Transistors, the building blocks of modern electronics, are typically made of silicon. Because it’s a semiconductor, this material can control the flow of electricity in a circuit. But silicon has fundamental physical limits that restrict how compact and energy-efficient a transistor can be.

MIT researchers have now replaced silicon with a magnetic semiconductor, creating a magnetic transistor that could enable smaller, faster, and more energy-efficient circuits. The material’s magnetism strongly influences its electronic behavior, leading to more efficient control of the flow of electricity. 

The team used a novel magnetic material and an optimization process that reduces the material’s defects, which boosts the transistor’s performance.

The material’s unique magnetic properties also allow for transistors with built-in memory, which would simplify circuit design and unlock new applications for high-performance electronics.

“People have known about magnets for thousands of years, but there are very limited ways to incorporate magnetism into electronics. We have shown a new way to efficiently utilize magnetism that opens up a lot of possibilities for future applications and research,” says Chung-Tao Chou, an MIT graduate student in the departments of Electrical Engineering and Computer Science (EECS) and Physics, and co-lead author of a paper on this advance.

Chou is joined on the paper by co-lead author Eugene Park, a graduate student in the Department of Materials Science and Engineering (DMSE); Julian Klein, a DMSE research scientist; Josep Ingla-Aynes, a postdoc in the MIT Plasma Science and Fusion Center; Jagadeesh S. Moodera, a senior research scientist in the Department of Physics; and senior authors Frances Ross, TDK Professor in DMSE; and Luqiao Liu, an associate professor in EECS, and a member of the Research Laboratory of Electronics; as well as others at the University of Chemistry and Technology in Prague. The paper appears today in Physical Review Letters.

Overcoming the limits

In an electronic device, silicon semiconductor transistors act like tiny light switches that turn a circuit on and off, or amplify weak signals in a communication system. They do this using a small input voltage.

But a fundamental physical limit of silicon semiconductors prevents a transistor from operating below a certain voltage, which hinders its energy efficiency.

To make more efficient electronics, researchers have spent decades working toward magnetic transistors that utilize electron spin to control the flow of electricity. Electron spin is a fundamental property that enables electrons to behave like tiny magnets.

So far, scientists have mostly been limited to using certain magnetic materials. These lack the favorable electronic properties of semiconductors, constraining device performance.

“In this work, we combine magnetism and semiconductor physics to realize useful spintronic devices,” Liu says.

The researchers replace the silicon in the surface layer of a transistor with chromium sulfur bromide, a two-dimensional material that acts as a magnetic semiconductor.

Due to the material’s structure, researchers can switch between two magnetic states very cleanly. This makes it ideal for use in a transistor that smoothly switches between “on” and “off.”

“One of the biggest challenges we faced was finding the right material. We tried many other materials that didn’t work,” Chou says.

They discovered that changing these magnetic states modifies the material’s electronic properties, enabling low-energy operation. And unlike many other 2D materials, chromium sulfur bromide remains stable in air.

To make a transistor, the researchers pattern electrodes onto a silicon substrate, then carefully align and transfer the 2D material on top. They use tape to pick up a tiny piece of material, only a few tens of nanometers thick, and place it onto the substrate.

“A lot of researchers will use solvents or glue to do the transfer, but transistors require a very clean surface. We eliminate all those risks by simplifying this step,” Chou says.

Leveraging magnetism

This lack of contamination enables their device to outperform existing magnetic transistors. Most others can only create a weak magnetic effect, changing the flow of current by a few percent or less. Their new transistor can switch or amplify the electric current by a factor of 10.

They use an external magnetic field to change the magnetic state of the material, switching the transistor using significantly less energy than would usually be required.

The material also allows them to control the magnetic states with electric current. This is important because engineers cannot apply magnetic fields to individual transistors in an electronic device. They need to control each one electrically.

The material’s magnetic properties could also enable transistors with built-in memory, simplifying the design of logic or memory circuits.

A typical memory device has a magnetic cell to store information and a transistor to read it out. Their method can combine both into one magnetic transistor.

“Now, not only are transistors turning on and off, they are also remembering information. And because we can switch the transistor with greater magnitude, the signal is much stronger so we can read out the information faster, and in a much more reliable way,” Liu says.

Building on this demonstration, the researchers plan to further study the use of electrical current to control the device. They are also working to make their method scalable so they can fabricate arrays of transistors.

This research was supported, in part, by the Semiconductor Research Corporation, the U.S. Defense Advanced Research Projects Agency (DARPA), the U.S. National Science Foundation (NSF), the U.S. Department of Energy, the U.S. Army Research Office, and the Czech Ministry of Education, Youth, and Sports. The work was partially carried out at the MIT.nano facilities.

When an artificial intelligence image generator produces a portrait, whose work went into it? The question sits at the center of lawsuits, licensing deals, and proposed regulations worldwide. Artists want credit. Companies want clarity. Policymakers want a way to assign responsibility.

New work from a team of researchers at MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) suggests that for models trained on large datasets, the question may often have no answer. It's not that the tools for finding it are inadequate. The connection itself has disappeared.

The scientists identified a phenomenon they call attribution decay, where the more data a generative model is trained on, the less any individual training example matters to any particular output. It feels counterintuitive, but at sufficiently large scales, they find, you can often remove any single image from the training data, or every image by a given artist, or every photograph of a given person, and the generated sample doesn't change.

And if removing something changes nothing, the researchers argue, it can't be said to be responsible for anything. 

"If you take away a piece of data and the output of the model doesn't change, then that piece of data didn't affect the output," says Zheng Dai SM ’21, PhD ’24, former MIT CSAIL researcher and lead author on the work. "So it doesn't make much sense to attribute the output to that piece of data. And if you then do this one at a time for every other piece of data and find that the output doesn’t change for any of them either, then it doesn't make much sense to attribute the output to any one of them."

"All previous methods were approximate," says MIT Professor David Gifford, who is an MIT CSAIL principal investigator. "They really could not absolutely show that deleting individual things did not change the output. This paper introduces the first method that is absolute. You're actually deleting the inputs and deleting all influences of the inputs. This is the first exact method for doing large-scale deletion efficiently and showing that the results don't change."

Dai and Gifford's project is described in an open-access paper published today in Nature Communications.

The retraining problem

Testing this idea directly meant answering a what-if question. What would this model have produced if it had never seen this particular image? Answering it honestly means retraining the model from scratch without that image, then doing it again for the next image, and the next. With millions of training examples, the math quickly becomes prohibitive, which is why prior work in the attribution field has relied on approximations that estimate a training example's influence, rather than actually removing it.

