Whitney Zhang ’21 believes in the importance of valuing workers regardless of where they fit into an organizational chart.

Zhang is a PhD student in MIT’s Department of Economics studying labor economics. She explores how the technological and managerial decisions companies make affect workers across the pay spectrum. 

“I’ve been interested in economics, economic impacts, and related social issues for a long time,” says Zhang, who majored in mathematical economics as an undergraduate. “I wanted to apply my math skills to see how we could improve policies and their effects.”

Zhang is interested in how to improve conditions for workers. She believes it’s important to build relationships with policymakers, focusing on an evidence-driven approach to policy, while always remembering to center those the policies may affect. “We have to remember the people whose lives are impacted by business operations and legislation,” she says. 

She’s also aware of the complex intermixture of politics, social status, and financial obligations organizations and their employees have to navigate.

“Though I’m studying workers, it’s important to consider the entire complex ecosystem when solving for these kinds of challenges, including firm incentives and global economic conditions,” she says.

The intersection of tech and labor policy

Zhang began investigating employee productivity, artificial intelligence, and related economic and labor market phenomena early in her time as a doctoral student, collaborating frequently with fellow PhD students in the department.

A collaboration with economics doctoral student Shakked Noy yielded the 2023 study investigating ChatGPT as a tool to improve productivity. Their research found it substantially increased workers’ productivity on writing tasks, most so for workers who initially performed the worst on the tasks.

“This was one of the earliest pieces of evidence on the productivity effects of generative AI, and contributed to providing concrete data on how impactful these types of tools might be in the workplace and on the labor market,” Zhang says.

In other ongoing research — “Determinants of Irregular Worker Schedules” — Zhang is using data from a payroll provider to examine scheduling unpredictability, investigating why companies employ unpredictable schedules and how these schedules affect low-wage employees’ quality of life.

The scheduling project, conducted with MIT economics PhD student Nathan Lazarus, is motivated, in part, by existing sociological evidence that low-wage workers’ unpredictable schedules are associated with worse sleep and well-being. “We’ve seen a relationship between higher turnover and inconsistent, inadequate schedules, which suggests workers dis-prefer these kinds of schedules,” Zhang says.

At an academic roundtable, Zhang presented her results to Starbucks employees involved in scheduling and staffing. The attendees wanted to learn more about how different scheduling practices impacted workers and their productivity. “These are the kinds of questions that could reveal useful information for small businesses, large corporations, and others,” she says.

By conducting this research, Zhang hopes to better understand whether or not scheduling regulations can improve affected employees’ quality of life, while also considering potential unintended consequences. “Why are these schedules set the way they’re set?” she asks. “Do businesses with these kinds of schedules require increased regulation?”

Another project, conducted with MIT economics doctoral student Arjun Ramani, examines the linkages between offshoring, remote work, and related outcomes. “Do the technological and managerial practices that have made remote work possible further facilitate offshoring?” she asks. “Do organizations see significant gains in efficiency? What are the impacts on U.S. and offshore workers?”

Her work is being funded through the National Science Foundation Graduate Research Fellowship Program and the Washington Center for Equitable Growth.

Putting people at the center

Zhang has observed the different kinds of people economics and higher education could bring together. She followed a dual enrollment track in high school, completing college-level courses with students from across a variety of demographic identities. “I enjoyed centering people in my work,” she says. “Taking classes with a diverse group of students, including veterans and mothers returning to school to complete their studies, made me more curious about socioeconomic issues and the policies relevant to them.”

She later enrolled at MIT, where she participated in the Undergraduate Research Opportunities Program (UROP). She also completed an internship at the World Bank, worked as a summer analyst at the Federal Reserve Bank of New York, and worked as an assistant for a diverse faculty cohort including MIT economists David Autor, Jon Gruber, and Nina Roussille. Autor is her primary advisor on her doctoral research, a mentor she cites as a significant influence.

“[Autor’s] course, 14.03 (Microeconomics and Public Policy), cemented connections between theory and practice,” she says. “I thought the class was revelatory in showing the kinds of questions economics can answer.”

Doctoral study has revealed interesting pathways of investigation for Zhang, as have her relationships with her student peers and other faculty. She has, for example, leveraged faculty connections to gain access to hourly wage data in support of her scheduling and employee impacts work. “Generally, economists have had administrative data on earnings, but not on hours,” she notes.

Zhang’s focus on improving others’ lives extends to her work outside the classroom. She’s a mentor for the Boston Chinatown Neighborhood Center College Access Program and a member of MIT’s Graduate Christian Fellowship group. When she’s not enjoying spicy soups or paddling on the Charles, she takes advantage of opportunities to decompress with her art at W20 Arts Studios.

“I wanted to create time for myself outside of research and the classroom,” she says.

Zhang cites the benefits of MIT’s focus on cross-collaboration and encouraging students to explore other disciplines. As an undergraduate, Zhang minored in computer science, which taught her coding skills critical to her data work. Exposure to engineering also led her to become more interested in questions around how technology and workers interact.

Working with other scholars in the department has improved how Zhang conducts inquiries. “I’ve become the kind of well-rounded student and professional who can identify and quantify impacts, which is invaluable for future projects,” she says. Exposure to different academic and research areas, Zhang argues, helps increase access to ideas and information.

U.S. Air Force Maj. Adam Fuhrmann ’11 was one of 10 individuals chosen from a field of 8,000 applicants for the 2025 U.S. astronaut candidate class, NASA announced in a live ceremony on Sept. 22. 

