When U.S. Rep. Earl L. “Buddy” Carter from Georgia’s 1st District visited Atlanta recently, one of his top priorities was meeting with the experts at Georgia Tech’s 20,000-square-foot Advanced Manufacturing Pilot Facility (AMPF).

Carter was recently named the House Energy and Commerce Committee’s chair of the Environment, Manufacturing, and Critical Materials Subcommittee, a group that concerns itself primarily with contamination of soil, air, noise, and water, as well as emergency environmental response, whether physical or cybersecurity.

Carter was recently named the House Energy and Commerce Committee’s chair of the Environment, Manufacturing, and Critical Materials Subcommittee, a group that concerns itself primarily with contamination of soil, air, noise, and water, as well as emergency environmental response, whether physical or cybersecurity.

Because AMPF’s focus dovetails with subcommittee interests, the facility was a fitting stop for Carter, who was welcomed for an afternoon tour and series of live demonstrations. Programs within Georgia Tech’s Enterprise Innovation Institute — specifically the Georgia Artificial Intelligence in Manufacturing (Georgia AIM) and Georgia Manufacturing Extension Partnership (GaMEP) — were well represented.

“Innovation is extremely important,” Carter said during his April 1 visit. “In order to handle some of our problems, we’ve got to have adaptation, mitigation, and innovation. I’ve always said that the greatest innovators, the greatest scientists in the world, are right here in the United States. I’m so proud of Georgia Tech and what they do for our state and for our nation.”

Carter’s AMPF visit began with an introduction by Tom Kurfess, executive director of the Georgia Tech Manufacturing Institute; Steven Ferguson, principal research scientist and managing director at Georgia AIM; research engineer Kyle Saleeby; and Donna Ennis, the Enterprise Innovation Institute’s director of community engagement and program development, and co-director of Georgia AIM.

Ennis provided an overview of Georgia AIM, while Ferguson spoke on the Manufacturing 4.0 Consortium and Kurfess detailed the AMPF origin story, before introducing four live demonstrations.

The first of these featured Chuck Easley, Professor of the Practice in the Scheller College of Business, who elaborated on supply chain issues. Afterward Alan Burl of EPICS: Enhanced Preparation for Intelligent Cybermanufacturing Systems and mechanical engineer Melissa Foley led a brief information session on hybrid turbine blade repair.

Finally, GaMEP project manager Michael Barker expounded on GaMEP’s cybersecurity services, and Deryk Stoops of Central Georgia Technical College detailed the Georgia AIM-sponsored AI robotics training program at the Georgia Veterans Education Career Transition Resource (VECTR) Center, which offers training and assistance to those making the transition from military to civilian life.

The topic of artificial intelligence, in all its subtlety and nuance, was of particular interest to Carter.

“AI is the buzz in Washington, D.C.,” he said. “Whether it be healthcare, energy [or] science, we on the Energy and Commerce Committee look at it from a sense [that there’s] a very delicate balance, and we understand the responsibility. But we want to try to benefit from this as much as we can.”

He continued: “I heard something today I haven’t heard before, and that is instead of calling it artificial intelligence, we refer to it as ‘augmented intelligence.’ I think that’s a great term, and certainly something I’m going to take back to Washington with me.”

Said Ennis, “It was a pleasure to host Rep. Carter for a firsthand look at AMPF, which is uniquely positioned to offer businesses the opportunity to collaborate with Georgia Tech researchers and students and to hear about Georgia AIM.”

She added, “At Georgia AIM, we’re committed to making the state a leader in artificial intelligence-assisted manufacturing, and we’re grateful for Congressman Carter’s interest and support of our efforts.”

Breast cancer is the second-most common cancer diagnosis for U.S. women, and the second-leading cause of female cancer deaths. In recent years, breast cancer treatments have improved significantly, thanks to targeted gene therapy and immunotherapy. However, for the small group of patients diagnosed with the most aggressive basal-like type of breast cancer, such approaches are less successful.

