For decades, bird flu was uncommon in the U.S., but that has changed. In the past several years, epidemics have threatened poultry and dairy cattle operations across the country. Higher egg prices, driven largely by bird flu-related supply disruptions, have cost American consumers billions of dollars in losses.
“The H5N1 strain of the bird flu, which has driven recent and current outbreaks, is a highly lethal virus that kills domestic chickens and other bird species in droves,” said David Pattie, GTRI research scientist and branch chief. “It can easily jump from birds to other animal species — and sometimes to humans.”
The research team will leverage artificial intelligence (AI) to design and test a probiotic avian flu vaccine that, if successful, could be served to chickens in their feed. Currently, vaccinating a flock means individually injecting every bird.
“We’re focusing on live bacterial vaccines, which means the vaccine comes from living bacteria you swallow, instead of an injection,” said Mike Farrell, GTRI principal research scientist and the project’s lead investigator.
“These probiotic vaccines would help protect birds and livestock from flu-like infections and lower the risk of those viruses spreading to humans,” he added.
In addition to Farrell and Pattie, the team includes researchers from an array of disciplines across the Institute: Faramarz Fekri, professor and John Pippin Chair in the School of Electrical and Computer Engineering; JC Gumbart, Dunn Family Professor in the School of Physics; Brian Hammer, associate professor in the School of Biological Sciences; and Anton Bryskin, director of the Molecular Evolution Core at the Parker H. Petit Institute for Bioengineering and Bioscience.
The project builds on Farrell’s ongoing research into developing probiotic vaccine adjuvants for human influenza. The goal is to use probiotic bacteria — the “good bacteria” found in foods like yogurt — to help create immunity for the flu vaccine.
If the researchers can get probiotic bacteria to display pieces of the flu virus (called antigens) on their surface, then they could be swallowed like a normal probiotic pill.
“The gut is a great place for building immunity. When these bacteria reach the gut, your body would recognize the virus pieces on the bacteria and start building flu antibodies,” Farrell explained. “That way, when the chickens get exposed to flu, their immune system would already be prepared to fight it.”
“The idea behind this oral bird flu vaccine is to leverage artificial intelligence and the vast historical database for H5N1 available to us, because it's a very well-studied virus,” Farrell said. “There is a ton of structural data out there.”
Gumbart is an expert in protein modeling and simulation. Part of his role is figuring out the best design for a viral protein piece (antigen) — one that looks and behaves like the real virus protein, so it triggers the right immune response. To do this, he will combine Fekri’s AI-generated predictions with computer modeling.
“That’s where my team adds real value,” Gumbart said. “We use simulations to test how stable and realistic these protein designs are, which allows us to choose the best ones for lab experiments.”
AI has already identified new medicines and antibiotics by studying chemical databases. If the team can use AI to help design virus proteins for vaccines, it could transform how vaccines are made.
Pattie says that any viral infectious disease with a high mortality rate has the potential to become a national security threat. “At that point, developing countermeasures becomes exceedingly important from a national security perspective,” he said.
This is the first time several of the team members are working on poultry research. For Gumbart, the project is a full-circle moment.
“I grew up in rural Illinois, and as a kid, one of my daily chores was to take care of chickens, and I kind of hated it,” he said. “It is some sort of universal irony that I am back to taking care of chickens again.”
]]>The questions surrounding AI may be an unprecedented new challenge, but at Georgia Tech, students are already trying to answer them. The AI Safety Initiative (AISI) is a student group aiming to steer AI research and policy for society’s benefit.
“AI introduces new kinds of challenges into our legal and societal frameworks,” said Rocio Perales Valdes, AISI co-director and second-year computer science student. “Its capabilities emerge fast and on a jagged, hard-to-predict edge, which leaves AI governance like chasing a moving target. The work ahead is building the governance and technical tools we need to evaluate these systems, set direction, and enforce them without hindering innovation.”
AISI focuses on developing and deploying AI responsibly, rather than avoiding it. The group offers guest talks from AI researchers, fellowships that immerse students in the latest safety research through reading and discussion groups, and independent projects that contribute directly to the field. Past projects from AISI include demonstrating large language model security risks on Capitol Hill, responding to U.S. Federal Requests for Information, and running a war game for GTRI faculty. Part lab and part learning community, AISI prepares students to become the next generation of AI safety researchers and practitioners. They have placed alumni at leading organizations such as Anthropic, RAND, Model Evaluations and Threat Research, the UK AI Security Institute, and the Horizon Institute for Public Service.
