts to restore consciousness,” she says.

Other authors include Erick Lopez, a former undergraduate researcher in Blumenfeld’s lab, and collaborators from Harvard Medical School.

Support for this work came, in part, from the National Institutes of Health.

Source: Yale

The post 1 region of the brain links all your senses appeared first on Futurity.

from Futurity https://ift.tt/gnTuL9N

How states can fight climate change without the feds

In the absence of an ambitious federal climate strategy, a new study shows state-led action can make a significant difference in reducing carbon emissions and addressing climate change.

The study also found that while state-led action is only slightly more expensive than a coordinated national effort, it would likely result in the adoption of different decarbonization technologies.

“Given that there is little expectation the Trump administration will promote a national effort to reduce greenhouse gas emissions to address climate change, we think there is significant value in assessing what kind of difference state-led efforts could make,” says Jeremiah Johnson, corresponding author of the study and an associate professor of civil, construction, and environmental engineering at North Carolina State University.

“For this study, we looked at a combination of 23 states that, based on political and policy indicators, seem most likely to consider joint action to reduce carbon emissions.

“Specifically, we looked at what the cost of such an effort would likely be, which decarbonization technologies would likely be adopted, and the extent to which these efforts could reduce our carbon footprint—and we compared all of these things to the cost, technology, and impact of a coordinated federal effort.”

The researchers drew on publicly available data across the full energy system for all 48 contiguous states, including everything from power generation and transportation to building operation and consumer needs, such as heating and cooling. This data was then fed into existing decarbonization models that were adapted to allow users to look at the impact of changes in individual states.

“We first looked at what the costs and technologies would be if the 23 states that already seemed inclined to strive for net zero carbon emissions actually achieved it,” says Gavin Mouat, first author of the paper and a former graduate student at NC State.

“That would reduce US carbon emissions by about 46% by 2050. We then looked at what the costs and technologies would be if all 48 contiguous states worked together to achieve that same 46% reduction.”

The researchers found costs were closer than anticipated between state-led and federal efforts; there was only a 0.7% difference in overall cost. However, the technologies adopted to reach the carbon emissions target were very different.

“That’s because different states have different resources,” Johnson says. “For example, some Great Plains states are excellent locations for establishing wind farms but are less likely to participate in a state-led initiative to address climate change.”

For example, in a state-led scenario, researchers found that industrial decarbonization—such as cleaner manufacturing technologies—played a far more prominent role than would be seen in a federally coordinated effort. On the other hand, a federally coordinated effort would rely more on clean energy production, such as wind and solar power generation.

The researchers also found there was potential for a state-led effort to affect climate-related pollution in neighboring states.

“Essentially, our model suggests it is possible that non-participating states could increase greenhouse gas emissions, because they might produce a product or service more cheaply for export to those states working to reduce their emissions,” Johnson says.

“However, the model also suggests that non-participating states might also reduce their greenhouse gas emissions. This could be due to the fact that clean energy technologies save them money, or because those states may be drawing power from power generation facilities in other states where emissions are falling.

“Ultimately the most important takeaway here is that state-led action can achieve substantial emission reductions, even without federal support, but that the world looks very different than one where there is federal coordination,” Johnson says.

“This has some important implications, not just for those states that choose to participate, but also for those who don’t.”

The paper appears in the journal Nature Communications.

Additional coauthors are from NC State and Carnegie Mellon University.

Support for this work came from the Alfred P. Sloan Foundation.

Source: North Carolina State University

The post How states can fight climate change without the feds appeared first on Futurity.

from Futurity https://ift.tt/Kyc8Vls

Text-generating AI models have different writing styles

A new study finds that large language models have distinctive styles.

It’s not unusual for people to have distinctive speech or writing styles. They can favor certain words and phrases or structure a sentence or a story uniquely.

It turns out that text-generating AI models have similar idiosyncrasies.

In a recent study, Carnegie Mellon University researchers found they could use characteristic word choices to determine which large language model (LLM) generated a particular bit of text with 97% accuracy.