This is NASA’s 24th class of astronaut candidates since the first Mercury 7 astronauts were chosen in 1959. Upon completion of his training, Fuhrmann will be the 45th MIT graduate to become a flight-eligible astronaut.

“As test pilots we don't do anything on our own, we work with amazing teams of engineers and maintenance professionals to plan, simulate, and execute complex and sometimes risky missions in aircraft to collect data and accomplish a mission, all while assessing risk and making smart calls as a team to do that as safely as possible,” Fuhrmann said at NASA’s announcement ceremony in Houston, Texas. “I'm happy to try to bring some of that experience to do the same thing with the NASA team and learn from everyone at Johnson Space Center how to apply those lessons to human spaceflight.”

His class now begins two years of training at the Johnson Space Center in Houston that includes instruction and skills development for complex operations aboard the International Space Station, Artemis missions to the moon, and beyond. Training includes robotics, land and water survival, geology, foreign language, space medicine and physiology, and more, while also conducting simulated spacewalks and flying high-performance jets.

From MIT to astronaut training

Fuhrmann, 35, is from Leesburg, Virginia, and has accumulated more than 2,100 flight hours in 27 aircraft, including the F-16 and F-35. He has served as a U.S. Air Force fighter pilot and experimental test pilot for nearly 14 years and deployed in support of operations Freedom’s Sentinel and Resolute Support, logging 400 combat hours.

Fuhrmann holds a bachelor’s degree in aeronautics and astronautics from MIT and master’s degrees in flight test engineering and systems engineering from the U.S. Air Force Test Pilot School and Purdue University, respectively. While at MIT, he was a member of Air Force ROTC Detachment 365 and was selected as the third-ever student leader of the Bernard M. Gordon-MIT Engineering Leadership Program (GEL) in spring 2011.

“We are tremendously proud of Adam for this notable accomplishment, and we look forward to following his journey through astronaut candidate school and beyond,” says Leo McGonagle, GEL founding and executive director.

“It’s always a thrill to learn that one of our own has joined NASA's illustrious astronaut corps,” says Julie Shah, head of the MIT Department of Aeronautics and Astronautics and the H.N. Slater Professor in Aeronautics and Astronautics. “Adam is Course 16’s 19th astronaut alum. We take very seriously the responsibility to provide the very best aerospace engineering education, and it's so gratifying to see that those fundamentals continue to set individuals from our community on the path to becoming an astronaut.”

Learning to be a leader at MIT

McGonagle recalls that Fuhrmann was a very early participant in GEL from 2009 to 2011.

“The GEL Program was still in its infancy during this time and was in somewhat of a fragile state as we were seeking to grow and cement ourselves as a viable MIT program. As the fall 2010 semester was winding down, it was evident that the program needed an effective GEL2 student leader during the spring semester, who could lead by example and inspire fellow students and who was an example of what right looks like. I knew Adam was already an emerging leader as a senior cadet in MIT’s Air Force ROTC Detachment, so I tapped him for the role of spring student leader of GEL,” said McGonagle.

Fuhrmann initially sought to decline this role, citing his time as a leader in ROTC. But McGonagle, having led the Army ROTC Program prior to GEL, felt that the GEL Student Leader role would challenge and develop Fuhrmann in other ways. In GEL, he would be charged with leading and inspiring students from a broad background of experiences, and focused exclusively on leading within engineering contexts, while engaging with engineering industry organizations.

“GEL needed strong student leadership at this time, so Adam took on the role, and it ended up being a win-win for both him and the program. He later expressed to me that the experience challenged him in ways that he hadn’t anticipated and complemented his Air Force ROTC leadership development. He was grateful for the opportunity, and the program stabilized and grew under Adam’s leadership. He was the right student at the right time and place,” said McGonagle.

Fuhrmann has remained connected to the GEL program. He asked McGonagle to administer his oath of commissioning into the U.S. Air Force, with his family in attendance, at the historic Bunker Hill Monument in Boston. “One of my proudest GEL memories,” said McGonagle, who is a former U.S. Army Lt. Colonel.

Throughout his time in service which included overseas deployments, Fuhrmann has actively participated in Junior Engineering Leader’s Roundtable leadership labs (ELLs) with GEL students, and he has kept in touch with his GEL2 cohort.

“Adam’s GEL2 cohort meets informally once or twice a year, usually via Zoom, to share and discuss professional challenges, lessons learned, life stories, to keep in touch with each other. This small but excellent group of GEL alum is committed to staying connected and supporting one another, as part of the broader GEL community,” said McGonagle.

At the center of nuclear reactors across the United States, a new type of chromium-coated fuel is being used to make the reactors more efficient and more resistant to accidents. The fuel is one of many innovations sprung from collaboration between researchers at MIT and the Idaho National Laboratory (INL) — a relationship that has altered the trajectory of the country’s nuclear industry.

Amid renewed excitement around nuclear energy in America, MIT’s research community is working to further develop next-generation fuels, accelerate the deployment of small modular reactors (SMRs), and enable the first nuclear reactor in space.

Researchers at MIT and INL have worked closely for decades, and the collaboration takes many forms, including joint research efforts, student and postdoc internships, and a standing agreement that lets INL employees spend extended periods on MIT’s campus researching and teaching classes. MIT is also a founding member of the Battelle Energy Alliance, which has managed the Idaho National Laboratory for the Department of Energy since 2005.

The collaboration gives MIT’s community a chance to work on the biggest problems facing America’s nuclear industry while bolstering INL’s research infrastructure.