Their workaround is an architecture they built themselves, called a "diffusion ensemble." Instead of one monolithic model, it's made up of many smaller components, each trained on a different slice of the data. Want to know what the model would do without a particular image? Just switch off the parts that saw it. No retraining, no approximation. What's left is a true counterfactual model, not an estimate of one.

Of course, a clever architecture only matters if it still works as a generator. So the team put the ensembles head to head with 24 conventional diffusion models trained on the exact same data. The images came out looking about as good by standard measures. 

One nice surprise in the numbers: The more training data, the better the ensembles held up against their single-model counterparts, a hint that they may actually be more data-efficient.

"When you have low amounts of data, they do very poorly," says Dai. "But if you have more data, it actually scales better compared to the vanilla diffusion model." 

Exploring a counterfactual universe

With ablation working, the researchers could finally ask their question at scale. Take one generated image, then imagine every alternate version of it, each produced by removing a different piece of the training data. The team calls this the image's counterfactual universe. The distance between the original and its most different alternate, the counterfactual radius, captures the most that any single piece of training data could have mattered.

They trained 24 ensembles on datasets from 256 images to more than 160,000, pulled from seven public collections including CIFAR-10, CelebA, MetFaces, and ArtBench. The pattern was consistent: The bigger the training set, the smaller the radius, shrinking along an inverse power law. It held whether differences were measured pixel by pixel or by semantic meaning, with statistical significance both ways.

The team also stress-tested their own result. Maybe ablation itself was the culprit? They redid it the brute-force way at small scale, training 1,282 separate models, and the decay showed up anyway. Maybe bigger datasets just make each removal proportionally smaller? They pinned the removed fraction in place, and it persisted. Fixed epochs, text-prompted models, class-conditioned models, four similarity metrics — the finding survived everything.

The privacy paradox

The implications run in a direction that surprised the researchers themselves.

Gifford sees the finding as bearing directly on the legal question of whether model outputs are derivative works. 

"One way to think about this is that these models are creative. They are not simply copying what they are fed, but creating brand new outputs. If those outputs have nothing to do with any individual piece of training data, that raises questions about fair use, about whether the outputs are themselves copyrightable as novel works, and about how authors get compensated when what comes out of a model isn't attributable to anything on the internet." 

Gifford also notes that the work shows how to produce outputs that are guaranteed to be unattributable, a capability he frames as an obligation for the industry, rather than a loophole. 

"In order for these companies to claim their outputs aren't derivative of the internet in a copyright-infringing way, they need to revise their models to take advantage of the advances in this work, so they can show they're not creating derivatives of individual people or items."

The work looks at diffusion models, now dominant in generating audiovisual media and prevalent in scientific applications including protein structure modeling and therapeutic discovery. Whether the same decay holds for the large language models at the center of the highest-profile copyright litigation is still an open question.

"If attribution worked, it would reliably tell us whether similarities between a model's output and a copyright-protected work are due to copying or coincidence," says James Grimmelmann, a law professor at Cornell Law School and Cornell Tech. "But this paper provides reason to think that attribution will fail for interesting models. Instead, technologists and courts will need to resort to other methods for assessing copying."

Dai and Gifford's work was supported by Schmidt Futures. 

MIT Principal Research Scientist and Haystack Observatory Assistant Director Emerita Anthea J. Coster was awarded the prestigious Appleton Prize at the International Union of Radio Science (URSI) General Assembly and Science Symposium in Krakow, Poland, on Aug. 16. 

The Appleton Prize recognizes career achievements and outstanding contributions to studies in ionospheric physics. Appleton awardees are regarded as pillars of the URSI atmospheric science community; the citation for Coster, an URSI Fellow, is for “pioneering research in GNSS [Global Navigation Satellite System] science, developing techniques to provide global-scale view of storm responses in the ionosphere, operationalizing novel algorithms, and providing novel ionospheric products to the community.” 

The Appleton Prize honors Sir Edward Victor Appleton, a Nobel Prize–winning physicist and former president of URSI (1934–52) who proved the existence of the ionosphere.

Coster joined MIT in 1984, originally at MIT Lincoln Laboratory, where she worked on satellite tracking applications within the Space Surveillance Complex situated at MIT Haystack Observatory. While at Lincoln, she was introduced to the Global Positioning System (GPS), the first GNSS; her GPS research at Lincoln eventually led to an appointment in Haystack’s geospace and atmospheric science research group. She continued and expanded her Lincoln-based GNSS research, focusing on ionospheric and atmospheric applications. At Haystack, Coster started as a research scientist, becoming an MIT principal research scientist in 2012; she also served as assistant director for the observatory from 2015 until 2024. 

Her career research focus spans the physics of the ionosphere, magnetosphere, and thermosphere, covering space weather and storm-time effects and coupling of these atmospheric regions, with particular expertise on GNSS positioning and measurement accuracy. Coster’s breakthrough contributions in GNSS applications to frontier geospace research span many areas, including ionosphere-magnetosphere coupling and mid-latitude ionospheric dynamics. A selected number of her accomplishments include the first real-time GNSS ionospheric monitoring system, as well as pioneering work in monitoring tropospheric water vapor with GNSS signals. She also was responsible for the first GNSS observations of storm-enhanced density, a bright and important feature that can span the heavily populated continental United States, with significant impacts to the Federal Aviation Administration Wide Area Augmentation System, which supplements traditional GPS navigation systems.

MIT Haystack Observatory director Phil Erickson says, "Dr. Coster's award from the International Radio Science Union is most well-deserved, and reflects her substantial international impact on the field of geospace remote sensing. Coster's pioneering application of GNSS signals to global and precise maps of total ionospheric electron density has produced a rich and insightful scientific output that anchors and greatly complements the multi-messenger, sensor fusion techniques at the forefront of the research field in near-Earth space weather dynamics. These areas are of critical importance to our increasingly spacefaring civilization."

Coster’s career also encompasses a lifetime of professional service contributions to the U.S. and international geophysical sciences community, including many leadership positions with the U.S. chapter of the Union of Radio Science, the Institute of Navigation, and the American Geophysical Union. She has served as co-chair of NASA's Living with a Star Program Analysis Group and is a current member of the U.S. National Academies of Science, Medicine, and Engineering Space Weather Roundtable. 