“AI safety is an urgent problem because there is a rapidly growing gap between what AI systems can do and what we understand about them; yet mitigating AI risks is systematically neglected by current market incentives,” said Yixiong Hao, third‑year computer science student and co‑director of AISI. “I think the set of things I can do to directly move the needle is quite limited in the next three to five years, and that’s why I run this group. I have higher leverage in convincing smart people to work on neglected problems in AI safety.”
Founded in 2022 by Gaurav Sett, who is now a Ph.D. student at the RAND School of Public Policy and a fellow at the Institute for Progress, AISI has grown quickly. Its 10‑member executive board supports a broad base of student involvement, with more than 70 students participating in the fellowship program each semester. Over the past two years, members have also published 13 papers at top conferences such as the International Conference on Learning Representations, with projects spanning AI security and algorithmic transparency.
From Discussion to Discovery
As a first‑year computer science student, Ishan Khire joined AISI looking for a deeper way to engage with AI safety and quickly found a pathway into research. After attending one general meeting, Khire enrolled in the group’s six‑week fellowship program, where students meet weekly to discuss current technical and policy challenges shaping the field.
“Finding a community that cares about AI safety was a big part of joining the fellowship,” Khire said. “Because AI safety is a broad subject, it was helpful to have an accountability group to discuss current issues.”
Thanks to the connections he made at AISI, Khire began conducting AI research with computing faculty member Giri Krishnan to predict the 3D structure of proteins.
“AI is going to be really transformative in the next five to 10 years, and we want to make that transformation go well,” Khire said. “AISI tries to upskill people and connect them to technical and policy research that helps them find impactful work.”
Student Advantage
AISI is entirely student‑run, with a small group of faculty advisors. That structure lends itself to uncertain research that can be difficult to fund through traditional academic labs, and faculty support has followed.
“Any cursory look at the news today will show there is significant angst about AI and whether it is being developed responsibly and with sufficient guardrails in place,” said Tom Conte, the College of Computing associate dean for Research. “AISI has Georgia Tech at the forefront of that conversation.”
AISI member and computer science Ph.D. student Glenn Matlin has recruited many undergraduate researchers from the group for his own projects.
“I consider AISI like a third lab,” he said. “I use it as a great place for recruiting students. I’m constantly sharing my own research, and it helps me stay up to date with what other researchers are talking about.”
Matlin also credits AISI with advancing his own work in AI safety. Through the fellowship, he synthesized research that helped him apply for opportunities such as the prestigious AI safety mentorship at the MATS Program, which has connected him to additional research funding.
In a future increasingly shaped by algorithms, AISI’s students are betting that the most important safeguards won’t come from code alone, but from the people guiding how that code is built, deployed, and governed.
“AI safety matters to everyone,” Matlin said. “AI is going to disrupt not just technology, but also politics and business — and its risks are creating urgent opportunities to make it safer.”
]]>
tess.malone@gatech.edu
]]>
The award was announced March 25, 2026 and is one of six Institute Research Awards given by Georgia Tech’s Office of the Executive Vice President for Research. The portfolio of awards honors achievements in research engagement, innovation, faculty advising, and impact.
Georgia AIM is a statewide coalition led by the Georgia Tech Enterprise Innovation Institute (EI2) and the Georgia Tech Manufacturing Institute (GTMI) to develop and deploy AI talent and innovation in manufacturing. The Georgia AIM coalition includes dozens of universities, technical colleges, nonprofits, and economic development organizations.
“It is an incredible experience to collaborate with technology and economic development leaders around the state to lead the nation and the world in AI for manufacturing,” said Aaron Stebner, Georgia AIM co-director and the Eugene C. Gwaltney Jr. Chair in Manufacturing at Georgia Tech.
“We are truly honored to receive this recognition from our peers at Georgia Tech,” said Tom Kurfess, GTMI Executive Director and HUSCO/Ramirez Distinguished Chair in Fluid Power and Motion Control.
Georgia AIM was initiated in 2021 by Stebner, EI2 Vice President David Bridges, Kurfess, Georgia AIM managing director and GTMI deputy director Steven Ferguson, and Georgia Tech executive director for strategic partnerships George White. The coalition received an initial $500,000 planning grant from the U.S. Economic Development Administration (EDA), which was followed by $65 million in additional grants from EDA and with additional federal, state, and private sector support now totals more than $100 million to enact projects across the state.
The Georgia AIM coalition counts many achievements on and off campus, including:
“Georgia AIM proves that innovation scales when built alongside workforce,” said Ferguson. “We built a seamless pipeline from education to industry, ensuring talent is ready to deploy AI in real manufacturing environments on day one.”