“The Idaho National Laboratory is the lead lab for nuclear energy technology in the United States today — that’s why it’s essential that MIT works hand in hand with INL,” says Jacopo Buongiorno, the Battelle Energy Alliance Professor in Nuclear Science and Engineering at MIT. “Countless MIT students and postdocs have interned at INL over the years, and a memorandum of understanding that strengthened the collaboration between MIT and INL in 2019 has been extended twice.”

Ian Waitz, MIT’s vice president for research, adds, “The strong collaborative history between MIT and the Idaho National Laboratory enables us to jointly contribute practical technologies to enable the growth of clean, safe nuclear energy. It’s a clear example of how rigorous collaboration across sectors, and among the nation’s top research facilities, can advance U.S. economic prosperity, health, and well-being.”

Research with impact

Much of MIT’s joint research with INL involves tests and simulations of new nuclear materials, fuels, and instrumentation. One of the largest collaborations was part of a global push for more accident-tolerant fuels in the wake of the nuclear accident that followed the 2011 earthquake and tsunami in Fukushima, Japan.

In a series of studies involving INL and members of the nuclear energy industry, MIT researchers helped identify and evaluate alloy materials that could be deployed in the near term to not only bolster safety but also offer higher densities of fuel.

“These new alloys can withstand much more challenging conditions during abnormal occurrences without reacting chemically with steam, which could result in hydrogen explosions during accidents,” explains Buongiorno, who is also the director of science and technology at MIT’s Nuclear Reactor Laboratory and the director of MIT’s Center for Advanced Nuclear Energy Systems. “The fuels can take much more abuse without breaking apart in the reactor, resulting in a higher safety margin.”

The fuels tested at MIT were eventually adopted by power plants across the U.S., starting with the Byron Clean Energy Center in Ogle County, Illinois.

“We’re also developing new materials, fuels, and instrumentation,” Buongiorno says. “People don’t just come to MIT and say, ‘I have this idea, evaluate it for me.’ We collaborate with industry and national labs to develop the new ideas together, and then we put them to the test,  reproducing the environment in which these materials and fuels would operate in commercial power reactors. That capability is quite unique.”

Another major collaboration was led by Koroush Shirvan, MIT’s Atlantic Richfield Career Development Professor in Energy Studies. Shirvan’s team analyzed the costs associated with different reactor designs, eventually developing an open-source tool to help industry leaders evaluate the feasibility of different approaches.

“The reason we’re not building a single nuclear reactor in the U.S. right now is cost and financial risk,” Shirvan says. “The projects have gone over budget by a factor of two and their schedule has lengthened by a factor of 1.5, so we’ve been doing a lot of work assessing the risk drivers. There’s also a lot of different types of reactors proposed, so we’ve looked at their cost potential as well and how those costs change if you can mass manufacture them.”

Other INL-supported research of Shirvan’s involves exploring new manufacturing methods for nuclear fuels and testing materials for use in a nuclear reactor on the surface of the moon.

“You want materials that are lightweight for these nuclear reactors because you have to send them to space, but there isn’t much data around how those light materials perform in nuclear environments,” Shirvan says.

People and progress

Every summer, MIT students at every level travel to Idaho to conduct research in INL labs as interns.

“It’s an example of our students getting access to cutting-edge research facilities,” Shirvan says.

There are also several joint research appointments between the institutions. One such appointment is held by Sacit Cetiner, a distinguished scientist at INL who also currently runs the MIT and INL Joint Center for Reactor Instrumentation and Sensor Physics (CRISP) at MIT’s Nuclear Reactor Laboratory.

CRISP focuses its research on key technology areas in the field of instrumentation and controls, which have long stymied the bottom line of nuclear power generation.

“For the current light-water reactor fleet, operations and maintenance expenditures constitute a sizeable fraction of unit electricity generation cost,” says Cetiner. “In order to make advanced reactors economically competitive, it’s much more reasonable to address anticipated operational issues during the design phase. One such critical technology area is remote and autonomous operations. Working directly with INL, which manages the projects for the design and testing of several advanced reactors under a number of federal programs, gives our students, faculty, and researchers opportunities to make a real impact.”

The sharing of experts helps strengthen MIT and the nation’s nuclear workforce overall.

“MIT has a crucial role to play in advancing the country’s nuclear industry, whether that’s testing and developing new technologies or assessing the economic feasibility of new nuclear designs,” Buongiorno says.

MIT has placed second in U.S. News and World Report’s annual rankings of the nation’s best universities, announced today. 

As in past years, MIT’s engineering program continues to lead the list of undergraduate engineering programs at a doctoral institution. The Institute also placed first in five out of 10 engineering disciplines.

U.S. News placed MIT first in its evaluation of undergraduate computer science programs, ranking it No. 1 in four out of 10 computer science disciplines.

MIT also topped the list of undergraduate business programs, a ranking it shares with the University of Pennsylvania. Among business subfields, MIT is ranked first in two out of 10 specialties.

Within the magazine’s rankings of “academic programs to look for,” MIT topped the list in the category of undergraduate research and creative projects. The Institute also ranks as the second most innovative national university and the fourth best value, according to the U.S. News peer assessment survey of top academics.

MIT placed first in five engineering specialties: aerospace/aeronautical/astronautical engineering; chemical engineering; computer engineering; materials engineering; and mechanical engineering. It placed within the top five in two other engineering areas: biomedical engineering and electrical/electronic/communication engineering.

Other schools in the top five overall for undergraduate engineering programs are Stanford University, the University of California at Berkeley, Georgia Tech, Caltech, the University of Illinois at Urbana-Champaign, and the University of Michigan at Ann Arbor.