She is an author or co-author on more than 200 peer-reviewed publications, and is the principal investigator of numerous federal scientific grants from NASA, the National Science Foundation, the Office of Naval Research, and the Air Force Office of Scientific Research. Prominent results of Coster’s work are heavily used, including scientifically rich GNSS total electron content (TEC) and scintillation data products available to the research community through NSF's CEDAR Madrigal database and the Millstone Hill Geospace Facility. 

Coster has also made a number of notable contributions to science outreach, such as deploying radio instrumentation with MIT graduate students in Brazil and Peru, presenting outreach talks to high school and middle school students in Rwanda and Zambia, and installing GNSS receivers in Inuit villages and along the remote Steese Highway in Alaska. For many years, she has taught U.N.-sponsored GNSS workshops aimed at workforce education and career advancement in disadvantaged countries.

Originally from Texas, Coster attended the University of Texas at Austin as an undergraduate and earned her master's and doctorate degrees at Rice University in Houston, where she was involved with ionospheric experiments at the Arecibo Observatory in Puerto Rico. She moved to Massachusetts in 1984 to join MIT Lincoln Laboratory. 

"Anthea Coster has made seminal contributions to the state of the profession, enabling the international science community to conduct ionospheric research at spatio-temporal scales that were previously unachievable," says Larisa Goncharenko, assistant director and head of the atmospheric and geospace group at Haystack. "Her pioneering work on introducing and relating GPS measurements to fundamental research has led the community to employ GNSS as an information-rich sensor for ionospheric remote sensing and space weather monitoring. Her effort enabled countless discoveries in the near-Earth space environment that has become increasingly important for human activities in space. I am truly in awe of Anthea's pioneering accomplishments, and incredibly proud of her receiving the Appleton Prize."

With this award, MIT Haystack Observatory is now home to three URSI prize recipients. Former director and research scientist John Evans received the Appleton Prize in 1975 with a citation for "ionospheric physics, including application of the incoherent scatter technique," and research scientist Alan Rogers received the 2008 John Howard Dellinger Gold Medal for outstanding contributions to radio astronomy. 

You’ve heard of gig workers, freelancers, and temporary employees. But do you know about marginal workers? 

Accounting for about one in six U.S. jobs, it’s a huge category of people, who are going nowhere fast in the workplace — and don’t really have much say about that. 

“Marginal workers are employees who have no career prospects at their organizations,” says MIT Professor Emeritus Paul Osterman, author of a new book on the subject. “They are employees of the organization for whom they work, but the organization does not intend to keep them, and these workers are much less attached to any career ladder.” 

As such, marginal workers are part of a larger trend in U.S. employment. According to Osterman’s analysis, 35 percent of U.S. workers are either marginal employees, freelancers, contractors, or gig employees finding work on online platforms like ridesharing services. 

“That’s a big number,” says Osterman, who is the Nanyang Technological University Professor Emeritus at the MIT Sloan School of Management, where he is also a professor emeritus of work and organization studies. “That’s over 55 million people in the American work force.”

Osterman scrutinizes this employment landscape in his new book, “Disposable Workers: The Transformation of Employment,” published this month by Harvard University Press. In it, he examines the different categories of “disposable” workers in the U.S., while making the case that they are all part of a still-growing movement by firms to control labor costs, leaving many workers in precarious positions. 

“I wanted to present a unified way of thinking about these trends,” Osterman says.

Cutting costs

Osterman is a longtime labor economist and author of several previous books, whose work has often focused on job quality and labor-market fairness. 

He was motivated to write “Disposable Workers,” he says, because of how significantly marginal workers have been overlooked. Indeed, the category and term “marginal workers” comes from Osterman. 

In researching the book, Osterman conducted an original survey of over 6,000 workers, which helped shed light on the concept of marginal workers. They can fit a range of professions: staff attorneys at a law firm, adjunct faculty, and many kinds of part-time employees with few opportunities for advancement.

Overall, Osterman finds that about 17 percent of U.S. employees are marginal workers. Roughly 12 percent are contract workers, who are often employed by staffing agencies but then assigned to work at varying locations. Another 5 percent are organizational freelancers, working for firms without being part of the permanent staff. This includes gig workers, who account for a little more than 1 percent of the workforce and draw work from online platforms such as rideshare services. (Beyond this, there are also freelancers who work individually for multiple clients.) 

The common denominator among these categories is that each has evolved as a result of firms trying to cut back on labor expenses while trying to gain flexibility and more managerial discretion. The result is fewer workers with promotion prospects, health benefits, and employment stability.

“I’m putting the discussion of freelancing, contracting, and marginal workers into a coherent story that shows they’re all of a piece, they’re all part of the same thing, in terms of how employers are thinking about it,” Osterman says. 

Long term versus short term

How employers think about it, to be clear, revolves primarily around employee costs. By deploying employees in a variety of marginal, freelance, and contract roles, and making some of those positions part-time, businesses have constructed a system in which fewer employees have rising wages or additional benefits, and the portion of firm revenues plowed back into paying for workers can shrink. 

“This is not a book that argues that there’s dishonesty or that anyone’s evil, but at the end of the day, firms only care about one thing, which is to maximize profits, period, end of story,” Osterman says. 

He adds: “I’m very careful to say it’s a good thing that firms create jobs and develop new products — all good.” Still, he notes, for people who prioritize the plight of workers, the expansion of a disposable work force is a significant issue. 

To be sure, many scholars have found that short-term labor cost reductions can be counterproductive. Many firms have appeared to benefit from having a more stable, committed, motivated work force, which seems to result in greater productivity. What Osterman finds is that firms are likely aware of this tradeoff, and still willing to have a less-committed, less-expensive staff.

“The firms are obviously making a decision that the costs outweigh the value of commitment,” Osterman says. He also notes that the evidence on the matter is not entirely clear-cut. 

“There’s a debate on both sides of that question,” Osterman says. “I can’t prove that firms are being smart or stupid. But I can just tell you what they’re doing. And what they’re doing is making the decision that they benefit from having a large fraction of their workforce be disposable.”

Making the issue matter

“Disposable Workers” has drawn praise from other scholars. David Weil, a professor in the Heller School for Social Policy and Management and the Department of Economics at Brandeis University, has called it “a carefully researched and engaging book documenting the degradation of employment in recent decades.” 

Indeed, as “Disposable Workers” makes clear, the workplace has been challenging for many employees for a while now. Add artificial intelligence into this setting, and the outlook would seem to get even tougher for employees. Indeed, Osterman thinks AI could increase the use of disposable workers, if only for indirect reasons.

