From art to medicine: How 青涩直播 researchers are using AI to solve real-world problems
As an R1 university, 青涩直播 is driving AI innovation across campus and beyond.
From drug discovery and education to economics and earth science, 青涩直播 researchers are using AI to tackle a wide range of real-world challenges. 青涩直播’s high-performance computing capabilities — made up of NVIDIA DGX SuperPOD and ManeFrame III — allow professors, students and researchers to use large-scale AI/ML (artificial intelligence/machine learning), generative AI and other types of AI to accelerate advances in research innovation, with help from 青涩直播’s O'Donnell Data Science and Research Computing Institute and the Office of Information Technology (OIT).
Learn about some of the AI research being done at 青涩直播:
Robert Hunt: Theology Meets AI
Robert Hunt, director of Global Theological Education at Perkins School of Theology, approaches AI as botfigures. Students have also created AI tutors using course materials to aid in studying.
Hunt recognizes that students often discover practical AI applications faster than faculty, advocating for more collaborative teaching approaches rather than traditional hierarchies.
Beyond practical uses, Hunt studies AI's broader societal implications, particularly how human-like AI agents will reshape human self-understanding. He worries that as AI becomes more social, humans risk adopting abstract, machine-like views of themselves. Hunt's research emphasizes authentic humanity exists through human relationships and collective imagination of what it means to be human, not technological companions.
"Those who are used to a slower pace of research and writing are going to find that they're left behind if they ignore AI,” he said. “Ironically, AI would relieve this pressure if used to do tasks that distract from reflection and developing student relationships.”
Jiun-Yu Wu: Chatbots for the Classroom
Jiun-Yu Wu, a professor in the Department of Teaching and Learning at Simmons School of Education and Human Development, has led AI-empowered education research for nearly a decade. Starting with machine learning applications in 2016, his work has evolved into comprehensive systems supporting students and educators.
Wu's research began with a practical problem in his statistics classrooms. Recognizing student anxiety in traditional settings, he developed a hybrid format combining online videos, social media discussions, and in-person meetings. This approach generated multimodal data requiring AI and statistical analysis to understand learning engagement patterns.
After ChatGPT's launch, Wu's team developed a statistics chatbot for messaging apps that students can download to their devices. The AI-powered chatbot, trained on course materials, offers 24/7 personalized academic support with specific content references.
"When you post your question to the chatbot, it will not only answer questions but also provide specific references to course materials, helping students understand both the what and why behind concepts," Wu said. "This approach addresses concerns about AI providing answers without context or learning opportunities."
Wu also uses AI for writing assistance and formatting, though he emphasizes careful verification. He advocates for proactive AI integration, comparing it to the adoption of the Internet and cell phones, while stressing that humans must lead AI collaboration.
Ira Greenberg of 青涩直播's Meadows School bridges art and AI through pioneering research and teaching.
Ira Greenberg: Coding Creativity
For Ira Greenberg, director of Creative Computation at Meadows School of the Arts, what began in 1992 as an exploration of computation in creative practice has evolved into AI research that spans from generative art to practical applications.
Greenberg’s artwork has been showcased worldwide, and he wrote the first major reference on the Processing programming language, which makes coding accessible to artists, designers, educators and beginners.
His current work focuses on developing AI systems using 青涩直播’s high-performance computing resources, where he builds agentic frameworks that provide structure for networks of LLMs to communicate and collaborate to solve complex challenges.
Greenberg is currently pioneering “Creative Computing Moonshots,” a teaching technique in which 青涩直播 students use AI to code their desired projects and then reverse-engineer them to gain a deeper understanding of their creations throughout the semester. This approach to AI enables him and his students to leverage existing models while creating specialized applications that can match or surpass human-created documentation.
“I want my students to be masters of AI coming out of school,” he said. “It’s a necessary tool for preparing them for their careers.”
Greenberg’s company, Gogentic AI, demonstrates this practical approach. The system can create highly technical courses that would traditionally take months to develop and often become obsolete before completion.