“The impact of Georgia AIM is grounded in collaboration — universities, industry, nonprofits and communities working together to shape the future of advanced manufacturing in Georgia,” said Bridges. “This recognition underscores what a coordinated statewide effort can accomplish.”
Because research covers a range of activities — from research and development to commercialization and public impacts — the annual awards recognize the many facets of work in this area. The peer-driven nomination process emphasizes measurable contributions and leadership across disciplines.
“The strength of Georgia Tech’s research enterprise begins with the talented people who push discovery forward every day,” said Tim Lieuwen, executive vice president for Research. “Congratulations to this year’s honorees, who demonstrate what it means to turn bold ideas into real-world impact, advancing knowledge from fundamental science to commercial and community applications. With these awards, we celebrate their leadership, creativity, and dedication to serving the public good.”
Read more about this year’s Institute Research Award winners.
]]>It was the beginning of the Covid-19 pandemic.
The world needed a vaccine that didn’t exist, and there was no clear timeline for one. No one knew how long the vaccine development process would take — or whether it would work at all.
Then, less than a year later, Pfizer and BioNTech set a record for how fast a drug moved from clinical trials to federal authorization — and to people waiting as the virus surged worldwide. That speed depended on more than scientific discovery. It hinged on trials, regulatory approval, and manufacturing at scale.
Startup BioNTech, a small biotech firm, had spent years developing mRNA technology. Pfizer, a huge pharmaceutical company, brought deep experience running large clinical trials, working with regulators, and manufacturing at scale. The two companies had worked together before, which meant they did not have to build trust, decision-making structures, or workflows in the middle of a crisis. Trials moved quickly. They knew what regulators required and how to meet those demands.
According to Georgia Tech research, that kind of business alignment is far from common — and can explain why many promising drugs never reach patients.
Manpreet Hora, senior associate dean for programs and professor of operations management in Georgia Tech’s Scheller College of Business, studies what happens after a drug leaves the lab. In a study published in Production and Operations Management, he and his coauthors analyzed nearly 300 biotech–pharma partnerships to understand why some drugs make it through and others stall.
“If you are a patient, this process is out of your control,” Hora said. “In some cases, it can cost lives.”
Drug development often depends on handoffs. Small biotech firms typically generate early discoveries. Larger pharmaceutical companies step in to run trials, work with regulators, and bring products to market.
But complications can arise when companies that lack similar experience levels try to develop the drug together.
Decision-making slows down. Roles become unclear. The process starts to erode.
"That's why partner choice matters," Hora said, comparing the process to a popular TV show. "It's like going on Shark Tank — just because someone is offering money doesn't mean they're the right partner."
Hora said the Pfizer–BioNTech partnership worked because both companies approached the work the same way, despite the difference in their size. Pfizer is one of the largest pharmaceutical companies in the world. BioNTech was a much smaller firm.
As of September 2025, 5 billion doses of the Pfizer–BioNTech Covid vaccine have been distributed globally.
Pfizer’s chairman and CEO, Albert Bourla, attributes the unprecedented success to a “world class collaboration” with BioNTech. He said, "I think it was because both companies had developed very similar cultures…We were both really very purpose-driven.”
Hora's research comes to the same conclusion: In an industry where drugs can take a decade to reach patients, the wrong partner can mean they never arrive at all.
]]>Manpreet Hora, a professor at Georgia Tech's Scheller College of Business, analyzed nearly 300 biotech–pharma partnerships and found that mismatched partners, such as companies with different experience levels, cultures, or decision-making styles, stall drug development. Roles blur. Trials drag. Promising treatments never reach patients.
The Pfizer–BioNTech Covid vaccine is the counterexample: two companies, vastly different in size, aligned on purpose and process. That alignment is what made speed possible.
Hora's takeaway: in an industry where drugs take a decade to reach patients, the wrong partner can mean they never arrive at all.
]]>Working with collaborators at University College London (UCL), Georgia Tech researchers used computer models to reveal how, during early development, forces generated by cells physically pull the neural tube closed — like a drawstring. This discovery offers new insight into a critical process that, when disrupted, can result in severe birth defects such as spina bifida.
“Understanding a complex developmental process like neural tube closure requires a highly interdisciplinary approach,” said Shiladitya Banerjee, an associate professor in the School of Physics. “By combining advanced biological imaging with theoretical physics, we were able to uncover the mechanical rules that drive cells to close the tube. My lab builds computational models to uncover the physical rules of living systems. The neural tube is an ideal focus because its formation requires incredible mechanical coordination.”
The researchers presented their findings in Current Biology.