“I think this trend is going to be exacerbated by AI, because AI introduces a lot of uncertainty to firms about what their staffing needs are, and if firms are uncertain, they’re going to want disposable workers,” Osterman says. However, he emphasizes, “Disposable Workers” is not a book about AI. 

In any case, if jobs in the U.S. have become more precarious, what can be done to reverse that trend? One answer might be more expansive worker protections stemming from union negotiations. But these days, Osterman notes, only about 6 percent of U.S. employees are in a union, so that will only go so far.

Still, Osterman points out that nonunion organizations can help the situations of workers, such as the advocacy groups that lobbied for a $15/hour minimum wage in many places several years ago. 

Then too, he observes, sometimes customer pressure gets firms, even large multinationals, to improve working conditions, either for the firm’s own workers, or along its supply chain. 

“There is no magic solution,” Osterman says. “There is a set of tools.”

A key reason he wrote “Disposable Workers” is to bring attention to the topic in the first place, and the full extent to which the U.S. now has a workforce without much security or prospects of upward mobility. Without recognition of that point, no effort to change things will unfold, Osterman believes.

“The bigger policy point is: This issue has to become salient,” Osterman says. “If it does, then public and political pressure will come to bear on firms. If it doesn’t, then it won’t.” 

Shannon Knight attributes her interest in neuroscience to an experience she had in high school. She and her sister attended a medical day for students at the nearby University of Illinois Chicago. As they were on their way out of the event, they walked past a room with a person holding a brain.

“We stopped and backpedaled into the room, and I was so fascinated,” says Knight. “I was able to hold the brain of a patient who had passed away of Alzheimer’s. The brain holds so much emotion, decision-making — everything. I realized that this man’s entire memory was in my hands, and something clicked for me. I decided that I really wanted to learn much more about this organ.”

Now in her sixth year of doctoral studies at MIT’s McGovern Institute for Brain Research, Knight is working on developing a novel gene therapy for childhood-onset epilepsy, specifically SYNGAP1 haploinsufficiency. This rare genetic disorder is caused by a mutation in the SYNGAP1 gene, rendering one of the two copies of the gene nonfunctional. 

SYNGAP1 is important for brain development and neuronal communication, and the disorder leads to seizures in children starting as young as 4 months old. Other symptoms include intellectual disabilities, challenges with eating and sleeping, and difficulties with movement.

While there are currently methods to address the symptoms of the disorder, such as anti-seizure medications and dietary restrictions, as the child ages, the seizures often become resistant to medications. Knight is working to develop a therapeutic using CRISPR, a biotechnology tool used to edit genes. This therapeutic aims to address the root cause of this medication resistance by focusing on the gene itself.

“The idea of leading science with empathy is something that I feel very deeply,” she says. “I hope my efforts in the lab work toward the benefit of the people affected, rather than just for the benefit of my own science.”

Researching gene therapies

Knight’s interest in the brain flourished as a neuroscience major at Bowdoin College, working with Professor Hadley Horch. While she had originally planned to be pre-med, Knight ultimately decided that it wasn’t the best fit. She enjoyed the research she did as part of her honors thesis, exploring the regeneration of neurons in the auditory system of crickets, and decided that she wanted to pursue more research in molecular neuroscience, as well as genetics.

After graduating, Knight worked at the Perrimon Lab at Harvard University, where she first learned about CRISPR, applying it in a fruit fly model. She worked for two years in the lab, co-authoring a few papers and applying to graduate schools. 

She ultimately landed in the lab of MIT Professor Guoping Feng, studying the potential of utilizing CRISPR to develop a gene therapy treatment for Phelan-McDermid Syndrome, a rare genetic disorder caused by a deletion or mutation on the 22nd chromosome.

“Many of our graduate students are passionate about making a positive impact to society through cutting-edge research, and Shannon is a perfect example,” says Feng, the James W. and Patricia T. Poitras Professor and associate director at the McGovern Institute. “She is developing gene therapy technologies that have the potential to help many kids with devastating neurodevelopmental disorders.”

Building off of the gene therapy research around Phelan-McDermid syndrome, which is now in clinical trials in patients, Knight is now in the early phases of testing gene therapy for SYNGAP1 disorder. The goal is to go through the same process for the SYNGAP1 gene therapy as for the Phelan-McDermid gene therapy — eventually obtaining U.S. Food and Drug Administration approval and beginning clinical trials. 

The testing of the gene therapy on mice with a version of SYNGAP1 disorder has shown promising preliminary results in alleviating seizures and all of the behavioral phenotypes. This work is being accelerated by the Rare Brain Disorders Nexus, an MIT initiative that launched in the fall of 2025.

“Something I think about a lot is the idea of who ‘deserves’ the attention of a gene therapy. I feel that, regardless of how rare a genetic disorder might be, it still deserves care,” says Knight. “SYNGAP1 disorder is extremely rare, only impacting one to four out of every 10,000 children. I am very fortunate to be at an institution like MIT that has so many labs and brilliant researchers working on diseases that impact large portions of society, and it was really important to me to spend my PhD years helping a small, often unseen population. Although I don’t actually have a relationship with someone who has SYNGAP1 disorder, I know so many people who feel invisible in systems, and it is really important to me to be able to focus on people who feel unseen and give them hope.”

Inspiring others in the lab

In addition to her passion for neuroscience and genetic research, Knight has also developed a love of teaching. She has been a teaching assistant for 9.12 (Experimental Molecular Neurobiology), leading the lab portion of the course. She has enjoyed working closely with small classes of students, introducing them to the fundamentals of neuroscience lab research.

“We walked through the process of looking at a specific protein in neurons, and talked about how you can go from cell culture all the way up to a mouse brain — and all the steps in between. It was so important to me to be able to teach the students and help them to consider all of the different types of experiments they could do,” she says. “I’ve talked to many of the students since then, and many said it was one of their favorite classes.”

Knight received the Goodwin Medal in 2025 in recognition of her commitment to excellent teaching.

“Shannon has a rare combination of scientific excellence, teaching talent, and compassion,” says Laura Frawley, senior lecturer and teaching and curriculum development specialist in the Department of Brain and Cognitive Sciences. “Students trust her because she is approachable and invested in their success, and they learn from her because she has an exceptional ability to make complex ideas accessible and engaging. Her influence extends far beyond the laboratory skills she teaches.”

Knight has also invited high school and other college students into the lab and worked with them during the summers.