Recently, scientists in the Georgia Tech Integrated Cancer Research Center (ICRC) have found that this particular breast cancer displays a unique interactive gene network structure. Using a type of mathematics called “graph theory,” which models relationships between a pair of objects, the researchers computationally detected changes in gene-gene interactions as this breast cancer occurs and develops.

“The discovery of novel gene networks associated with basal-like breast cancers has helped us identify potential new gene targets to treat this very aggressive type of breast cancer,” said John McDonald, ICRC founding director, professor emeritus in the School of Biological Sciences, and the study’s corresponding author. “We would not have discovered these possible treatments through analyses of gene expression alone.”

While causing just 10-20% of breast cancer diagnoses, basal-like breast cancer is much more aggressive than other subtypes — and if not identified early, when it can be treated by surgery and/or radiation therapy, effective anti-cancer drug treatment can be challenging. The basal-like subtype does not respond to traditional hormonal therapies.

One theory as to why, advocated by many cancer researchers, is that individual genes do not function autonomously; as such, changes in how genes interact with one another in cancer may be as important as the cancer-driving genes themselves.

“The components of any complex system, like the human genome, are certainly important,” said McDonald. “The way in which these independent components interact with one another is also critical.”

For this study, the researchers analyzed three major subtypes of breast cancer, with particular emphasis on the most aggressive basal-like subtype. The researchers found that gene-gene interactive networks are quite different in the aggressive basal-like subtype, compared to the more prevalent luminal A and luminal B subtypes.

Many of the genes comprising these unique networks were found to be involved in functions not previously associated with breast cancer. Stephen Housley, a neurobiology researcher in the School of Biological Sciences and a co-author on the paper, noted that “an unexpected and intriguing result from our study is that neural processes appear to play a prominent role in distinguishing the highly aggressive basal-like tumors from the less aggressive luminal A and luminal B subtypes.”

In total, the researchers examined more than 300 million pairs of genes, comparing healthy women to those with breast cancer. Study co-author Zainab Ashard, a computational biologist who recently worked in McDonald’s lab, explained, “Differences in the gene network structure between healthy individuals and breast cancer patients allowed us to identify changes in patterns of gene-gene interactions within breast cancer development.”[s1] 

The team’s results are detailed in a new paper, “Changes in Gene Network Interactions in Breast Cancer Onset and Development,” which appeared in the April 2024 issue of GEN Biotechnology. Based on the results of this study and their previously published analyses of eight other types of cancer, the researchers believe they have established the usefulness of network analysis in identifying potential new candidates for the diagnosis of and targeted gene therapy treatment for breast and other types of cancers.

In addition to McDonald, Housley, and Ashard, Kara Keun Lee, a former bioinformatics Ph.D. student who worked in McDonald’s lab, is also a co-author on the paper.

The results shown here are in whole or in part based on data generated by the TCGA Research Network. The Genotype-Tissue Expression (GTEx) Project was supported by the Common Fund of the Office of the Director of the National Institutes of Health, and by NCI, NHGRI, NHLBI, NIDA, NIMH, and NINDS.

This research was supported by the Mark Light Integrated Cancer Research Center Student Fellowship, the Deborah Nash Endowment Fund, Northside Hospital (Atlanta), and the Ovarian Cancer Institute (Atlanta).

Citation: “Changes in Gene Network Interactions in Breast Cancer Onset and Development,” Zainab Arshad, Stephen N. Housley, Kara Keun Lee, and John F. McDonald, GEN Biotechnology, April 2024,
DOI: https://doi.org/10.1089/genbio.2024.0002

The College of Sciences is funding two research centers through a new seed grant program. 

Selected from a finalist pool of nine proposals, Associate Professors Yuanzhi Tang and Thackery Brown’s ideas were chosen for their high potential for novel interdisciplinary research and impact. 

Tang’s center will focus on sustainable mineral research, and Brown’s on spatial computation and navigation. Applications for the research will span the development of more sustainable batteries, as well as seeking to improve human health and well-being.