“It was quite surprising that we achieved this level of accuracy,” says Mingjie Sun, a PhD student in the computer science department.

Sun and the other researchers achieved only 60% to 70% accuracy when trying to distinguish between just two LLMs on their own. The team’s specialized classifier program achieved superior results despite tackling a much more demanding task—differentiating between not two, but five LLMs: ChatGPT, Claude, Gemini, Grok, and DeepSeek.

The computer analysis revealed distinct profiles for each LLM. ChatGPT, for instance, tended to offer detailed, explanatory texts, while Claude favored concise, straightforward answers.

These idiosyncrasies aren’t superficial but deeply embedded in each model. Even when texts were scrambled, rephrased, translated, or summarized, the personality or style of each LLM remained distinct.

One implication of these findings is the need for caution when using synthetic data—text generated by LLMs—to train a new generation of models. This practice could pass the source model idiosyncrasies to the next generation of LLMs, potentially affecting the behavior of these AIs in unforeseen ways.

Zico Kolter, professor and director of CMU’s machine learning department, says that while using synthetic data for training was once a widespread method, its use has been on the decline.

One thing that the researchers did not do was try to discern the difference between AI-generated and human-generated text. Such an ability might be used in ferreting out academic fraud, but other research groups have already done a great deal of work on this topic. So, Sun, Kolter, and colleagues at the University of California, Berkeley; the University of Pennsylvania; and Princeton University focused their research on obtaining a deeper scientific understanding of LLMs.

“This work is much more about understanding the distinctive characteristics, the natures of different LLMs, the same way we think about different styles of writing by people,” Kolter says.

“Given how much content is being produced by LLMs on the internet these days, it is valuable to understand the distinguishing characteristics of these various models.”

This pre-print paper has not undergone peer review and its findings are preliminary.

Source: Carnegie Mellon University

The post Text-generating AI models have different writing styles appeared first on Futurity.

from Futurity https://ift.tt/KVExu4q

Can a pea-sized part of the brain help treat addiction?

Researchers are investigating using a pea-sized node in the brain to potentially treat drug addiction.

Major scientific discoveries can arise from simple decisions: say, by simply looking where virtually no one else has. Such was the case for Ines Ibañez-Tallon, a research associate professor in the Laboratory of Molecular Biology at The Rockefeller University, who over the past decade has revealed how one small, understudied region of the brain plays an outsized role in addiction and substance abuse.

It’s work that’s sparked a federally funded search for new medications that may help people beat chemical dependencies.

Known as the habenula, this narrow strip of gray and white matter—so tiny it’s considered a microstructure—is an ancient piece of the brain, first appearing in vertebrates about 360 million years ago.

Digging deep into this little node, Ibañez-Tallon uncovered an extremely complex and highly connected command center—one that’s part finely tuned sensor and part lightning-fast switchboard, detecting and sending chemical signals to other brain regions, including those that produce pleasure-inducing and modulatory neurotransmitters such as dopamine, acetylcholine, serotonin, and norepinephrine.

She’s also documented that the habenula helps regulate emotional states and cognitive behaviors, including motivation, disappointment, depression, and stress.

In addition to identifying a potential drug target that could directly address opioid addiction, her insights also point to how positive behaviors could boost healthy reward responses.

Here, she explains how she brought this little-known brain region to light:

The post Can a pea-sized part of the brain help treat addiction? appeared first on Futurity.

from Futurity https://ift.tt/QqN7MPu

‘MagicTime’ makes text-to-video AI more realistic

Using time-lapse videos as training data, computer scientists have developed video generators that simulate the physical world more accurately.

While text-to-video artificial intelligence models like OpenAI’s Sora are rapidly metamorphosing in front of our eyes, they have struggled to produce metamorphic videos.

Simulating a tree sprouting or a flower blooming is harder for AI systems than generating other types of videos because it requires the knowledge of the physical world and can vary widely.