In computer science, MIT placed first in four specialties: artificial intelligence (shared with Carnegie Mellon University); biocomputing/bioinformatics/biotechnology; computer systems; and theory. It placed in the top five of six other disciplines: cybersecurity; data analytics/science; game/simulation development (shared with Carnegie Mellon); mobile/web applications; programming languages; and software engineering.

Other schools in the top five overall for undergraduate computer science programs are Carnegie Mellon, Stanford, UC Berkeley, Princeton University, and Georgia Tech.

Among undergraduate business specialties, the MIT Sloan School of Management leads in production/operations management and quantitative analysis. It also placed within the top five in five other categories: analytics; entrepreneurship; finance; management information systems; and supply chain management/logistics.

Other undergraduate business programs ranking in the top five include UC Berkeley, the University of Michigan at Ann Arbor, and New York University.

Recently, U.S. News & World Report ranked medium to large undergraduate economics programs based on a peer assessment survey; MIT’s economics program has placed first in this ranking.

MIT Department of Mathematics researchers David Roe ’06 and Andrew Sutherland ’90, PhD ’07 are among the inaugural recipients of the Renaissance Philanthropy and XTX Markets’ AI for Math grants. 

Four additional MIT alumni — Anshula Gandhi ’19, Viktor Kunčak SM ’01, PhD ’07; Gireeja Ranade ’07; and Damiano Testa PhD ’05 — were also honored for separate projects.

The first 29 winning projects will support mathematicians and researchers at universities and organizations working to develop artificial intelligence systems that help advance mathematical discovery and research across several key tasks.

Roe and Sutherland, along with Chris Birkbeck of the University of East Anglia, will use their grant to boost automated theorem proving by building connections between the L-Functions and Modular Forms Database (LMFDB) and the Lean4 mathematics library (mathlib).

“Automated theorem provers are quite technically involved, but their development is under-resourced,” says Sutherland. With AI technologies such as large language models (LLMs), the barrier to entry for these formal tools is dropping rapidly, making formal verification frameworks accessible to working mathematicians. 

Mathlib is a large, community-driven mathematical library for the Lean theorem prover, a formal system that verifies the correctness of every step in a proof. Mathlib currently contains on the order of 105 mathematical results (such as lemmas, propositions, and theorems). The LMFDB, a massive, collaborative online resource that serves as a kind of “encyclopedia” of modern number theory, contains more than 109 concrete statements. Sutherland and Roe are managing editors of the LMFDB.

Roe and Sutherland’s grant will be used for a project that aims to augment both systems, making the LMFDB’s results available within mathlib as assertions that have not yet been formally proved, and providing precise formal definitions of the numerical data stored within the LMFDB. This bridge will benefit both human mathematicians and AI agents, and provide a framework for connecting other mathematical databases to formal theorem-proving systems.

The main obstacles to automating mathematical discovery and proof are the limited amount of formalized math knowledge, the high cost of formalizing complex results, and the gap between what is computationally accessible and what is feasible to formalize.

To address these obstacles, the researchers will use the funding to build tools for accessing the LMFDB from mathlib, making a large database of unformalized mathematical knowledge accessible to a formal proof system. This approach enables proof assistants to identify specific targets for formalization without the need to formalize the entire LMFDB corpus in advance.

“Making a large database of unformalized number-theoretic facts available within mathlib will provide a powerful technique for mathematical discovery, because the set of facts an agent might wish to consider while searching for a theorem or proof is exponentially larger than the set of facts that eventually need to be formalized in actually proving the theorem,” says Roe.

The researchers note that proving new theorems at the frontier of mathematical knowledge often involves steps that rely on a nontrivial computation. For example, Andrew Wiles’ proof of Fermat’s Last Theorem uses what is known as the “3-5 trick” at a crucial point in the proof.

“This trick depends on the fact that the modular curve X_0(15) has only finitely many rational points, and none of those rational points correspond to a semi-stable elliptic curve,” according to Sutherland. “This fact was known well before Wiles’ work, and is easy to verify using computational tools available in modern computer algebra systems, but it is not something one can realistically prove using pencil and paper, nor is it necessarily easy to formalize.”

While formal theorem provers are being connected to computer algebra systems for more efficient verification, tapping into computational outputs in existing mathematical databases offers several other benefits.

Using stored results leverages the thousands of CPU-years of computation time already spent in creating the LMFDB, saving money that would be needed to redo these computations. Having precomputed information available also makes it feasible to search for examples or counterexamples without knowing ahead of time how broad the search can be. In addition, mathematical databases are curated repositories, not simply a random collection of facts. 

“The fact that number theorists emphasized the role of the conductor in databases of elliptic curves has already proved to be crucial to one notable mathematical discovery made using machine learning tools: murmurations,” says Sutherland.

“Our next steps are to build a team, engage with both the LMFDB and mathlib communities, start to formalize the definitions that underpin the elliptic curve, number field, and modular form sections of the LMFDB, and make it possible to run LMFDB searches from within mathlib,” says Roe. “If you are an MIT student interested in getting involved, feel free to reach out!” 

The artificial intelligence models that turn text into images are also useful for generating new materials. Over the last few years, generative materials models from companies like Google, Microsoft, and Meta have drawn on their training data to help researchers design tens of millions of new materials.

But when it comes to designing materials with exotic quantum properties like superconductivity or unique magnetic states, those models struggle. That’s too bad, because humans could use the help. For example, after a decade of research into a class of materials that could revolutionize quantum computing, called quantum spin liquids, only a dozen material candidates have been identified. The bottleneck means there are fewer materials to serve as the basis for technological breakthroughs.