“Improving the human condition, fostering community, and pursuing research excellence are at the forefront of Georgia Tech’s mission, and these new centers will play a critical role in furthering that goal,” says Laura Cadonati, associate dean for Research in the College of Sciences and a professor in the School of Physics. “The College of Sciences is thrilled to support these new initiatives, and is excited to continue to develop the seed grant program.” 

A second call for research center proposals is planned for January 2025, with funding to start in July 2025.

The new Center for Sustainable and Decarbonized Critical Energy Mineral Solutions (CEMS), to be led by Yuanzhi Tang, an associate professor in the School of Earth and Atmospheric Sciences, will serve as a hub for sustainable procurement solutions for critical energy mineral resources, including rare earth elements and metals used for battery production.

Thackery Brown, an associate professor in the School of Psychology, will lead the second center, the Center for Research and Education in Navigation (CRaNE). CRaNE will investigate problems related to spatial computation, cognition, and navigation — which has implications for human health, animal conservation, smart architecture and urban design.

“This generous support from the College of Sciences will enable us to host a conference on spatial cognition, computation, design, and navigation; to provide collaborative multi-lab seed grants; and to establish the first of a series of explicitly co-mentored, interdisciplinary graduate student Fellowships,” Brown says. “Collectively, these are the seeds of a high-impact and self-sustaining center.”

About the Center for Sustainable Decarbonized Critical Energy Mineral Solutions (CEMS)

Yuanzhi Tang, School of Earth and Atmospheric Sciences 

Co-sponsored by the College of Sciences, Strategic Energy Institute (SEI), Brook Byers Institute for Sustainable Systems (BBISS), Institute for Electronics and Nanotechnology (IEN), and Institute for Materials (iMat), CEMS began as a joint BBISS-SEI initiative lead project that has since grown into a joint center focused on critical elements and materials for sustainable energy.

Sustainably sourcing these materials provides a critical foundation for both high-tech industry and green economy. “Rare earth elements and battery metals like lithium, copper, and nickel are in high demand, but low domestic resources and production have resulted in a heavy reliance on imports,” Tang explains. “How can we domestically produce these resources, and how can we do this sustainably? Georgia Tech and the College of Sciences are in a unique position for developing a large regional research umbrella to connect these dots.”

CEMS will leverage on three key pillars: science and technology development, strengthening collaboration among the University System of Georgia (USG) universities, and developing regional resources and economy, Tang says. “By leveraging collaboration among Georgia universities, and fostering engagement with regional industries, the Center will develop new science and technology, leading the way in research on how to procure these ‘essential vitamins’ for clean energy transition in a sustainable and decarbonized manner.”

About the Center for Research and Education in Navigation (CRaNE)

Thackery Brown, School of Psychology 

CRaNE will focus on solving problems related to spatial computation, cognition, and navigation. “How do we treat catastrophic loss of one’s ability to get from A to B in Alzheimer's disease? How do we build smarter cities that are easier and more carbon efficient to navigate? How can we develop robots,” Brown says, “which navigate with the flexibility and efficiency of our own minds? CRaNE will bring together experts from many different fields to help address these problems with truly creative and integrative scientific and technological solutions.”

CRaNE will support interdisciplinary collaborative research, including developing a graduate student fellowship program, and conducting K-12 outreach.

“Our goal for CRaNE is to position the College of Sciences, Georgia Tech, and our extended network of collaborator institutions as a center of gravity for cutting-edge work on how the mind, brain, and artificial systems process space — how they can be made better at it, and how we can engineer our world around us in ways that support the humans and animals that need to navigate it to survive,” Brown says.

Emphasizing the collaborative nature of CRaNE, Brown adds that “by targeting collaborative grants, research, and education, and by promoting outreach and education earlier in the STEM pipeline, we hope to accelerate progress at the frontiers of these fields — and to invest in future science that cannot be easily addressed by a single lab or discipline.”

This article was first published in the University of Illinois Urbana-Champaign newsroom. Read the full story here.

Researchers at Georgia Institute of Technology and the University of Illinois Urbana-Champaign have developed a first-of-its-kind technique called electron videography to capture moving images at the molecular scale. In the first demonstration of the technique, the team took a microscopic moving picture of the delicate dance between proteins and lipids found in cell membranes. The study, “Electron videography of a lipid–protein tango” was published last week in the journal Science Advances.