Vishal Ahuja: Smarter Healthcare Decisions
青涩直播's AI for Healthcare Leaders program equips senior executives to leverage artificial intelligence for enhanced business performance and patient outcomes. Led by Vishal Ahuja, academic director of the Cox School of Business's Healthcare Initiative, the program emphasizes understanding the causes of problems rather than focusing solely on specific AI tools.
Ahuja's research develops decision-making tools that help healthcare professionals and policymakers improve care quality, patient health, and delivery efficiency.
"My focus is to highlight simple, nonclinical approaches that impact patient outcomes, particularly for diabetes, which costs over $412 billion annually," he said. "Using Veterans Affairs data, we've demonstrated that continuity of primary care significantly reduces hospitalizations and unnecessary diagnostic tests."
Ahuja's findings have also shown that co-locating mental and physical health providers decreases severe outcomes among diabetic patients with mental health conditions, including extended hospital stays from substance abuse or suicide attempts. The biggest improvements occurred among severely ill patients.
Ahuja is collaborating with several North Texas healthcare leaders to translate AI innovations into practical solutions, ensuring emerging technologies create meaningful improvements in patient care and hospital operations. The findings have important implications for policymaking around patient scheduling and care coordination.
Corey Clark's interdisciplinary approach uses gaming technology to address societal challenges and combines human-machine collaboration through AI applications.
Corey Clark: Gaming Tech for Real-World Solutions
Corey Clark, deputy director of research at 青涩直播 Guildhall and associate professor of computer science at the Lyle School of Engineering, uses gaming technology to address societal challenges through AI applications. His interdisciplinary approach combines human-machine collaboration, explainable AI, and bias mitigation for problems ranging from drug discovery to human trafficking.
Clark, 青涩直播 chemist Elfi Kraka and their students developed the SmartCADD platform, combining AI with quantum mechanics to accelerate drug discovery. At the Intelligent Systems and Bias Examination Lab, he leads efforts to address facial recognition fairness using synthetic datasets. Clark's team has created over 1 million diverse facial images to reduce demographic bias in AI systems.
“The rapid advancements in both gaming technology and AI have enable amazing new opportunities to advance many fields, and their convergence has created new paradigms on humans and machines can collaborate,” Clark said.
Clark's gaming research includes Dark Shadows, created with economist Beth Wheaton-Páramo and 青涩直播's human trafficking research team. This video game assists law enforcement in analyzing trafficking data by having players investigate imaginary crimes while organizing actual trafficking information from Department of Justice press releases, thereby enhancing AI algorithms that extract details while preserving victim privacy.
As CTO of BALANCED Media Technology, Clark developed the HEWMEN platform, which enables players to contribute computational power and human intuition through gameplay, creating a "Community as a Service" model.
Ömer Özak: Economics Gets an AI Boost
As a macroeconomist with over 20 years in computational economics, Ömer Özak has no trouble writing code to build computational models that can answer key economic questions.
But 青涩直播’s associate professor of economics at the Dedman College of Humanities and Sciences regularly uses AI to assist him with some of the necessary, but mundane tasks that come with writing code and data analysis. This allows him to focus on what he really wants to do: explore the deep historical origins and long-run consequences of some of the most fundamental cultural, human and economic characteristics that are at the roots of contemporary comparative economic development across countries, regions and ethnic groups.
For instance, he plans to use AI’s large language models – systems capable of understanding and generating human language by processing lots of text data – to help him implement and classify data from complex datasets to try to determine the role of culture and identity in secessionist conflicts around the world.
AI, Özak said, doesn’t always give him a solution that works. "But somehow, it takes you closer to where you want to go in the end, which ends up being quite helpful,” he said.
Jia Zhang: AI Assistants for Science
Jia Zhang, the Interim Chair of Computer Science at 青涩直播’s Lyle School of Engineering, wants to give researchers an AI assistant, so they can speed up scientific advances.
Her work includes an AI-powered “Alexa-like” virtual assistant, created with NASA collaborators, that assists Earth scientists sift through decades of geospatial data with natural-language questions.