Now, MIT researchers have developed a technique that lets popular generative materials models create promising quantum materials by following specific design rules. The rules, or constraints, steer models to create materials with unique structures that give rise to quantum properties.

“The models from these large companies generate materials optimized for stability,” says Mingda Li, MIT’s Class of 1947 Career Development Professor. “Our perspective is that’s not usually how materials science advances. We don’t need 10 million new materials to change the world. We just need one really good material.”

The approach is described today in a paper published by Nature Materials. The researchers applied their technique to generate millions of candidate materials consisting of geometric lattice structures associated with quantum properties. From that pool, they synthesized two actual materials with exotic magnetic traits.

“People in the quantum community really care about these geometric constraints, like the Kagome lattices that are two overlapping, upside-down triangles. We created materials with Kagome lattices because those materials can mimic the behavior of rare earth elements, so they are of high technical importance.” Li says.

Li is the senior author of the paper. His MIT co-authors include PhD students Ryotaro Okabe, Mouyang Cheng, Abhijatmedhi Chotrattanapituk, and Denisse Cordova Carrizales; postdoc Manasi Mandal; undergraduate researchers Kiran Mak and Bowen Yu; visiting scholar Nguyen Tuan Hung; Xiang Fu ’22, PhD ’24; and professor of electrical engineering and computer science Tommi Jaakkola, who is an affiliate of the Computer Science and Artificial Intelligence Laboratory (CSAIL) and Institute for Data, Systems, and Society. Additional co-authors include Yao Wang of Emory University, Weiwei Xie of Michigan State University, YQ Cheng of Oak Ridge National Laboratory, and Robert Cava of Princeton University.

Steering models toward impact

A material’s properties are determined by its structure, and quantum materials are no different. Certain atomic structures are more likely to give rise to exotic quantum properties than others. For instance, square lattices can serve as a platform for high-temperature superconductors, while other shapes known as Kagome and Lieb lattices can support the creation of materials that could be useful for quantum computing.

To help a popular class of generative models known as a diffusion models produce materials that conform to particular geometric patterns, the researchers created SCIGEN (short for Structural Constraint Integration in GENerative model). SCIGEN is a computer code that ensures diffusion models adhere to user-defined constraints at each iterative generation step. With SCIGEN, users can give any generative AI diffusion model geometric structural rules to follow as it generates materials.

AI diffusion models work by sampling from their training dataset to generate structures that reflect the distribution of structures found in the dataset. SCIGEN blocks generations that don’t align with the structural rules.

To test SCIGEN, the researchers applied it to a popular AI materials generation model known as DiffCSP. They had the SCIGEN-equipped model generate materials with unique geometric patterns known as Archimedean lattices, which are collections of 2D lattice tilings of different polygons. Archimedean lattices can lead to a range of quantum phenomena and have been the focus of much research.

“Archimedean lattices give rise to quantum spin liquids and so-called flat bands, which can mimic the properties of rare earths without rare earth elements, so they are extremely important,” says Cheng, a co-corresponding author of the work. “Other Archimedean lattice materials have large pores that could be used for carbon capture and other applications, so it’s a collection of special materials. In some cases, there are no known materials with that lattice, so I think it will be really interesting to find the first material that fits in that lattice.”

The model generated over 10 million material candidates with Archimedean lattices. One million of those materials survived a screening for stability. Using the supercomputers in Oak Ridge National Laboratory, the researchers then took a smaller sample of 26,000 materials and ran detailed simulations to understand how the materials’ underlying atoms behaved. The researchers found magnetism in 41 percent of those structures.

From that subset, the researchers synthesized two previously undiscovered compounds, TiPdBi and TiPbSb, at Xie and Cava’s labs. Subsequent experiments showed the AI model’s predictions largely aligned with the actual material’s properties.

“We wanted to discover new materials that could have a huge potential impact by incorporating these structures that have been known to give rise to quantum properties,” says Okabe, the paper’s first author. “We already know that these materials with specific geometric patterns are interesting, so it’s natural to start with them.”

Accelerating material breakthroughs

Quantum spin liquids could unlock quantum computing by enabling stable, error-resistant qubits that serve as the basis of quantum operations. But no quantum spin liquid materials have been confirmed. Xie and Cava believe SCIGEN could accelerate the search for these materials.

“There’s a big search for quantum computer materials and topological superconductors, and these are all related to the geometric patterns of materials,” Xie says. “But experimental progress has been very, very slow,” Cava adds. “Many of these quantum spin liquid materials are subject to constraints: They have to be in a triangular lattice or a Kagome lattice. If the materials satisfy those constraints, the quantum researchers get excited; it’s a necessary but not sufficient condition. So, by generating many, many materials like that, it immediately gives experimentalists hundreds or thousands more candidates to play with to accelerate quantum computer materials research.”

“This work presents a new tool, leveraging machine learning, that can predict which materials will have specific elements in a desired geometric pattern,” says Drexel University Professor Steve May, who was not involved in the research. “This should speed up the development of previously unexplored materials for applications in next-generation electronic, magnetic, or optical technologies.”

The researchers stress that experimentation is still critical to assess whether AI-generated materials can be synthesized and how their actual properties compare with model predictions. Future work on SCIGEN could incorporate additional design rules into generative models, including chemical and functional constraints.

“People who want to change the world care about material properties more than the stability and structure of materials,” Okabe says. “With our approach, the ratio of stable materials goes down, but it opens the door to generate a whole bunch of promising materials.”