"This is the first time we are looking at a protein on an individual scale and haven't frozen it or tagged it," says Aditi Das, a corresponding author and associate professor in the School of Chemistry and Biochemistry at Georgia Tech.

Electron microscopy techniques image at the molecular or atomic scale, yielding detailed, nanometer-scale pictures. However, they often rely on samples that have been frozen or fixed in place, leaving scientists to try to infer how molecules move and interact — like trying to map the choreography of a dance sequence from a single frame of film.

"Usually, we have to crystalize or freeze a protein, which poses challenges in capturing high-resolution images of flexible proteins. Alternately, some techniques use a molecular tag that we track, rather than watching the protein itself,” Das says. “In this study we are seeing the protein as it is, behaving how it does in a liquid environment, and seeing how lipids and proteins interact with each other."

The technique can be used to study the dynamics of other biomolecules, breaking free of constraints that have limited microscopy to still images of fixed molecules. In this study, the team examined nanoscale discs of lipid membranes and how they interacted with proteins normally found on the surface of or embedded in cell membranes.

These membrane proteins are significant for medical treatments, and are involved in processes including muscle contraction, brain function, and immune system functions. Moving forward, the researchers plan to use their electron videography technique to study other types of membrane proteins and other classes of molecules and nanomaterials.

DOI: 10.1126/sciadv.adk0217

With new vehicle models being developed by major brands and a growing supply chain, the electric vehicle (EV) revolution seems well underway. But, as consumer purchases of EVs have slowed, car makers have backtracked on planned EV manufacturing investments. A major roadblock to wider EV adoption remains the lack of a fully realized charging infrastructure. At just under 51,000 public charging stations nationwide, and sizeable gaps between urban and rural areas, this inconsistency is a major driver of buyer hesitance.

How do we understand, at a large scale, ways to make it easier for consumers to have confidence in public infrastructure? That is a major issue holding back electrification for many consumer segments.

- Omar Asensio, Associate Professor at Georgia Institute of Technology and Climate Fellow, Harvard Business School | Director, Data Science & Policy Lab

Omar Asensio, associate professor in the School of Public Policy and director of the Data Science and Policy Lab at the Georgia Institute of Technology, and his team have been working to solve this trust issue using the Microsoft CloudHub partnership resources. Asensio is also currently a visiting fellow with the Institute for the Study of Business in Global Society at the Harvard Business School.

The CloudHub partnership gave the Asensio team access to Microsoft’s Azure OpenAI to sift through vast amounts of data collected from different sources to identify relevant connections. Asensio’s team needed to know if AI could understand purchaser sentiment as negative within a population with an internal lingo outside of the general consumer population. Early results yielded little. The team then used specific example data collected from EV enthusiasts to train the AI for a sentiment classification accuracy that now exceeds that of human experts and data parsed from government-funded surveys.

The use of trained AI promises to expedite industry response to consumer sentiment at a much lower cost than previously possible. “What we’re doing with Azure is a lot more scalable,” Asensio said. “We hit a button, and within five to 10 minutes, we had classified all the U.S. data. Then I had my students look at performance in Europe, with urban and non-urban areas. Most recently, we aggregated evidence of stations across East and Southeast Asia, and we used machine learning to translate the data in 72 detected languages.”

We are excited to see how access to compute and AI models is accelerating research and having an impact on important societal issues. Omar's research sheds new light on the gaps in electric vehicle infrastructure and AI enables them to effectively scale their analysis not only in the U.S. but globally.

- Elizabeth Bruce, Director, Technology for Fundamental Rights, Microsoft

Asensio's pioneering work illustrates the interdisciplinary nature of today’s research environment, from machine learning models predicting problems to assisting in improving EV infrastructure. The team is planning on applying the technique to datasets next, to address equity concerns and reduce the number of “charging deserts.” The findings could lead to the creation of policies that help in the adoption of EVs in infrastructure-lacking regions for a true automotive electrification revolution and long-term environmental sustainability in the U.S.