The UCL team studied mouse embryos, which develop similarly to humans, and Georgia Tech researchers used that data to construct their models. From the data, they identified the fundamental physics mechanism that enables neural tube closure in part of the brain. This mechanism, called a “purse string,” is made of actin, a pivotal protein that forms a cell’s skeletal structure. As the purse strings tighten, the tube closes.
“These actin molecules are very important because they give rigidity and shape to cells,” Banerjee said.
“During neural tube closure, actin filaments form a ring around the opening and engage molecular motors — proteins that generate forces inside cells,” he said. “As these motors pull on the actin, they generate tension that tightens the ring and draws the tube closed.”
As the actin ring tightens, cells stretch and elongate, causing them to align and move together in a synchronized pattern, like a school of fish. This coordination allows the cells to move faster and more efficiently, increasing tension and driving a feedback loop that helps seal the neural tube.
The team built a computer model to show how this feedback loop leads to successful neural tube formation. Further research using the model could help explain why the neural tube fails to close.
“Physics-based modeling of cell and tissue mechanics allows us to connect the dots between developmental stages in a way that is both robust and quantitative, simulating experiments that are impossible in biological tissues,” said Gabriel Galea, the study co-author and UCL group leader. “In this case, it allowed us to explain how a cell’s mechanical experience impacts its current and future shapes during a critical step of brain development.”
Beyond neural tube development, the findings highlight the power of physics-based modeling to explain complex biological processes that can’t be observed directly. The researchers say this approach could be applied to other stages of human development where forces, motion, and timing are just as critical.
The computational research at Banerjee Lab is funded by the National Institute of General Medical Sciences
Fernanda Pérez-Verdugo, Eirini Maniou, Gabriel L. Galea, Shiladitya Banerjee, “Mechanosensitive feedback organizes cell shape and motion during hindbrain neuropore morphogenesis,” Current Biology, 2026.
DOI: 10.1016/j.cub.2026.02.068
]]>
The paper, The Borderlands of Foldability: Lessons from Simplified Proteins, is a meta-analysis of six decades of protein research and reveals that ancient proteins may have been far more complicated and dynamic than previously thought.
Recently published in the journal Trends in Chemistry, the study includes Georgia Tech researchers Lynn Kamerlin, professor in the School of Chemistry and Biochemistry and Georgia Research Alliance Vasser-Woolley Chair in Molecular Design, and Quantitative Biosciences Ph.D. candidate Alfie-Louise Brownless.
Co-authors also include Institute of Science Tokyo graduate student Koh Seya and Liam M. Longo, who serves as a specially appointed associate professor at Science Tokyo and as an affiliate research scientist at the Blue Marble Space Institute of Science.
The research has implications ranging from the origins of life and the search for life in the universe to cutting-edge medical innovation. “One of the biggest unanswered questions in science is how life first began,” says Kamerlin, who is a corresponding author of the study. “Understanding how the first protein-like molecules formed and what the earliest proteins may have been like is a key part of that puzzle.”
“Proteins power our bodies — and all life on Earth,” she adds. “Simply put, the evolution of proteins is the reason that we’re able to have this conversation at all.”
If proteins are the scaffolding of life, amino acids are the components that make up that scaffolding. “Today, an average protein is constructed from a chain of about 300 amino acids, involving 20 different types of amino acids,” Kamerlin shares. Proteins fold when these chains twist into a specific 3-dimensional shape, creating structures critical for biology.
However, while these folds are essential, exactly how a protein knows which way to fold remains a mystery. “We know that proteins didn’t just fold randomly,” Kamerlin shares, “because randomly trying all possible configurations would take a protein longer than the age of the universe.”
It’s a cornerstone problem in biological science called “Levinthal’s Paradox,” and highlights a fundamental mystery: Proteins fold incredibly quickly into very specific combinations — but like a sheet of paper spontaneously folding into an origami swan, researchers don’t know how proteins “choose” the folds they make.
“We can predict what a protein will look like, but can’t tell you how it got there,” Kamerlin adds. “That’s what we’re interested in exploring: how small early proteins developed into the complex proteins that support every living thing on today’s Earth.”
Early proteins likely had access to just half of today’s amino acids. “About 10-12 amino acids were likely available on early Earth,” Kamerlin says. Like writing a story with just the letters “A” through “L,” researchers assumed that the ‘vocabulary’ proteins could build from such a limited amino acid alphabet would also be constrained.
“There is a language to protein folding,” Kamerlin explains. “That language is hidden in their structures. Our research is in trying to understand the rules — the grammar and vocabulary that dictate a protein fold.”