The work was supported, in part, by the U.S. Department of Energy, the National Energy Research Scientific Computing Center, the National Science Foundation, and Oak Ridge National Laboratory.

The Martin Trust Center for MIT Entrepreneurship strives to teach students the craft of entrepreneurship. Over the last few years, no technology has changed that craft more than artificial intelligence.

While many are predicting a rapid and complete transformation in how startups are built, the Trust Center’s leaders have a more nuanced view.

“The fundamentals of entrepreneurship haven’t changed with AI,” says Trust Center Entrepreneur in Residence Macauley Kenney. “There’s been a shift in how entrepreneurs accomplish tasks, and that trickles down into how you build a company, but we’re thinking of AI as another new tool in the toolkit. In some ways the world is moving a lot faster, but we also need to make sure the fundamental principles of entrepreneurship are well-understood.”

That approach was on display during this summer’s delta v startup accelerator program, where many students regularly turned to AI tools but still ultimately relied on talking to their customers to make the right decisions for their business.

Students in this year’s cohort used AI tools to accelerate their coding, draft presentations, learn about new industries, and brainstorm ideas. The Trust Center is encouraging students to use AI as they see fit while also staying mindful of the technology’s limitations.

The Trust Center itself has also embraced AI, most notably through Jetpack, its generative AI app that walks users through the 24 steps of disciplined entrepreneurship outlined in Managing Director Bill Aulet’s book of the same name. When students input a startup idea, the tool can suggest customer segments, early markets to pursue, business models, pricing, and a product plan.

The ways the Trust Center wants students to use Jetpack is apparent in its name: It’s inspired by the acceleration a jetpack provides, but users still need to guide its direction.

Even with AI technology’s current limitations, the Trust Center’s leaders acknowledge it can be a powerful tool for people at any stage of building a business, and their use of AI will continue to evolve with the technology.

“It’s undeniable we’re in the midst of an AI revolution right now,” says Entrepreneur in Residence Ben Soltoff. “AI is reshaping a lot of things we do, and it’s also shaping how we do entrepreneurship and how students build companies. The Trust Center has recognized that for years, and we’ve welcomed AI into how we teach entrepreneurship at all levels, from the earliest stages of idea formation to exploring and testing those ideas and understanding how to commercialize and scale them.”

AI’s strengths and weaknesses

For the past few years, when the Trust Center’s delta v staff get together for strategic retreats, AI has been a central topic. The delta v program’s organizers think about how students can get the most out of the technology each year as they plan their summer-long curriculum.

Everything starts with Orbit, the mobile app designed to help students find entrepreneurial resources, network with peers, access mentorship, and identify events and jobs. Jetpack was added to Orbit last year. It is trained on Aulet’s “Disciplined Entrepreneurship” as well as former Trust Center Executive Director Paul Cheek’s “Startup Tactics” book.

The Trust Center describes Jetpack’s outputs as first drafts designed to help students brainstorm their next steps.

“You need to verify everything when you are using AI to build a business,” says Kenney, who is also a lecturer at MIT Sloan and MIT D-Lab. “I have yet to meet anyone who will base their business on the output of something like ChatGPT without verifying everything first. Sometimes, the verification can take longer than if you had done the research yourself from the beginning.”

One company in this year’s cohort, Mendhai Health, uses AI and telehealth to offer personalized physical therapy for women struggling with pelvic floor dysfunction before and after childbirth.

“AI has definitely made the entrepreneurial process more efficient and faster,” says MBA student Aanchal Arora. “Still, overreliance on AI, at least at this point, can hamper your understanding of customers. You need to be careful with every decision you make.”

Kenney notes the way large language models are built can make them less useful for entrepreneurs.

“Some AI tools can increase your speed by doing things like automatically sorting your email or helping you vibe code apps, but many AI tools are built off averages, and those can be less effective when you’re trying to connect with a very specific demographic,” Kenney says. “It’s not helpful to have AI tell you about an average person, you need to personally have strong validation that your specific customer exists. If you try to build a tool for an average person, you may build a tool for no one at all.”

Students eager to embrace AI may also be overwhelmed by the sheer volume of tools available today. Fortunately, MIT students have a long history of being at the forefront of any new technology, and this year’s delta v cohort featured teams leveraging AI at the core of their solutions and in every step of their entrepreneurial journeys.

MIT Sloan MBA candidate Murtaza Jameel, whose company Cognify uses AI to simulates user interactions with websites and apps to improve digital experiences, describes his firm as an AI-native business.

“We’re building a design intelligence tool that replaces product testing with instant, predictive simulations of user behavior,” Jameel explains. “We’re trying to integrate AI into all of our processes: ideation, go to market, programming. All of our building has been done with AI coding tools. I have a custom bot that I’ve fed tons of information about our company to, and it’s a thought partner I’m speaking to every single day.”

The more things change…

One of the fundamentals the Trust Center doesn’t see changing is the need for students to get out of the lab or the classroom to talk to customers.

“There are ways that AI can unlock new capabilities and make things move faster, but we haven’t turned our curriculum on its head because of AI,” Soltoff says. “In delta v, we stress first and foremost: What are you building and who are you building it for? AI alone can’t tell you who your customer is, what they want, and how you can better serve their needs. You need to go out into the world to make that happen.”

Indeed, many of the biggest hurdles delta v teams faced this summer looked a lot like the hurdles entrepreneurs have always faced.

“We were prepared at the Trust Center to see a big change and to adapt to that, but the companies are still building and encountering the same challenges of customer identification, beachhead market identification, team dynamics,” Kenney says. “Those are still the big meaty challenges they’ve always been working on.”