- Christa M. Ernst

Source Paper: Reliability of electric vehicle charging infrastructure: A cross-lingual deep learning approach - ScienceDirect

A Georgia Tech biomedicine pioneer is poised to transform pediatric healthcare.

Following an internal search, Stanislav Emelianov, Joseph M. Pettit Endowed Chair, Georgia Research Alliance Eminent Scholar, and professor of electrical and computer engineering and biomedical engineering at Georgia Tech, has been named co-director of the Children’s Healthcare of Atlanta Pediatric Technology Center (PTC).

The PTC is a unique partnership that supports interdisciplinary research among clinicians, engineers, and scientists from Georgia Tech, Children’s, and Emory University. Together, they harness the power of artificial intelligence, data science, and cutting-edge medical devices to address the most pressing challenges in pediatric healthcare.

Emelianov is driven by a profound dedication to improving children's well-being. He said, “I am committed to accelerating pediatric research and innovation at Georgia Tech and Children’s, with a sincere aspiration for the patient-centric PTC to transform pediatric healthcare not only in Atlanta and Georgia but also beyond. Together, we aim to set a new standard, positively impacting the lives of children far and wide.”

Emelianov brings a wealth of experience to the PTC. He has served as the director of Georgia Tech’s Ultrasound Imaging and Therapeutics Research Laboratory since 2015. Among his many accolades and distinctions, he is a nationally renowned trailblazer in advanced imaging methods to diagnose, monitor, and treat cancer and other diseases.

“Professor Emelianov’s history of scholarship and innovation in medical imaging and his vision to engage and support Georgia Tech faculty and students embarking on pediatric research will be a great asset to the Pediatric Technology Center,” said Julia Kubanek, Georgia Tech vice president for Interdisciplinary Research. “We are also enormously grateful to Professor M.G. Finn who has served as the leader of the PTC at Georgia Tech for seven years, building collaborations and enabling our research community to tackle research problems of importance to kids’ health.”

Emelianov will now hold the James A. Carlos Family Chair for Pediatric Technology, which had been held by M.G. Finn since 2017. The proceeds of the chair will be used for research in the PTC. Georgia Tech extends its gratitude to the Carlos family for their generous contributions to pediatrics.

Stanislav Emelianov’s PTC co-director will be appointed by Children’s.

When U.S. Rep. Earl L. “Buddy” Carter from Georgia’s 1st District visited Atlanta recently, one of his top priorities was meeting with the experts at Georgia Tech’s 20,000-square-foot Advanced Manufacturing Pilot Facility (AMPF).

Carter was recently named the House Energy and Commerce Committee’s chair of the Environment, Manufacturing, and Critical Materials Subcommittee, a group that concerns itself primarily with contamination of soil, air, noise, and water, as well as emergency environmental response, whether physical or cybersecurity.

Because AMPF’s focus dovetails with subcommittee interests, the facility was a fitting stop for Carter, who was welcomed for an afternoon tour and series of live demonstrations. Programs within Georgia Tech’s Enterprise Innovation Institute — specifically the Georgia Artificial Intelligence in Manufacturing (Georgia AIM) and Georgia Manufacturing Extension Partnership (GaMEP) — were well represented.

“Innovation is extremely important,” Carter said during his April 1 visit. “In order to handle some of our problems, we’ve got to have adaptation, mitigation, and innovation. I’ve always said that the greatest innovators, the greatest scientists in the world, are right here in the United States. I’m so proud of Georgia Tech and what they do for our state and for our nation.”

Carter’s AMPF visit began with an introduction by Thomas Kurfess, Regents' Professor and HUSCO/Ramirez Distinguished Chair in Fluid Power and Motion Control in the George W. Woodruff School of Mechanical Engineering and executive director of the Georgia Tech Manufacturing Institute; Steven Ferguson, principal research scientist and managing director at Georgia AIM; research engineer Kyle Saleeby; and Donna Ennis, the Enterprise Innovation Institute’s director of community engagement and program development, and co-director of Georgia AIM.