The grammar they discovered was surprising: with a combination of creative techniques and environmental support, complex structures can arise from limited amino acid alphabets.
“We found that it is possible to develop complex folds with very simple tools — and certain environments, like salty ones, can help support that,” Kamerlin shares. “Early proteins could also cross-link and associate, interacting like LEGO blocks to create more complex structures.”
Now, the team is conducting research in environments that could mimic conditions on early Earth — aiming to discover more about how these regions could have given rise to today’s complex proteins. “This aspect of our research also ties into the amazing space research happening at Georgia Tech,” Kamerlin says. “While we’re interested in understanding early life on Earth, our work could help inform where best to look for evidence of life beyond our planet.”
Kamerlin specializes in creating computer models that simulate possible scenarios – creating an opportunity to quickly and efficiently test many theories. The most compelling of these can then be tested by her collaborator and co-author at Science Tokyo, Liam Longo, in lab experiments.
Protein folding is also at the forefront of medical innovation, ranging from diagnostic tools to cancer treatments and neurodegenerative diseases. “In the broader scope, we’re interested in discovering what we can design, what we can stress test, and what we can reconstruct with AI and other computational tools,” Kamerlin says. “Because if you can understand how proteins fold, you gain the ability to design them.”
Funding: NASA, the Human Frontier Science Program, and the Knut and Alice Wallenberg Foundation
DOI: https://doi.org/10.1016/j.trechm.2026.03.001
]]>Selena Langner
College of Sciences
Georgia Institute of Technology
GT² is a newly established research initiative at Georgia Tech that focuses on discovery science, engineering innovation, and AI-enabled decision tools to address urgent challenges at the intersection of environmental and community resilience in the Southeast. The initiative fosters research in direct service to regional communities through public-private partnerships, and it provides opportunities for graduate student engagement.
The BHIC-GT² research fund and partnership will pursue shared initiatives in the fields of coastal sustainability, ecosystem health, and environmental resilience. By combining BHIC’s applied, field-based conservation work with Georgia Tech’s expertise in technological innovation and data analysis, new opportunities for impactful research will be created through graduate student projects and community engagement.
About the Partnership
Like the GT² initiative, BHIC’s Johnston Center for Coastal Sustainability was created to translate research into real-world impact. BHIC established the Johnston Center as a research partnership and education hub for sustainability initiatives on Bald Head Island, with the broader goal of advancing coastal sustainability across the Southeast. Seed funding for the Center was provided in 2021 by Dick and Pat Johnston, longtime supporters of BHIC.
Dick, a Georgia Tech IM 1962 alumnus, and Pat Johnston shared their enthusiasm for the BHIC and Georgia Tech collaboration, noting:
“We are delighted to see our two favorite institutions come together through this partnership. It brings additional resources, expertise, and leadership to our shared focus on keeping the historic tagline ‘Living in Harmony with Nature’ in the hearts of future generations.”
Joel Kostka, Faculty Director of GT² who also serves as Tom and Marie Patton Distinguished Professor and associate chair for Research in the School of Biological Sciences with a joint appointment in the School of Earth and Atmospheric Sciences at Georgia Tech added:
“The Bald Head Island Conservancy and its Johnston Center for Coastal Sustainability exemplify how place‑based conservation and rigorous science can work together to create real impact. The Bald Head Island Conservancy’s long‑term stewardship, research infrastructure, and commitment to translating science into action make it an ideal partner for Georgia Tech for Georgia’s Tomorrow as we advance collaborative research that strengthens coastal resilience across the Southeast.”
This partnership will focus on Georgia Tech graduate student research projects that use innovative technology and data analyses to directly support the conservation work of BHIC.
Graduate student research already plays an important role in BHIC’s conservation efforts. Gabie Krueger, a Georgia Tech Ph.D. student in Ocean Sciences and Engineering and BHIC’s 2025-26 Johnston Graduate Fellow in Coastal Sustainability, has been working with BHIC scientists on a salt marsh ecology project that examined how ribbed mussels and fiddler crabs influence the health of Bald Head Island’s dominant salt marsh grass Spartina alterniflora. These flora-fauna interactions serve as primary indicators of marsh health, so her research is important for understanding the resilience of Bald Head Island’s salt marsh to environmental concerns such as sea-level rise and development.
Through the BHIC-GT² partnership, Georgia Tech student researchers who work with the Conservancy will also gain invaluable experience with local conservation efforts and community engagement.