But now, these models have taken an evolutionary step.

Computer scientists at the University of Rochester; Peking University; University of California, Santa Cruz; and National University of Singapore developed a new AI text-to-video model that learns real-world physics knowledge from time-lapse videos.

The team outlines their model, MagicTime, in a paper in IEEE Transactions on Pattern Analysis and Machine Intelligence.

(Credit: University of Rochester GIF created using MagicTime)

“Artificial intelligence has been developed to try to understand the real world and to simulate the activities and events that take place,” says Jinfa Huang, a PhD student supervised by Professor Jiebo Luo from the University of Rochester’s computer science department, both of whom are among the paper’s authors.

“MagicTime is a step toward AI that can better simulate the physical, chemical, biological, or social properties of the world around us.”

Previous models generated videos that typically have limited motion and poor variations. To train AI models to more effectively mimic metamorphic processes, the researchers developed a high-quality dataset of more than 2,000 time-lapse videos with detailed captions.

Currently, the open-source U-Net version of MagicTime generates two-second, 512 -by- 512-pixel clips (at 8 frames per second), and an accompanying diffusion-transformer architecture extends this to ten-second clips. The model can be used to simulate not only biological metamorphosis but also buildings undergoing construction or bread baking in the oven.

But while the videos generated are visually interesting and the demo can be fun to play with, the researchers view this as an important step toward more sophisticated models that could provide important tools for scientists.

“Our hope is that someday, for example, biologists could use generative video to speed up preliminary exploration of ideas,” says Huang.

“While physical experiments remain indispensable for final verification, accurate simulations can shorten iteration cycles and reduce the number of live trials needed.”

Source: University of Rochester

The post ‘MagicTime’ makes text-to-video AI more realistic appeared first on Futurity.

from Futurity https://ift.tt/DkpEN2r

Why are city kids more prone to allergies?

Researchers have discovered that a previously uncharacterized subset of immune cells may play a critical role in the development of allergic diseases and explain differences between urban and rural populations.

The finding in the journal Allergy provides new insight into how the immune system is shaped in early life—and why urban children are more prone to allergies than children from rural areas.

Led by researchers from the University of Rochester Medical Center (URMC) pediatrics department, including MD/PhD student Catherine Pizzarello and senior author Kirsi Järvinen-Seppo, the study uncovered a unique subpopulation of T cells known as helper 2 (Th2) cells with distinct molecular characteristics.

T-cells are the foundational immune cells that fight off infections, but there is evidence that this specific subtype is recognizing certain foods as allergenic and attacking them, according to Jarvinen-Seppo.

“These pro-allergic T cells are more inflammatory than anything previously described in this context,” says Järvinen-Seppo, chief of Pediatric Allergy and Immunology at UR Medicine Golisano Children’s Hospital.

“They were found more frequently in urban infants who later developed allergies, suggesting they may be a predictive biomarker or even a mechanistic driver of allergic disease.”

The study compared blood samples from urban infants with those from infants in a farming community, specifically the Old Order Mennonites (OOM) of New York’s Finger Lakes region—known for their low rates of allergies. Researchers found that while urban infants had higher levels of the aggressive Th2 cells, OOM infants had more regulatory T cells that help keep the immune system in balance and reduce the likelihood of allergic responses.

While additional research is needed to identify a possible cause, Jarvinen-Seppo speculates that differences in the development of the gut microbiome between the two populations, and more exposure to “healthy” bacteria in rural children, may be a factor.

“The farming environment, which is rich in microbial exposure, appears to support the development of a more tolerant immune system. Meanwhile, the urban environment may promote the emergence of immune cells that are primed for allergic inflammation,” says Jarvinen-Seppo.

The work is part of a broader, NIH-funded investigation into how early-life exposures influence long-term immune outcomes.

The goal is to continue this foundational work to uncover protective factors that could be translated into preventive therapies, including probiotics or microbiome-supporting interventions.