Ennis provided an overview of Georgia AIM, while Ferguson spoke on the Manufacturing 4.0 Consortium and Kurfess detailed the AMPF origin story, before introducing four live demonstrations.

The first of these featured Chuck Easley, Professor of the Practice in the Scheller College of Business, who elaborated on supply chain issues. Afterward, Alan Burl of EPICS: Enhanced Preparation for Intelligent Cybermanufacturing Systems and mechanical engineer Melissa Foley led a brief information session on hybrid turbine blade repair.

Finally, GaMEP project manager Michael Barker expounded on GaMEP’s cybersecurity services, and Deryk Stoops of Central Georgia Technical College detailed the Georgia AIM-sponsored AI robotics training program at the Georgia Veterans Education Career Transition Resource (VECTR) Center, which offers training and assistance to those making the transition from military to civilian life.

The topic of artificial intelligence, in all its subtlety and nuance, was of particular interest to Carter.

“AI is the buzz in Washington, D.C.,” he said. “Whether it be healthcare, energy, [or] science, we on the Energy and Commerce Committee look at it from a sense [that there’s] a very delicate balance, and we understand the responsibility. But we want to try to benefit from this as much as we can.”

“I heard something today I haven’t heard before," Carter continued, "and that is instead of calling it artificial intelligence, we refer to it as ‘augmented intelligence.’ I think that’s a great term, and certainly something I’m going to take back to Washington with me.”

“It was a pleasure to host Rep. Carter for a firsthand look at AMPF," shared Ennis, "which is uniquely positioned to offer businesses the opportunity to collaborate with Georgia Tech researchers and students and to hear about Georgia AIM.

“At Georgia AIM, we’re committed to making the state a leader in artificial intelligence-assisted manufacturing, and we’re grateful for Congressman Carter’s interest and support of our efforts."

ATLANTA — Annoviant Inc. a health technology company and member startup in the Center for MedTech Excellence at Georgia Tech's Enterprise Innovation Institute, is receiving a $2.99 million National Institutes of Health (NIH) grant to further scale the development and commercialization of its TxGuard™ pulmonary-valved conduit for pediatric heart disease.

The award follows two Phase I NIH grants the company received, the most recent being in 2021.

Annoviant's patented TxGuard™ stands at the forefront of technological innovation in conduit replacements for treating congenital heart disease (CHD), the most prevalent birth defect globally and a leading cause of birth-related mortality, the company said.

CHD encompasses a broad range of abnormalities that disrupt blood flow to and from the heart. It affects approximately 40,000 newborns annually — or 1% of births in the U.S. — and 1.35 million worldwide. With an estimated 2.9 million CHD patients in the U.S. alone, the need for advanced solutions is paramount.

"This marks a significant milestone for Annoviant as we accelerate our pursuit of impactful innovation to save lives," said Annoviant CEO and co-founder Ajay Houde, Ph.D. "It validates our hypothesis and shows the NIH's confidence in our ability to make good progress. Because we are a small startup, it gives private investors the confidence to invest with us and more companies working with us across the broader ecosystem."

Addressing critical shortcomings observed in current commercial devices, TxGuard™ offers clinical advantages, notably its resistance to calcification, thrombosis, infection, and the host cell integration. This cutting-edge technology marks a new era in pediatric cardiac interventions, providing durable pulmonary valved grafts that adapt and regenerate alongside patients, minimizing the need for multiple re-operations over their lifetimes.

"Heart disease is the leading killer of men and women in the U.S. and is the most common birth defect in our newborns," said Center for MedTech Excellence Director Nakia Melecio, who worked with Annoviant to help it scale and reviewed its federal funding submissions.

The Center for MedTech Excellence, which launched in 2022, works with early-stage life sciences startups that have specific obstacles that young tech companies in other sectors don't face.

"This is a critical milestone for the company, and validates its research and work, thus far," Melicio said. "Annoviant's technology is tackling several challenges that the market currently faces and elevating the possibility for better patient outcomes in management of congestive heart failure."