G. Christopher Shank, Ph.D., Executive Director of BHIC, commented:
“The Bald Head Island Conservancy is thrilled about this opportunity to create a formal research partnership with Georgia Tech, one of the nation’s most esteemed research universities. It is recognition of the quality of conservation studies we are currently pursuing at the Conservancy and it also augments the impact of our work for BHI and beyond because of the technological and data analysis talent that Georgia Tech for Georgia’s Tomorrow will bring to this partnership.”
Why This Matters
This research fund and partnership represents an important step forward in strengthening connections between academic research and applied conservation institutions. Together, BHIC and GT² aim to inform coastal management decisions, support resilience planning, engage students, and advance research that benefits coastal ecosystems and communities across the southeastern U.S.
Looking Ahead
Additional details about joint initiatives, research priorities, and collaborative opportunities will be shared in the coming months.
]]>
Chris Shank
Executive Director
Bald Head Island Conservancy
shank@bhic.org
That’s one of the main findings from a study by AI Caring on what older adults expect from explainable AI (XAI).
AI Caring is one of three AI Institutions led by Georgia Tech and funded by the National Science Foundation (NSF). The institution supports AI research that benefits older adults and their caregivers.
Niharika Mathur, a Ph.D. candidate in the School of Interactive Computing, was the lead author of a paper based on the study. The paper will be presented in April at the 2026 ACM Conference on Human Factors in Computing Systems (CHI) in Barcelona.
Mathur worked with the Cognitive Empowerment Program at Emory University to interview 23 older adults who live alone and use voice-activated AI assistants like Amazon’s Alexa and Google Home.
Many of them told her they feel excluded from the design of these products.
“The assumption is that all people want interactions the same way and across all kinds of situations, but that isn’t true,” Mathur said. “How older people use AI and what they want from it are different from what younger people prefer.”
One example she gave is that young people tend to be informal when talking with AI. Older people, on the other hand, talk to the agent like they would a person.
“If Older adults are talking to their family members about Alexa, they usually refer to Alexa as ‘she’ instead of ‘it,’” Mathur said. “They tend to humanize these systems a lot more than young people.”
The study evaluated AI explanations that drew information from four sources of data:
Mathur said older users trust the agent more when it bases its explanations on data from the first three sources. However, internal reasoning creates skepticism.
Internal reasoning means the AI doesn’t have enough data from the other sources to give an explanation. It provides a percentage to reflect its confidence based on what it knows.
“The overwhelming response was negative toward confidence scores,” Mathur said. “If the AI says it’s 92% confident, older adults want to know what that’s based on.”
This is another example that Mathur said points to generational preferences.
“There’s a lot of explainable AI research that shows younger people like to see numbers in explanations, and they also tend to rely too much on explanations that contain numerical confidence. Older adults are the opposite. It makes them trust it less.”
Mathur said that AI agents interacting with older adults should serve a dual purpose. They should provide users with companionship and support independence while reducing the caretaking burden often placed on family members.
Some studies have shown that engineers have tended to favor caretakers in the design of these tools. They prioritize daily tasks and routines, leaving some older adults to feel like they are merely a box to be checked.
She discovered that in urgent situations, older users prefer the AI to be straightforward, while in casual settings, they desire more conversation.
“How people interact with technological systems is grounded in what the stakes of the situation are,” she said. “If it had anything to do with their immediate sense of safety, they did not want conversational elaboration. They want the AI to be very direct and factual.”
Mathur said AI agents that interact with older adults are ideally constructed with a dual purpose. They should provide companionship and autonomy for the users while alleviating the burden of caretaking that is often placed on their family members.
Some studies have shown that engineers have strayed toward favoring caretakers in the design of these tools. They prioritize daily tasks and routines, leaving some older adults to feel like they are a box to be checked.
“They’re not being thought of as consumers,” Mathur said. “A lot of products are being made for them but not with them.”
She also said psychological well-being is one of the most important outcomes these tools should produce.
Showing older adults that they are listened to can significantly help in gaining their trust. Some interviewees told Mathur they want agents who are deliberate about understanding their preferences and don’t dismiss their questions.
Meeting these needs reduces the likelihood of protesting and creating conflict with family members.
“It highlights just how important well-designed explanations are,” she said. “We must go beyond a transparency checklist.”
]]>Georgia Tech researchers are making AI easier to understand through their work on Transformer Explainer. The free, online tool shows non-experts how ChatGPT, Claude, and other large language models (LLMs) process language.
Transformer Explainer is easy to use and runs on any web browser. It quickly went viral after its debut, reaching 150,000 users in its first three months. More than 563,000 people worldwide have used the tool so far.