“If we can identify the conditions for this disparity between the different T cell subpopulations, we can potentially find solutions in allergic disease development,” Järvinen-Seppo says.

Source: University of Rochester

The post Why are city kids more prone to allergies? appeared first on Futurity.

from Futurity https://ift.tt/xfmEnCP

Watch: Legless robot can jump 10 feet high

Inspired by the movements of a tiny parasitic worm, engineers have created a 5-inch soft robot that can jump as high as a basketball hoop.

Their device, a silicone rod with a carbon-fiber spine, can leap 10 feet high even though it doesn’t have legs. The researchers made it after watching high-speed video of nematodes pinching themselves into odd shapes to fling themselves forward and backward.

The research appears in Science Robotics.

The researchers say their findings could help develop robots capable of jumping across various terrain, at different heights, in multiple directions.

“Nematodes are amazing creatures with bodies thinner than a human hair,” says Sunny Kumar, lead coauthor of the paper and a postdoctoral researcher in the School of Chemical and Biomolecular Engineering (ChBE) at Georgia Tech.

“They don’t have legs but can jump up to 20 times their body length. That’s like me laying down and somehow leaping onto a three-story building.”

Nematodes, also known as round worms, are among the most abundant creatures on Earth. They live in the environment and within humans, other vertebrates, and plants. They can cause illnesses in their host, which sometimes can be beneficial. For instance, farmers and gardeners use nematodes instead of pesticides to kill invasive insects and protect plants.

One way they latch onto their host before entering their bodies is by jumping. Using high-speed cameras, Victor Ortega-Jimenez—a lead author and former Georgia Tech research scientist who’s now a faculty member at the University of California, Berkeley—watched the creatures bend their bodies into different shapes based on where they wanted to go.

“It took me over a year to develop a reliable method to consistently make these tiny worms leap from a piece of paper and film them for the first time in great detail” Ortega-Jimenez says.

To hop backward, nematodes point their head up while tightening the midpoint of their body to create a kink. The shape is similar to a person in a squat position. From there, the worm uses stored energy in its contorted shape to propel backward, end over end, just like a gymnast doing a backflip.

To jump forward, the worm points its head straight and creates a kink on the opposite end of its body, pointed high in the air. The stance is similar to someone preparing for a standing broad jump. But instead of hopping straight, the worm catapults upward.

“Changing their center of mass allows these creatures to control which way they jump. We’re not aware of any other organism at this tiny scale that can effectively leap in both directions at the same height,” Kumar says.

And they do it despite nearly tying their bodies into a knot.

“Kinks are typically dealbreakers,” says Ishant Tiwari, a ChBE postdoctoral fellow and lead coauthor of the study. “Kinked blood vessels can lead to strokes. Kinked straws are worthless. Kinked hoses cut off water. But a kinked nematode stores energy that is used to propel itself in the air.”

After watching their videos, the team created simulations of the jumping nematodes. Then they built soft robots to replicate the leaping worms’ behavior, later reinforcing them with carbon fibers to accelerate the jumps.

Kumar and Tiwari work in Associate Professor Saad Bhamla’s lab. They collaborated on the project with Ortega-Jimenez and researchers at the University of California, Riverside.

The group found that the kinks allow nematodes to store more energy with each jump. They rapidly release it—in a tenth of a millisecond—to leap, and they’re tough enough to repeat the process multiple times.

The study suggests that engineers could create simple elastic systems made of carbon fiber or other materials that could withstand and exploit kinks to hop across various terrain.

“A jumping robot was recently launched to the moon, and other leaping robots are being created to help with search and rescue missions, where they have to traverse unpredictable terrain and obstacles,” Kumar says.

“Our lab continues to find interesting ways that creatures use their unique bodies to do interesting things, then build robots to mimic them.”

Support for the work came from the National Institutes of Health and the National Science Foundation. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the view of any funding agency.

Source: Georgia Tech

The post Watch: Legless robot can jump 10 feet high appeared first on Futurity.

from Futurity https://ift.tt/QqZ1ROX