“It’s so exciting to bring in kids with no previous experience in a wet lab, and watch them be so amazed by all of the things that you can do,” she says. “Experiments that might seem so routine and relatively simple to me, at this point, are so exciting for them.”

Following the completion of her PhD program, Knight plans to do postdoctoral research and would ultimately like to be a faculty member at a small liberal arts college.

“It’s amazing to see students gain confidence over time, and then seeing them progress in their careers as scientists,” she says. “That’s very rewarding for me.”

A functional nervous system depends on the cooperation of many kinds of cells. So as developing organisms build their nervous systems, their neurons must take on different forms and functions to fulfill their designated roles. That carefully orchestrated process gives rise to thousands of different cell types in the human brain.

In the tiny worm known as C. elegans, the nervous system is far simpler, comprising a mere 118 classes of neurons.

At MIT, scientists in H. Robert Horvitz’s lab are studying the worms to learn about how nervous systems develop. Horvitz is the David H. Koch Professor of Biology at MIT, an investigator at the McGovern Institute for Brain Research at MIT, and an investigator at the Howard Hughes Medical Institute. His team has just discovered that a protein complex called cohesin, which helps shape the three-dimensional structure of the genome in both worms and humans, is critical for establishing some neurons’ identities as development unfolds.

The open-access findings, reported July 31 in the journal Science Advances, could help scientists find a way to treat a rare developmental disorder called Cornelia de Lange syndrome, which is caused by mutations that interrupt the cohesin complex.

Model organism

MIT postdoc Dongyeop Lee explains that C. elegans is a powerful model for studying neurodevelopment not just because its nervous system has been comprehensively mapped, but also because of the ease and speed with which scientists can study the function of its genes.

Because many of the worm’s genes have been retained through evolution, findings from studies of C. elegans often reveal important aspects of human biology. The current study began with worms that, because of a genetic mutation, make too many neurons of a certain type.

Adrenergic neurons, named for the kind of neurotransmitter they use to communicate with other neurons, are vital for enabling worms to respond to both their environment and their own internal state. Normally, C. elegans has just two pairs of adrenergic neurons: two RIM neurons and two RIC neurons. But the worms Lee studied had extras of both.

Takashi Hirose, a former member of the Horvitz lab, first observed this change in 2007.

Lee later continued the study and discovered that worms carrying a mutation in a gene called coh-1 have extra adrenergic neurons. The coh-1 gene encodes one part of the cohesin complex.

When Lee tested other mutations that disrupt cohesin, he found the same effect: Worms without fully functional cohesin had too many RIM neurons and too many RIC neurons.

Molecular switch

With a series of experiments designed to tease apart how cohesin impacts neurons’ identities, Lee discovered that cohesin cooperates with a gene-regulating protein called EOR-1 (known in humans as PLZF) to direct some neurons to develop into neurons that communicate with the inhibitory neurotransmitter GABA.

By reorganizing the structure of the genome, cohesin can change the way gene regulators like EOR-1 interact with DNA. Lee’s experiments showed that when either cohesin or EOR-1 couldn’t do its job, cells that should have become GABA-producing neurons become adrenergic neurons instead.

“What we found is that there are two alternative possible fates of certain neurons, and cohesin acts as a molecular switch that decides one of the possible neuronal fates,” Lee explains. “This means the structure of genomic DNA in the nucleus is important for neuronal fate determination.”

Disease connection

Lee adds that extra adrenergic neurons were not the only abnormality he observed in worms with cohesin mutations. Cohesin is important for shaping cells and tissues throughout the body. “The mutants have severe developmental defects,” Lee says. “They grow slowly. They don’t move well, and they also have defects in reproduction.”

Notably, the problems Lee saw in the worms echo aspects of Cornelia de Lange syndrome, a rare genetic disorder that impacts physical, cognitive, and behavioral development. Cornelia de Lange syndrome can be caused by mutations in cohesin genes, and Lee says that the discovery of how cohesin mutations affect worm development and behavior opens new opportunities to study the disease and search for potential therapeutic targets in C. elegans.

The Horvitz lab already has some promising leads. Taking advantage of the quick genetic screens that are possible in worms, Lee has found additional mutations that can counteract impaired cohesin, improving the health of worms with cohesin mutations. The team is now working to identify the genes where these suppressor mutations occur, so they can investigate whether they might make good therapeutic targets in humans.

Meanwhile, the team is also exploring a potential role for cohesin in shaping the fates of other neuron types, as well as searching broadly for additional molecules that work with cohesin to guide development. “We expect we have opened up a new biology,” Lee says. “This paper is just the beginning.”

Innovation is a concept that has become mythologized in the modern era: what it is, how to manage it, how to teach it, and how to get it to work for us. Despite these explorations, it remains fundamentally misunderstood, writes Eugene Fitzgerald, the Merton C. Flemings SMA Professor in MIT’s Department of Materials Science and Engineering, in his latest book, “The Invisible Engine: Why Innovation Evades Control.”

Fitzgerald draws on a decade of work leading international research programs, including the MIT and Masdar Institute Cooperative Program and the MIT-Singapore Alliance for Research and Technology, where he explored innovation as the integration of market applications, technology, and implementation.

Written at a moment when artificial intelligence is reshaping how we think about knowledge, research, and innovation, “The Invisible Engine” examines a deeper question: How does innovation actually happen, and how should society invest in it?

In this interview, Fitzgerald discusses his own experiences with innovation — including his co-invention of strained silicon at AT&T Bell Laboratories in the 1990s, which helped extend Moore’s Law, the semiconductor industry’s long-standing trend of increasing the number of transistors on chips roughly every two years — while exploring common misconceptions about innovation, how to create the conditions for it, and novel ways to prepare institutions for future uncertainty.

Q: What inspired you to write this book? 

A: The book really grew out of the last 10 years of work in research-to-market activity, from the MIT Masdar program to the MIT-Singapore Alliance. In science, we have professional journals and things like that that capture discoveries within individual fields, but these larger-scale projects — where science, economics, industry, and society all intersect — don’t really have an academic thread that connects them.

I wanted to write a book that condensed all of those connections, because the innovation process at that scale is really the intersection of many different fields. The dominant ones are science and economics, because those are the underlying principles that drive how innovation happens.

So I was interested in marking this moment in history and documenting the experiments we’ve done at scale — trying to understand how knowledge of the innovation process can be incorporated into large collaborative research programs.

What started as a practical effort to make these programs work became a broader and somewhat unexpected interest in the innovation process itself.