Global interest in Transformer Explainer continues when the team presents the tool at the 2026 Conference on Human Factors in Computing Systems (CHI 2026). CHI, the world’s most prestigious conference on human-computer interaction, will take place in Barcelona, April 13-17.
“There are moments when LLMs can seem almost like a person with their own will and personality, and that misperception has real consequences. For example, there have been cases where teenagers have made poor decisions based on conversations with LLMs,” said Ph.D. student Aeree Cho.
“Understanding that an LLM is fundamentally a model that predicts the probability distribution of the next token helps users avoid taking its outputs as absolute. What you put in shapes what comes out, and that understanding helps people engage with AI more carefully and critically.”
A transformer is a neural network architecture that changes data input sequence into an output. Text, audio, and images are forms of processed data, which is why transformers are common in generative AI models. They do this by learning context and tracking mathematical relationships between sequence components.
Transformer Explainer demystifies how transformers work. The platform uses visualization and interaction to show, step by step, how text flows through a model and produces predictions.
Using this approach, Transformer Explainer impacts the AI landscape in four main ways:
“When I first learned about transformers, I felt overwhelmed. A transformer model has many parts, each with its own complex math. Existing resources typically present all this information at once, making it difficult to see how everything fits together,” said Grace Kim, a dual B.S./M.S. computer science student.
“By leveraging interactive visualization, we use levels of abstraction to first show the big picture of the entire model. Then users click into individual parts to reveal the underlying details and math. This way, Transformer Explainer makes learning far less intimidating.”
Many users don’t know what transformers are or how they work. The Georgia Tech team found that people often misunderstand AI. Some label AI with human-like characteristics, such as creativity. Others even describe it as working like magic.
Furthermore, barriers make it hard for students interested in transformers to start learning. Tutorials tend to be too technical and overwhelm beginners with math and code. While visualization tools exist, these often target more advanced AI experts.
Transformer Explainer overcomes these obstacles through its interactive, user-focused platform. It runs a familiar GPT model directly in any web browser, requiring no installation or special hardware.
Users can enter their own text and watch the model predict the next word in real time. Sankey-style diagrams show how information moves through embeddings, attention heads, and transformer blocks.
The platform also lets users switch between high-level concepts and detailed math. By adjusting temperature settings, users can see how randomness affects predictions. This reveals how probabilities drive AI outputs, rather than creativity.
“Millions of people around the world interact with transformer-driven AI. We believe that it is crucial to bridge the gap between day-to-day user experience and the models' technical reality, ensuring these tools are not misinterpreted as human-like or seen as sentient,” said Ph.D. student Alex Karpekov.
“Explaining the architecture helps users recognize that language generated by models is a product of computation, leading to a more grounded engagement with the technology.”
Cho, Karpekov, and Kim led the development of Transformer Explainer. Ph.D. students Alec Helbling, Seongmin Lee, Ben Hoover, and alumni Zijie (Jay) Wang (Ph.D. ML-CSE 2024) and Minsuk Kahng (Ph.D. CS-CSE 2019) assisted on the project.
Professor Polo Chau supervised the group and their work. His lab focuses on data science, human-centered AI, and visualization for social good.
Acceptance at CHI 2026 stems from the team winning the best poster award at the 2024 IEEE Visualization Conference. This recognition from one of the top venues in visualization research highlights Transformer Explainer’s effectiveness in teaching how transformers work.
“Transformer Explainer has reached over half a million learners worldwide,” said Chau, a faculty member in the School of Computational Science and Engineering.
“I'm thrilled to see it extend Georgia Tech's mission of expanding access to higher education, now to anyone with a web browser.”
]]>Now, the team is working with the Linux Foundation and the Open Source Security Foundation (OpenSSF) to ensure that its breakthrough doesn’t remain confined to a competition environment. The team’s new initiative, OSS-CRS, aims to standardize and operationalize cyber reasoning systems (CRSs) for real-world use.
“The AI Cyber Challenge pushed the boundaries of autonomous software security, with seven teams developing systems capable of finding and remediating vulnerabilities at scale,” said Andrew Chin, a Georgia Tech Ph.D. student and lead on the OSS-CRS program.
“However, after the competition’s conclusion, it has been difficult to apply these advancements to the open-source community due to infrastructure incompatibilities and the lack of long-term maintenance for the open-sourced CRS implementations.”
To address this gap, Georgia Tech’s Systems Software Lab (SSLab), directed by Professor Taesoo Kim, is leading the development of OSS-CRS, which provides both a common framework for CRS development and the infrastructure needed to deploy these systems seamlessly across open-source projects.