Amid endless hype about AI agents and the future of work, many founders this summer still said the human side of delta v is what makes the program special.

“I came to MIT with one goal: to start a technology company,” Jameel says. “The delta v program was on my radar when I was applying to MIT. The program gives you incredible access to resources — networks, mentorship, advisors. Some of the top folks in our industry are advising us now on how to build our company. It’s really unique. These are folks who have done what you’re doing 10 or 20 years ago, all just rooting for you. That’s why I came to MIT.”

In recent years, power outages caused by extreme weather or substation attacks have exposed the vulnerability of the electric grid. For the nation’s military bases, which are served by the grid, being ready for outages is a matter of national security. What better way to test readiness than to cut the power?

Lincoln Laboratory is doing just that with its Energy Resilience Readiness Exercises (ERREs). During an exercise, a base is disconnected from the grid, testing the ability of backup power systems and service members to work through failure. Lasting up to 15 hours, each exercise mimics a real outage event with limited forewarning to the base population.

“No one thought that this kind of real-world test would be accepted. We’ve now done it at 33 installations, impacting over 800,000 people,” says Jean Sack ’13, SM ’15, who leads the program with Christopher Lashway and Annie Weathers in the laboratory's Energy Systems Group.

According to a Department of Energy report, 70 percent of the nation’s transmission lines are approaching end of life. This aging infrastructure, combined with increasing power demands and interdependencies, threatens cascading failures. In response, the Department of Defense (DoD) has sharpened its focus on energy resilience, or the ability to anticipate, withstand, and recover from outages. On a base, an outage could disrupt critical missions, open the door to physical or cyberattacks, and cut off water supplies.

“Threats to this already-fragile system are increasing. That's why this work is so important,” Sack says. 

Safely cutting power

Before an exercise, the laboratory team works closely with base leadership and infrastructure personnel to carefully plan how it will safely disconnect from utility power. Over multiple site visits, they study each building and mission to understand power capabilities, ensure health and safety, and develop contingency plans.

“We get people together who may never have spoken before, but depend on one another. We like to say ‘connecting mission owners to their utility providers,’” says Lashway, a former electrician turned energy-systems researcher. “The planning process is a huge learning opportunity, and a chance to fix issues ahead of the outage.”

On the day of the outage, laboratory staff are on site to ensure the process runs smoothly, but the base is meant to run the exercise. Since beginning in 2018, the ERRE campaign has reached huge installations, including Fort Bragg, a U.S. Army base in North Carolina that sees nearly 150,000 people daily, and sites as far away as England and Japan.

The key is to not limit its scope. All facilities and missions, especially those that are critical, should be included, and service members are tasked with working through issues. To make exercises even more useful as an evaluation of readiness, some are modified with scripted scenarios simulating real-world incidents. These scenarios might challenge personnel to handle a cyberattack to control systems, shutdown of a backup power plant, or a rocket launch during an outage.

“We can do all the tabletop exercises in the world, but when you actually pull the plug, the question is, what actually goes on?” former assistant secretary of defense for sustainment Robert McMahon said at a joint House Armed Services subcommittee hearing about initial exercises. “Perhaps the most important lesson that I've seen is a lack of appreciation and understanding by our senior leaders at the installation level, all the way up to my level, of what we thought was going to happen versus what actually occurred, and then being able to apply those lessons learned.”

Illuminating issues

The ERREs have brought to light common issues across bases. One of them is a reliance on fragile or faulty backup systems. For example, electronic equipment experiences a hard shutdown if it isn't supported by a backup battery to bridge power transitions. In some instances, these battery systems failed or unexpectedly depleted due to age or generator issues. “We see a giant comms room drop out, and then phones and computers don’t work. It emphasizes the need for redundancies,” Lashway says.

Generators also present issues. Some fail because they aren’t regularly serviced or refueled through the long outage. Sometimes, personnel mistakenly assumed a generator would support their entire building, requiring reconfigurations after the fact. Air conditioning systems are often excluded from generator-supported emergency circuits, but rooms with a large number of computers generate a lot of heat, and overheated equipment quickly shuts down.

The exercises also unveiled interdependencies and chain reactions. In one case, a fire-suppression system accidentally went off, dousing a hangar in foam. The cause was a pressure drop at the same exact moment a switch reset.

“Executing an operation at this scale stresses how each of these factors need to work harmoniously and efficiently to ensure that the base, and ultimately missions, remain functional,” Lashway says.

Beyond resolving technical issues, the exercises have been valuable for practicing coordination and following chains of command. They’ve also revealed social challenges of operating through outages. For instance, some DoD guidance restricts the use of generators at daycare centers, so parents needed to coordinate care while maintaining their mission. 

After an exercise, the laboratory compiles all findings in a report for the base. It provides time stamps of significant events by building, identifies links between issues, and summarizes common problems site-wide. It then provides recommendations to address vulnerabilities. “Our goal is to provide as much justification as possible for the base to get the resources they need to fix a problem,” Sack says. 

The researchers also want to help bases prevent issues and avoid costly repairs. Recently, they’ve been using power meters to capture electrical data before, during, and after an exercise. These monitoring tools reveal power-quality issues that are otherwise hidden.

“Not all power is created equal, and standards must be followed to ensure equipment, especially specialized military equipment, operates properly and doesn’t get damaged over the long term. Power metering provides a view into that,” says Lashway.