Pediatric patients with CHD often undergo multiple cardiovascular surgeries throughout their lives, with associated costs totaling billions for the U.S. healthcare industry. TxGuard™ offers a transformative solution to this ongoing challenge, promising extended durability and reduced healthcare burden for patients and providers alike.

He credited the company's work with the Center for MedTech Excellence and being a health tech startup in the Advanced Technology Development Center (ATDC), the Enterprise Innovation Institute's startup incubator, as being pivotal in Annoviant's growth.

ATDC SBIR/STTR Catalyst Connie Casteel, who works with the incubator's portfolio companies to help the prepare for these federal, non-dilutive funding grants, had worked with Annoviant on its federal funding approach and strategy.

"We went through the 16-week program with the MedTech Center and it really helped us think through the various aspects of the commercialization process and operational challenges we would face," Houde said. "Greg Jungles at ATDC was also instrumental in helping us.  I'm really thankful for Nakia and his work with the MedTech Center and Greg and the team at ATDC."

More and more often, automated lending systems powered by artificial intelligence (AI) reject qualified loan applicants without explanation.

Even worse, they leave rejected applicants with no recourse.

People can have similar experiences when applying for jobs or petitioning their health insurance providers. While AI tools determine the fate of people in difficult situations daily, Upol Ehsan says more thought should be given to challenging these decisions or working around them.

Ehsan, a Georgia Tech explainable AI (XAI) researcher, says many rejection cases are not the applicant’s fault. Rather, it’s more likely a “seam” in the design process — a mismatch between what designers thought the AI could do and what happens in reality.

Ehsan said “seamless design” is the standard practice of AI designers. While the goal is to create a process by which users get what they need without interruption or barriers, seamless design has a way of doing just the opposite. 

No amount of thought or design input will keep AI tools from making mistakes. When mistakes happen, those impacted by them want to know why they happened.

Because seamless design often includes black-boxing — the act of concealing the AI’s reasoning — answers are never provided.

But what if there were a way to challenge an AI’s decisions and turn its mistakes into benefits for end users? Ehsan believes that can be done through “seamful design.”

n his latest paper, Seamful Explainable AI: Operationalizing Seamful Design in XAI, Ehsan proposes a strategic way of anticipating AI harms, learning their reasonings, and leveraging mistakes instead of concealing them. 

GIVING USERS MORE OPTIONS

In his research, Ehsan worked with loan officers who used automated lending support systems. The seams, or flaws, he discovered in these tools’ processes impacted applicants and lenders.

“The expectation is that the lending system works for everyone,” Ehsan said. “The reality is that it doesn’t. You’ve found the seam once you’ve figured out the difference between expectation and reality. Then we ask, ‘How can we show this to end users so they can leverage it?’”

To give users options when AI negatively impacts them, Ehsan suggests three things for designers to consider:

• Actionability: Does the information about the flaw help the user take informed actions on the AI’s recommendation?
• Contestability: Does the information provide the resources necessary to justify saying no to the AI?
• Appropriation: Does identifying these seams help the user to adapt and appropriate the AI’s output in a way that is different from the provided design but helps the user make the right decision?

Ehsan uses the example of someone who was rejected for a loan despite having a good credit history. The rejection may have been due to a seam, such as a flawed discriminating algorithm, in the AI that screens the applications.

A post-deployment process is needed in cases like this to mitigate damage and empower affected end users. Loan applicants, for instance, should be allowed to contest the AI’s decision based on known issues with an algorithm. 

AGAINST THE GRAIN

Ehsan said his idea for seamful design is outside of the mainstream vernacular. However, his challenge to current accepted principles is gaining traction.

He is now working with cybersecurity, healthcare, and sales companies that are adopting his process.

These companies may pioneer a new way of thinking in AI design. Ehsan believes this new mindset can allow designers to switch to a proactive mindset instead of being stuck in a reactive state of conducting damage control.

“You want to stay a little ahead of the curve so you’re not always caught off guard when things happen,” Ehsan said. “The more proactive you can be and the more passes you can take at your design process, the safer and more responsible your systems will be.”