Q: What is the “invisible engine?” 

A: The invisible engine is this decentralized collective intelligence of different actors, which are people and companies that eventually create surprise in the marketplace, which brings great profit.

This concept of “surprise” comes from Frank Knight, an economist from the early 1900s who was trying to understand the Industrial Revolution happening around him. So he takes a close look at the entrepreneur and asks, “What does the entrepreneur do?” And his answer is that the entrepreneur takes on uncertainty. They bring something into the world without knowing exactly what will happen, and their reward is surprise — everyone is surprised that people want it and that it can be done. Because the entrepreneur is the first to discover that opportunity, they can earn a profit.

Q: How did your experience developing semiconductor technologies shape the ideas in the book? 

A: It started with Bell Labs. My colleague and I made an important discovery — we found a way of straining silicon in a thin-film form with very few defects, which had never been done before. From the physics point of view, it was a big result. But I was always interested in having impact in the world, not just scientific recognition, so I went to my manager and asked, “What do we do next?”

He said, “Go talk to the marketing people at AT&T.” In hindsight, that made perfect sense. Bell Labs, like a lot of great industrial labs, created a lot of stuff, but they couldn’t always commercialize it.

Then I came to MIT, which was an open aperture after Bell Labs. Here I could keep uncertainty open across all the elements and find convergence in different directions. Eventually I started a company, and going between institutions to stimulate things was an eye-opening experience. We eventually reached a settlement with Intel over a patent dispute because the industry discovered that strained silicon was needed to extend Moore’s Law — something we never expected.

A lot of people want things to be organized and say, “Oh yeah, look at all that chaos.” But no — the path from Bell Labs to MIT to a startup, and then to industry adoption, was the innovation process.

Q: What’s the biggest misconception about innovation? 

A: People think that all research investment works the same way if the goal is economic impact. But there are actually three different kinds of research investment, and they’re meant for different things.

There’s the one we all know about, which I call “altruistic science.” The purpose of altruistic science — in investing in an academic institution — is to produce educated people. It’s not done in the context of the world that ideas eventually have to succeed in. And if you honestly look at the direct economic yield over all these years, it’s basically zero.

Strategic research is the second investment category. As opposed to a single area of technology or science, it’s organized around a goal. A new F-35, for example, may need advancements in several fields, so the customer — in this case the government — wants them to come together. Basically, they’re taking economics out of the equation because they’re the only customer, but they have much broader uncertainty because they have multiple domains of technology that they have to deal with.

The third category is what I call “fundamental innovation.” It’s meant to represent the whole process from research to economic growth, even if it’s on 10-, 15-, or 20-year time horizons. Fundamental innovation is different because it has three variables: technology — what is physically possible; implementation — how it can be built and delivered; and market — who will adopt it, and why. Fundamental innovation involves all the necessary elements the whole time to converge on possible value. So you’re thinking about market applications the whole time, you’re thinking about new science and technology that could create new innovation options. Then you’re working in the real world saying, “OK, here’s how implementation would happen today, but maybe this could change, maybe that could change.” Not only are you doing your research, but the world is changing at the same time.

So that’s really the biggest misconception — that innovation is about an idea. It isn’t. It’s a process of working with things in the world until they become valuable.

Q: Who did you write the book for? 

A: I wrote it for multiple audiences: individual innovators and students; researchers and faculty; corporate leaders; research funders; and policy-makers. So, people who have a stake in trying to figure out, either with their careers or with their investments — whether it’s government or private — how to invest in the far future.

Q: What’s one lesson you hope readers take away?

A: For the policy people, I would say: Understand how innovation works in the economy, stop getting in its way, come up with new methods to drive it more efficiently, and realize there are three different streams of investment — altruistic, strategic, and this fundamental innovation stream that is not purposely being funded.

For students, I think understanding this is how you can actually have impact. What I point out in the book is that being involved in the innovation process makes you T-shaped: You have technical depth in one area and a broad working knowledge of many areas. If you’re doing research under these conditions, you start to learn about the world and all these different dimensions. It inherently includes business, economics, and applications. You’ve become broader, but then you still have the technical depth to drill down into any area.

For universities, this is who we should be. We should be teaching people how to do this and how to participate in these research corporations that I’m talking about. I call them third places: places that bring everybody together for this purpose — for investment, for everything else. Universities are the ones that can really trigger that, because companies aren’t going to have enough time. The government and universities should be targeting these third places for innovation, and students and faculty will be able to become more T-shaped through that interaction.

Innovation is a concept that has become mythologized in the modern era: what it is, how to manage it, how to teach it, and how to get it to work for us. Despite these explorations, it remains fundamentally misunderstood, writes Eugene Fitzgerald, the Merton C. Flemings SMA Professor in MIT’s Department of Materials Science and Engineering, in his latest book, “The Invisible Engine: Why Innovation Evades Control.”

Fitzgerald draws on a decade of work leading international research programs, including the MIT and Masdar Institute Cooperative Program and the MIT-Singapore Alliance for Research and Technology, where he explored innovation as the integration of market applications, technology, and implementation.

Written at a moment when artificial intelligence is reshaping how we think about knowledge, research, and innovation, “The Invisible Engine” examines a deeper question: How does innovation actually happen, and how should society invest in it?

In this interview, Fitzgerald discusses his own experiences with innovation — including his co-invention of strained silicon at AT&T Bell Laboratories in the 1990s, which helped extend Moore’s Law, the semiconductor industry’s long-standing trend of increasing the number of transistors on chips roughly every two years — while exploring common misconceptions about innovation, how to create the conditions for it, and novel ways to prepare institutions for future uncertainty.

Q: What inspired you to write this book? 

A: The book really grew out of the last 10 years of work in research-to-market activity, from the MIT Masdar program to the MIT-Singapore Alliance. In science, we have professional journals and things like that that capture discoveries within individual fields, but these larger-scale projects — where science, economics, industry, and society all intersect — don’t really have an academic thread that connects them.

I wanted to write a book that condensed all of those connections, because the innovation process at that scale is really the intersection of many different fields. The dominant ones are science and economics, because those are the underlying principles that drive how innovation happens.

So I was interested in marking this moment in history and documenting the experiments we’ve done at scale — trying to understand how knowledge of the innovation process can be incorporated into large collaborative research programs.

What started as a practical effort to make these programs work became a broader and somewhat unexpected interest in the innovation process itself.

Q: What is the “invisible engine?” 