As part of this effort, the team has ported its competition-winning system, Atlantis, into the OSS-CRS framework. The move makes it compatible with laptops and other everyday machines with flexible resource and budget configurations.
Interoperability is also central to the framework’s design. Atlantis can be combined with other CRSs to improve performance, including systems developed by fellow AIxCC finalists and newer agentic, command-line-based tools. This modular approach reflects a key lesson the team learned from the competition: collaboration between systems can outperform any single solution.
OSS-CRS has been accepted as a sandbox project within OpenSSF’s AI/ML Security Working Group, a milestone that brings added technical guidance and community support to the project. This includes:
The collaboration will also foster stronger connections with open-source maintainers, helping streamline vulnerability disclosure and remediation workflows.
]]>They would normally get this in their natural habitats while foraging for food and staying alert to predators that might target calves.
However, the four elephants reside at Zoo Atlanta, so they don’t have to worry about these things.
That’s why zoo caretakers are always on the lookout for better ways to help their elephants exercise their brains.
The caretakers at Zoo Atlanta found one when they met Arianna Mastali, a Ph.D. student in Georgia Tech’s School of Interactive Computing. Mastali designed an audio enrichment wall to help stimulate Zoo Atlanta’s elephants.
Many zoos build concrete enrichment walls to foster elephant problem-solving and critical thinking. The walls usually have holes for the elephants to reach through with their trunks as they search for food, treats, or playful objects on the other side.
Mastali enhanced Zoo Atlanta’s enrichment wall by adding an interactive audio component. A nearby speaker system emits distinctive low-frequency tones when an elephant sticks its trunk into a hole.
“They’re intelligent creatures that require a lot of complexity in their habitat,” Mastali said. “We wanted to add to that complexity while giving them more control.”
Mastali’s system uses cameras and computer vision to detect when an elephant’s trunk is inside a hole and then sends a signal to the speakers to play a sound.
Mastali is a member of the Georgia Tech Animal Lab, directed by School of IC professor Melody Jackson. The lab often uses sensing technology to enhance animal wellness.
Mastali said she tried incorporating sensing devices into her project several times. She constructed an insert made of PVC pipe and attached a sensor to its base that used infrared beams to detect the elephant’s trunk.
However, she said it was difficult to account for the elephants’ strength. Their trunks would break the insert after a day or two.
She pivoted toward computer vision to remove the risk of damage and keep the enrichment wall as close to natural as possible.
“A big lesson we learned was that using existing materials the elephants are already familiar with was the best way to do things, and it simplified our design process,” she said.
Shane Rosse, a student in Georgia Tech’s Online Master of Science in Computer Science (OMSCS) program, assisted Mastali with the computer vision component.
Mastali observed the elephants’ behavior at the wall seven days before and seven days after the installation of the audio enrichment system.
The number of times the elephants approached the wall after installation increased by 176%, and time spent at the wall increased by 71%
“We weren’t sure at first if they would care that much, so it was great to see how much time they spent at the wall, especially our less dominant females,” said Kirby Miller, senior elephant caretaker at Zoo Atlanta. “They seem to like it the most.”
Miller said the elephants used to only approach the wall when they knew there was food behind it. That started to change after the audio enrichment system was installed.
“We would be off somewhere else, and we’d hear the speaker playing the sounds, and we knew there wasn’t any food back there,” Miller said. “Tara had her trunk in one of the holes, just listening to the sound. That let us know they do like it, and they’re very curious about it.”
Miller said because elephants have sharp memories and acute senses of hearing and smell, their habitats must be designed with that in mind.
Zoo Atlanta’s African Savanna elephant habitat was redesigned in 2019. In addition to the enrichment wall, it includes a bathing pond, two waterfalls, and swing boom devices that hold hay for elephants to eat as they would in the wild.
Miller said elephants sheltered at any zoo or conservation would benefit from enrichment devices enhanced by technology.
“I think anything they can participate in that gives them choice and control is great for all zoo elephants,” she said. “It depends on the elephants, but with our elephants, they can hear much higher frequencies than we can. That noise isn’t that loud for us, but for them, they’re feeling that noise, and they can hear much more, which makes it more stimulating for them.”
]]>The elephant enrichment wall at Zoo Atlanta recently received an upgrade thanks to a Georgia Tech Ph.D. student. Arianna Mastali designed an audio enrichment system that uses computer vision to detect when an elephant sticks its trunk into the enrichment wall as it searches for food. The system then sends a signal to play a unique tone from a nearby speaker that corresponds to each hole. So far, Mastali has found that elephant wall interactions have increased by 176%, and the elephants are visiting the wall even when there isn't food behind it.
]]>