Sparking resiliency ahead

Lincoln Laboratory’s ERRE campaign has resulted in legislation. In 2021, Congress passed a law requiring each military branch to perform at least five ERREs, or "Black Start Exercises," per year through 2027. That law was recently reauthorized until 2032. The team has transitioned the ERRE process to two private companies, as well as within the Air Force and Army, to conduct exercises in the coming years.

“It's very exciting that this got Congress' attention and has scaled across the DoD,” says Nick Judson, who leads the portfolio of energy, water, and natural hazard resilience efforts within the Energy Systems Group. “This idea started out as a way to enable change on DoD installations, and included a lot of difficult conversations about turning the power off to critical missions, and now we're seeing significant improvements to the readiness of bases and their missions.”

It may even be encouraging some healthy competition across the services, Lashway says. At a recent regional event in Colorado, three U.S. Space Force installations each vied to push the scope and duration of their exercises.

The team’s focus is now turning to related analysis, such as water resiliency. Water and wastewater systems are vulnerable to disruptions beyond power outages, including equipment failure, sabotage, or water source depletion.

“We are conducting tabletop exercises and workshops uniting stakeholders around the importance of water and wastewater systems to enable missions,” says Amelia Servi, who leads this work. “So far, we’ve seen great engagement from groups managing water systems who have been seeking funds to fix these aging systems, and from missions who have previously taken water for granted.”

They are also working on long-term energy planning, including ways for installations to be less dependent on the grid. One way is to install microgrids, which are self-sufficient systems that can tap into stored energy. According to Sack, microgrids are highly customized and complicated to operate, so one goal is to design a standardized system. The team's recent power-metering data is providing useful initial inputs into such a design.

The researchers are also considering how this work could improve energy resiliency for civilians. Large-scale exercises might not be feasible for the public, but they could be conducted in areas important to public safety, or in places that rely on military resources. During one exercise in Georgia, city residents partially depended upon a base's power plant, so that exercise included working with the city to ensure its resiliency to the outage.

“Striking that balance of testing readiness without causing harm is a big challenge in this field and a huge motivation for us,” Sack says. “We are encouraged by the outcomes. Our work is impacting the services at the highest level, rewriting infrastructure policy, and making sure people can better sustain operations during grid disruptions.”

When OpenAI introduced ChatGPT to the world in 2022, it brought generative artificial intelligence into the mainstream and started a snowball effect that led to its rapid integration into industry, scientific research, health care, and the everyday lives of people who use the technology.

What comes next for this powerful but imperfect tool?

With that question in mind, hundreds of researchers, business leaders, educators, and students gathered at MIT’s Kresge Auditorium for the inaugural MIT Generative AI Impact Consortium (MGAIC) Symposium on Sept. 17 to share insights and discuss the potential future of generative AI.

“This is a pivotal moment — generative AI is moving fast. It is our job to make sure that, as the technology keeps advancing, our collective wisdom keeps pace,” said MIT Provost Anantha Chandrakasan to kick off this first symposium of the MGAIC, a consortium of industry leaders and MIT researchers launched in February to harness the power of generative AI for the good of society.

Underscoring the critical need for this collaborative effort, MIT President Sally Kornbluth said that the world is counting on faculty, researchers, and business leaders like those in MGAIC to tackle the technological and ethical challenges of generative AI as the technology advances.

“Part of MIT’s responsibility is to keep these advances coming for the world. … How can we manage the magic [of generative AI] so that all of us can confidently rely on it for critical applications in the real world?” Kornbluth said.

To keynote speaker Yann LeCun, chief AI scientist at Meta, the most exciting and significant advances in generative AI will most likely not come from continued improvements or expansions of large language models like Llama, GPT, and Claude. Through training, these enormous generative models learn patterns in huge datasets to produce new outputs.

Instead, LuCun and others are working on the development of “world models” that learn the same way an infant does — by seeing and interacting with the world around them through sensory input.

“A 4-year-old has seen as much data through vision as the largest LLM. … The world model is going to become the key component of future AI systems,” he said.

A robot with this type of world model could learn to complete a new task on its own with no training. LeCun sees world models as the best approach for companies to make robots smart enough to be generally useful in the real world.

But even if future generative AI systems do get smarter and more human-like through the incorporation of world models, LeCun doesn’t worry about robots escaping from human control.

Scientists and engineers will need to design guardrails to keep future AI systems on track, but as a society, we have already been doing this for millennia by designing rules to align human behavior with the common good, he said.

“We are going to have to design these guardrails, but by construction, the system will not be able to escape those guardrails,” LeCun said.

Keynote speaker Tye Brady, chief technologist at Amazon Robotics, also discussed how generative AI could impact the future of robotics.

For instance, Amazon has already incorporated generative AI technology into many of its warehouses to optimize how robots travel and move material to streamline order processing.

He expects many future innovations will focus on the use of generative AI in collaborative robotics by building machines that allow humans to become more efficient.

“GenAI is probably the most impactful technology I have witnessed throughout my whole robotics career,” he said.

Other presenters and panelists discussed the impacts of generative AI in businesses, from largescale enterprises like Coca-Cola and Analog Devices to startups like health care AI company Abridge.

Several MIT faculty members also spoke about their latest research projects, including the use of AI to reduce noise in ecological image data, designing new AI systems that mitigate bias and hallucinations, and enabling LLMs to learn more about the visual world.

After a day spent exploring new generative AI technology and discussing its implications for the future, MGAIC faculty co-lead Vivek Farias, the Patrick J. McGovern Professor at MIT Sloan School of Management, said he hoped attendees left with “a sense of possibility, and urgency to make that possibility real.”