A: The invisible engine is this decentralized collective intelligence of different actors, which are people and companies that eventually create surprise in the marketplace, which brings great profit.

This concept of “surprise” comes from Frank Knight, an economist from the early 1900s who was trying to understand the Industrial Revolution happening around him. So he takes a close look at the entrepreneur and asks, “What does the entrepreneur do?” And his answer is that the entrepreneur takes on uncertainty. They bring something into the world without knowing exactly what will happen, and their reward is surprise — everyone is surprised that people want it and that it can be done. Because the entrepreneur is the first to discover that opportunity, they can earn a profit.

Q: How did your experience developing semiconductor technologies shape the ideas in the book? 

A: It started with Bell Labs. My colleague and I made an important discovery — we found a way of straining silicon in a thin-film form with very few defects, which had never been done before. From the physics point of view, it was a big result. But I was always interested in having impact in the world, not just scientific recognition, so I went to my manager and asked, “What do we do next?”

He said, “Go talk to the marketing people at AT&T.” In hindsight, that made perfect sense. Bell Labs, like a lot of great industrial labs, created a lot of stuff, but they couldn’t always commercialize it.

Then I came to MIT, which was an open aperture after Bell Labs. Here I could keep uncertainty open across all the elements and find convergence in different directions. Eventually I started a company, and going between institutions to stimulate things was an eye-opening experience. We eventually reached a settlement with Intel over a patent dispute because the industry discovered that strained silicon was needed to extend Moore’s Law — something we never expected.

A lot of people want things to be organized and say, “Oh yeah, look at all that chaos.” But no — the path from Bell Labs to MIT to a startup, and then to industry adoption, was the innovation process.

Q: What’s the biggest misconception about innovation? 

A: People think that all research investment works the same way if the goal is economic impact. But there are actually three different kinds of research investment, and they’re meant for different things.

There’s the one we all know about, which I call “altruistic science.” The purpose of altruistic science — in investing in an academic institution — is to produce educated people. It’s not done in the context of the world that ideas eventually have to succeed in. And if you honestly look at the direct economic yield over all these years, it’s basically zero.

Strategic research is the second investment category. As opposed to a single area of technology or science, it’s organized around a goal. A new F-35, for example, may need advancements in several fields, so the customer — in this case the government — wants them to come together. Basically, they’re taking economics out of the equation because they’re the only customer, but they have much broader uncertainty because they have multiple domains of technology that they have to deal with.

The third category is what I call “fundamental innovation.” It’s meant to represent the whole process from research to economic growth, even if it’s on 10-, 15-, or 20-year time horizons. Fundamental innovation is different because it has three variables: technology — what is physically possible; implementation — how it can be built and delivered; and market — who will adopt it, and why. Fundamental innovation involves all the necessary elements the whole time to converge on possible value. So you’re thinking about market applications the whole time, you’re thinking about new science and technology that could create new innovation options. Then you’re working in the real world saying, “OK, here’s how implementation would happen today, but maybe this could change, maybe that could change.” Not only are you doing your research, but the world is changing at the same time.

So that’s really the biggest misconception — that innovation is about an idea. It isn’t. It’s a process of working with things in the world until they become valuable.

Q: Who did you write the book for? 

A: I wrote it for multiple audiences: individual innovators and students; researchers and faculty; corporate leaders; research funders; and policy-makers. So, people who have a stake in trying to figure out, either with their careers or with their investments — whether it’s government or private — how to invest in the far future.

Q: What’s one lesson you hope readers take away?

A: For the policy people, I would say: Understand how innovation works in the economy, stop getting in its way, come up with new methods to drive it more efficiently, and realize there are three different streams of investment — altruistic, strategic, and this fundamental innovation stream that is not purposely being funded.

For students, I think understanding this is how you can actually have impact. What I point out in the book is that being involved in the innovation process makes you T-shaped: You have technical depth in one area and a broad working knowledge of many areas. If you’re doing research under these conditions, you start to learn about the world and all these different dimensions. It inherently includes business, economics, and applications. You’ve become broader, but then you still have the technical depth to drill down into any area.

For universities, this is who we should be. We should be teaching people how to do this and how to participate in these research corporations that I’m talking about. I call them third places: places that bring everybody together for this purpose — for investment, for everything else. Universities are the ones that can really trigger that, because companies aren’t going to have enough time. The government and universities should be targeting these third places for innovation, and students and faculty will be able to become more T-shaped through that interaction.

The scales of a pine cone open in low humidity to scatter seeds, but close in damp conditions to protect seeds from moisture. An artificial material with the same behavior could be useful in applications like moisture-responsive shingles for passive cooling.

MIT researchers have now developed a system that simplifies the process of designing this type of bioinspired material. 

Their framework captures how mechanisms across length scales in a natural system, like the cells, fibers, and tissues inside a pine cone, work together to achieve unique properties. It then formally translates that behavior in an engineered system. 

The framework organizes biological behavior into building blocks that can be used to design synthetic structures that can be mathematically validated to perform the same way, and fabricated using a 3D printer. 

By taking much of the guesswork out of this design process, the framework could help engineers more readily create new adaptive materials while cutting development time and eliminating costs from failed prototypes. This framework could one day be used to design soft robotic grippers that respond automatically to their environment without any complex electronics, or morphing structures for airplane wings that predictably change their shape in response to temperature shifts.

“I’ve always been fascinated with natural materials and how complex behavior emerges from very simple building blocks,” says Lee Marom, an MIT graduate student in the departments of Mechanical Engineering and Architecture and lead author of a paper on this framework. “What really excites me about this work is going beyond bio-inspiration to what we could call ‘bio-derivation,’ where we move past observing a unique behavior to capturing the relationships and mechanisms that are actually producing that behavior, and then finding a systematic way to translate them into an engineered system.”

Marom is joined on the paper by corresponding author Markus Buehler, the Jerry McAfee Professor of Engineering in the departments of Civil and Environmental Engineering and Mechanical Engineering; Gioele Zardini, the Rudge and Nancy Allen Assistant Professor of Civil and Environmental Engineering, a principal investigator in the Laboratory for Information and Decision Systems, and an affiliate faculty with the Institute for Data, Systems, and Society; and Skylar Tibbits, an associate professor in the Department of Architecture. The research appears in the Journal of the Mechanics and Physics of Solids.

Biological building blocks

Pine cones can open and close their scales in response to humidity because of complex interactions within the organism’s structure. 