Ehsan collaborated with researchers from Georgia Tech, the University of Maryland, and Microsoft. They will present their paper later this year at the 2024 Association for Computing Machinery’s Conference on Computer-Supported Cooperative Work and Social Computing (CSCW) in Costa Rica. 

“Seamful design embraces the imperfect reality of our world and makes the most out of it,” he said. “If it becomes mainstream, it can help us address the hype cycle AI suffers from now. We don’t need to overhype AI’s capacity or impose unachievable goals. That’d be a gamechanger in calibrating people’s trust in the system.” 

Computing research at Georgia Tech is getting faster thanks to a new state-of-the-art processing chip named after a female computer programming pioneer.

Tech is one of the first research universities in the country to receive the GH200 Grace Hopper Superchip from NVIDIA for testing, study, and research.

Designed for large-scale artificial intelligence (AI) and high-performance computing applications, the GH200 is intended for large language model (LLM) training, recommender systems, graph neural networks, and other tasks. 

Alexey Tumanov and Tushar Krishna procured Georgia Tech’s first pair of Grace Hopper chips. Spencer Bryngelson attained four more GH200s, which will arrive later this month.

“We are excited about this new design that puts everything onto one chip and accessible to both processors,” said Will Powell, a College of Computing research technologist.

“The Superchip’s design increases computation efficiency where data doesn’t have to move as much and all the memory is on the chip.” 

A key feature of the new processing chip is that the central processing unit (CPU) and graphics processing unit (GPU) are on the same board.

NVIDIA’s NVLink Chip-2-Chip (C2C) interconnect joins the two units together. C2C delivers up to 900 gigabytes per second of total bandwidth, seven times faster than PCIe Gen5 connections used in newer accelerated systems.  

As a result, the two components share memory and process data with more speed and better power efficiency. This feature is one that the Georgia Tech researchers want to explore most.

Tumanov, an assistant professor in the School of Computer Science, and his Ph.D. student Amey Agrawal, are testing machine learning (ML) and LLM workloads on the chip. Their work with the GH200 could lead to more sustainable computing methods that keep up with the exponential growth of LLMs.

The advent of household LLMs, like ChatGPT and Gemini, pushes the limit of current architectures based on GPUs. The chip’s design overcomes known CPU-GPU bandwidth limitations. Tumanov’s group will put that design to the test through their studies. 

Krishna is an associate professor in the School of Electrical and Computer Engineering and associate director of the Center for Research into Novel Computing Hierarchies (CRNCH).

His research focuses on optimizing data movement in modern computing platforms, including AI/ML accelerator systems. Ph.D. student Hao Kang uses the GH200 to analyze LLMs exceeding 30 billion parameters. This study will enable labs to explore deep learning optimizations with the new chip.  

Bryngelson, an assistant professor in the School of Computational Science and Engineering, will use the chip to compute and simulate fluid and solid mechanics phenomena. His lab can use the CPU to reorder memory and perform disk writes while the GPU does parallel work. This capability is expected to significantly reduce the computational burden for some applications.

“Traditional CPU to GPU communication is slower and introduces latency issues because data passes back and forth over a PCIe bus,” Powell said. “Since they can access each other’s memory and share in one hop, the Superchip’s architecture boosts speed and efficiency.” 

Grace Hopper is the inspirational namesake for the chip. She pioneered many developments in computer science that formed the foundation of the field today.  

Hopper invented the first compiler, a program that translates computer source code into a target language. She also wrote the earliest programming languages, including COBOL, which is still used today in data processing. 

Hopper joined the U.S. Navy Reserve during World War II, tasked with programming the Mark I computer. She retired as a rear admiral in August 1986 after 42 years of military service.

Georgia Tech researchers hope to preserve Hopper’s legacy using the technology that bears her name and spirit for innovation to make new discoveries.

“NVIDIA and other vendors show no sign of slowing down refinement of this kind of design, so it is important that our students understand how to get the most out of this architecture,” said Powell. 

“Just having all these technologies isn’t enough. People must know how to build applications in their coding that actually benefit from these new architectures. That is the skill.”