She is also working on a system that uses advanced machine learning techniques to automatically recommend chainable software modules and algorithms, known as Units of Work. This intelligent system – called RANGER – seeks to save researchers valuable time and boost productivity by handling the technical challenges of software integration.
“As scientific workflows continue to grow in complexity, RANGER stands poised to revolutionize how researchers approach big data analytics, potentially accelerating scientific discoveries across various disciplines,” Zhang said.
Zhang is also working to make AI models more trustworthy and explainable. One project involves an "AI vaccine" that trains models to better detect bad data by exposing them to it, similar to how a biological vaccine works. She also wants to develop algorithms that can explain how AI reaches conclusions, helping build user trust.
Carla L. Reyes: Law in the Age of AI
While AI can improve how legal practitioners function and how people access the justice system, Carla L. Reyes, a recognized leader on issues raised by the intersection of business law and emerging technology who teaches at the 青涩直播 Dedman School of Law, warned that lawyers also need to be wary of possible unintended consequences.
“The prompts used to generate answers matter,” Reyes said. “The way the person who receives the answers reviews them, confirms their content, and then uses the answers matter.”
Not all lawyers realize this, she said, which has resulted in them submitting material generated by a large language model to a court without confirming its accuracy, and then finding out later that the cases cited do not exist or the material was otherwise inaccurate.
Exploring these kinds of AI “ripple effects” is a core part of Reyes’ research.
For instance, she has studied legal frameworks for granting personhood to autonomous entities like AI systems or code-driven corporations. She’s also analyzed how terminology around AI in criminal justice – like using “AI judge” or “AI tool” shape public understanding, policy standards and trust.
Another worry for Reyes is that “we will make bad law due to our lack of understanding or in our rush to make AI-specific law.”
Neena Imam leads efforts to make federated learning more secure against hacked devices and develops the AREAD initiative—creating AI systems that smartly adjust their computing complexity in real-time based on available resources.
Neena Imam: Resilient, Resource-Smart AI
When a lot of decentralized devices – like smartphones or laptops – work together to train a shared AI model to perform a task, it’s called federated learning (FL). But sometimes, some of these devices might be hacked or give wrong information on purpose.
Neena Imam, Director of the O’Donnell Data Science and Research Computing Institute at 青涩直播, is leading a team to try to make FL more resilient against these types of threats.
Her team is also working to address another issue with FL: heterogeneity. That’s when different devices use different models, making it hard for FL to generate one accurate AI model without needing everyone to use the same setup.
In addition, Imam is the lead researcher for the AREAD (Advancing Resource-Efficient AI and Smart Data Management) initiative, which seeks to develop a software framework that lets AI models smartly adjust their computing complexity in real-time based on what they are processing and available resources.
In 2025, Lyda Hill Holdings gave Imam $100,000 in funding for the AREAD initiative.
“AI is transforming sectors like healthcare, finance, and smart cities by enabling faster diagnoses, smarter risk management, and efficient urban systems, all the while lowering costs,” she said. “Responsible AI deployment is essential to building a more intelligent, equitable, and trustworthy future.”
Carrie Johnston: Librarians as AI Guides
Carrie Johnston, Director of Research and Scholarly Initiatives for 青涩直播 Libraries, approaches AI as an evolution of the libraries’ core mission: helping people organize and find the best resources for teaching and research. Rather than wholesale adoption or rejection, the libraries take a critical, collaborative approach to emerging AI tools.
"What libraries have always taught is critical information literacy," Johnston said. "Just like you would evaluate a website or journal article as a source, you're also going to evaluate an AI tool and the results it's providing using those same principles that librarians have been teaching for years."
青涩直播 librarians are actively learning about new tools to suggest best practices while respecting faculty and students across the spectrum of AI adoption. Johnston emphasizes that understanding AI's impact requires campus-wide collaboration, particularly in identifying which tools provide accurate information versus those that may hallucinate sources or provide incorrect results.
"We want to engage with people who are using it and with those who are skeptical about it, because that is all part of taking a critical approach," she said. "We're here for everybody, no matter their orientation to AI or their comfort level."