EurekAlert! In the Spotlight - Artificial Intelligence

This month, we’re focusing on artificial intelligence (AI), a topic that continues to capture attention everywhere. Here, you’ll find the latest research news, insights, and discoveries shaping how AI is being developed and used across the world.

Researchers create multimodal sentiment analysis method that improves detection of human emotions while reducing computational cost

Research
Peer-Reviewed Publication

Multimodal sentiment analysis is an information processing technique that attempts to predict human emotional states from multiple modalities like text, audio, and video. Due to challenges in aligning multiple modalities, existing methods are limited to analysis at course or fine granularity, which risks missing nuances in human emotional expression. Researchers have now developed an innovative approach to MSA that reduces computational time required to sentiment prediction while offering improved performance.

Journal: Research
Funder: National Natural Science Foundation of China, Sichuan Provincial Natural Science Foundation, Fundamental Research Funds for the Central Universities

14-Aug-2025

Deep learning model successfully predicted ignition in inertial confinement fusion experiment

American Association for the Advancement of Science (AAAS)
Peer-Reviewed Publication

Brian Spears and colleagues built a generative machine learning model that was used to successfully predict the outcome of a recent fusion ignition experiment at the U.S. National Ignition Facility (NIF). Their model predicted, with a probability greater than 70%, that ignition was the most likely outcome of the experiment. The findings could guide researchers working on future inertial confinement fusion experiments, which use energetic lasers to compress and heat a capsule of hydrogen isotopes to create nuclear reactions that produce fusion energy. Ignition refers to the outcome where the fusion energy produced exceeds the laser energy used to conduct the experiment – a feat that NIF researchers accomplished in a small experiment in 2022. This success was predicted by the generative learning model developed by Spears et al., which was built using experimental data, radiation hydrodynamics simulations, and Bayesian statistics. Having a successful prediction model will provide swift guidance to fusion energy researchers as they modify experimental designs and determine whether future upgrades in laser energy and other variables might improve fusion output and efficiency, the authors note.

Journal: Science

14-Aug-2025

Maternal antibodies in breast milk regulate early immune responses in mouse gut

American Association for the Advancement of Science (AAAS)
Peer-Reviewed Publication

The first weeks outside the womb are a massive learning experience for the neonatal immune system, especially in the gut as it learns to distinguish helpful from harmful antigens. In mice, maternal antibodies ingested in breast milk in the first week after birth help to regulate immune responses in the newborn gut, according to a new study by Meera Shenoy and colleagues. This process builds the important partnership between the microbiome and host at an early stage, paving the way for nutrient assimilation, avoiding inappropriate inflammation and resistance against pathogens. Shenoy et al. found that mouse maternal immunoglobin G antibodies bind to bacteria in the neonatal gut to modulate the immune response in the gut mucosa. This antibody-bacteria complex binds to antibody receptors in the gut to limit the response of T cells to food antigens during weaning and in response to experimentally induced colitis during early life. “Translating these findings to human infants will require studies integrating information on the timing of exposure to maternal antibodies, along with the development of key immune components and the intestinal microbiota,” Michael Silverman and Kathryn Knoop write in a related Perspective.

Journal: Science

14-Aug-2025

Building a broader cybersecurity ecosystem

University of Pittsburgh
Meeting Announcement

On Tuesday, August 12, 2025, the University of Pittsburgh Cyber Energy Center and Pitt Cyber hosted “Transforming Cybersecurity: A Multidisciplinary Approach to Risk, Technology, and Policy.” The in-person, day-long workshop brought together experts from across industries and disciplines to assess the current state of cybersecurity through a multidisciplinary lens.

14-Aug-2025

UH researchers to develop AI to aid in emergency food distribution

University of Houston
Grant and Award Announcement

Families, particularly those already vulnerable to food insecurity, can face difficulties obtaining food in the aftermath of natural disasters. University of Houston researchers will utilize artificial intelligence to develop an online resource for food pantries, aiming to streamline stakeholder collaboration and distribute resources to families in need.

14-Aug-2025

Texas A&M researchers map America’s power outage hot spots using AI

Texas A&M University
Peer-Reviewed Publication

Researchers from the Urban Resilience AI Lab at Texas A&M University have used machine learning to create a nationwide Power System Vulnerability Index (PSVI) that identifies areas at increased risk of power outages.

Journal: Applied Energy
Funder: U.S. National Science Foundation

14-Aug-2025

SETI Institute awards Davie Postdoctoral Fellowship for AI/ML-driven exoplanet discovery

SETI Institute
Grant and Award Announcement

The SETI Institute awarded the Davie Postdoctoral Fellowship for AI/ML-driven exoplanet discovery to Isabel Angelo. Machine learning is changing the way we search for exoplanets and making it possible to discover patterns in massive datasets. Angelo’s research will refine and expand ML-driven pipelines for detecting exoplanets, and she will work with SETI Institute researcher Dr. Vishal Gajjar and his team and collaborators at the SETI Institute and IIT Tirupati in India.

The project will enhance supervised Convolutional Neural Network (CNN) architectures and integrate anomaly-detection techniques to identify subtle or unconventional exoplanet candidates hidden in massive datasets. These could include ringed or disintegrating worlds, exocomets, complex multi-planet systems and possibly signs of alien megastructures.

Welcome to In the Spotlight, where each month we shine a light on something exciting, timely, or simply fascinating from the world of science.

Latest News Releases

Researchers create multimodal sentiment analysis method that improves detection of human emotions while reducing computational cost

Deep learning model successfully predicted ignition in inertial confinement fusion experiment

Maternal antibodies in breast milk regulate early immune responses in mouse gut

Building a broader cybersecurity ecosystem

UH researchers to develop AI to aid in emergency food distribution

Texas A&M researchers map America’s power outage hot spots using AI

SETI Institute awards Davie Postdoctoral Fellowship for AI/ML-driven exoplanet discovery