Shifts in humidity cause changes in microscopic cellulose fibers, which then cause transformations in larger groupings of fibers called laminas, which impact tissue layers, and so on, all the way up to the pinecone we see hanging from a tree branch.

“We instantiated the framework on the pine cone because it gives us a relatively simple, well-understood mechanism to demonstrate how the framework works. But its value becomes even greater as we apply it to more complex systems,” Marom says.

For engineers, the challenge is not necessarily reproducing an individual behavior, but translating the mechanisms and relationships that produce it across length scales. Without an explicit framework, these relationships need to be reformulated for each new system. 

To streamline the material design process, MIT researchers created a mathematical framework that captures how the components at each scale in a natural object work together to exhibit a certain behavior. The framework carries the design all the way to fabrication, translating the engineered behavior into verified manufacturing specifications and executable code that is used to 3D-print the object.

“What we were missing was a way to connect the mathematical description of a natural system all the way to its physical realization. The goal of this framework is to make that entire chain explicit so we can reason about what has to be preserved at each step,” Marom says.

The framework utilizes tools from category theory, which is a systematic method to compose larger systems from smaller ones in a way that is guaranteed to succeed.

Using category theory, the system maps out how a stimulus, such as humidity, causes a response at each level of the biological hierarchy within an organism like a pine cone. It models each level of the biological hierarchy as a separate building block that is independently validated.

Then the framework constructs a larger system from these building blocks by employing mathematical rules to ensure there is a valid transition between each step in the hierarchy. 

It assigns each building block in the natural system to a synthetic counterpart. In this way, the engineered material preserves the stimulus-response interactions that cause the natural organism’s unique behavior.

The work extends a research program in Buehler’s laboratory spanning more than a decade. 

Earlier studies used category theory to describe hierarchical materials and determine when building blocks could be replaced while preserving higher-level function. In subsequent work, Buehler and colleagues introduced “categorical prototyping,” using the same mathematics to preserve selected molecular-scale mechanics when translating computational models into large-scale 3D-printed prototypes. 

The new framework takes the next step by closing the entire chain, from multiscale biological mechanics, through an engineered realization and fabrication specification, to an experimentally validated, machine-executable design.

“Biological materials derive their extraordinary functionality from relationships that span scales, from molecular and fiber-level mechanisms to whole structures. Category theory gives us a way to make those relationships explicit and transferable. Once that design logic is captured mathematically, nature becomes a library of composable mechanisms that can be translated, recombined, and realized in new material systems,” Buehler says.

Compositional structure 

“Once we know that the relationships we mapped are valid, we can start recombining them in new ways. That means the framework isn’t only describing existing systems, it can also help us reason about ones we haven’t built before,” Marom explains.

For instance, the engineers mapped the humidity-driven bending behavior in a pine cone and the humidity-driven twisting behavior of a wheat awn as separate sets of building blocks. 

Then they combined some building blocks from each to design and fabricate a new type of actuator that exhibits thermal twisting behavior, without the need to do any new design work. When tested, the twisting actuator performed as the researchers expected.

In the future, engineers could use this framework to reliably combine verified components into new, bio-inspired designs for adaptive materials in applications like robotics, biomedical devices, or wearable technology.

“The systematization of our framework allows you to reuse pieces without needing to start from scratch each time, saving a huge amount of computation. That’s the real-world payoff,” Zardini says.

Now that the researchers have laid the groundwork with this mathematical framework, they can apply it to objects with more complex mechanics. They also plan to incorporate artificial intelligence models into their pipeline to expedite the discovery of new adaptive materials. 

“We have shown that the boundaries between disciplines do not matter as much as we think they do. Some of the principles from category theory can be used to guide and empower materials design. These mathematical structures seem to really have no boundaries,” Zardini says.

“The larger vision is physical AI: intelligence that can reason in terms of physical mechanisms and then turn those ideas into matter. Here we are beginning to build the infrastructure for that — composable physical knowledge, mathematical rules for determining what can be combined, and a path from a new design concept all the way to machine instructions and fabrication. Ultimately, this could allow AI not only to discover new materials and mechanisms, but to physically realize and test what it discovers,” Buehler says.  

This research was supported, in part, by the MIT Lemelson Engineering Fellowship, Singapore DSO National Laboratories, and the MIT Generative AI Impact Consortium.

MIT researchers have engineered bacteria that can function as transistors, allowing the team to create living “circuit boards” that can be printed onto a growth medium in a Petri dish. 

In electrical circuits, transistors function as switches that can turn current on or off. In the biological circuits that the researchers have created, bacterial switches control the flow of small molecules, which send signals to downstream circuit components. 

The research team designed two different transistors, along with three bacterial strains that relay information between the transistors, giving them the building blocks they need to design nearly any type of circuit. In a new study, they used these cells to create circuits that can add two or three inputs, or send one input to a specific location in the circuit.

“We’ve built some initial computer architecture components that are commonly used, but any operation can be built with these five strains,” says Hamid Doosthosseini PhD ’25, an MIT postdoc and the lead author of the new study. 

Using this approach, the researchers hope to develop circuits that one day could coat plant leaves or roots, where they could compute to sense and respond to environmental conditions such as drought or attack by pests. 

Christopher Voigt, head of MIT’s Department of Biological Engineering, is the senior author of the paper, which was recently published in Nature Chemical Biology. Former MIT postdoc Haorong Chen is also an author of the paper.

Cells as transistors

When designing synthetic biology circuits, researchers typically engineer cells to express proteins and transcription factors that interact to perform a task such as sensing a target molecule, which then triggers production of a specific output.

These simple circuits can perform various logic functions, but they must use unique transcription factors to avoid crosstalk within the circuit. There is a limited number of transcription factors that can be used for these circuits, which limits the overall complexity that can be achieved in a single cell. Additionally, putting too many circuits in one cell can overburden the cell’s protein production machinery.

In the new paper, the researchers took a different approach: Instead of building an entire circuit into one cell, they designed cells that could act as transistors. These transistors can then be combined in different ways to create a variety of circuits.

To create the transistors, the researchers chose a bacterium called Pantoea agglomerans, which commonly grows on surfaces, including plants. Using these cells, they made two types of transistors that can be switched on or off by a molecule called OC-6. One of the transistors is switched on by this input, and the other is switched off. Each transistor also detects the presence of a target molecule, in this case, OC-12. Depending on whether that molecule is present, and whether the switch is active, the transistors produce an output molecule known as OHC-14.

The researchers also used three strains of Pantoea agglomerans to create relays, which translate the OHC-14 signal into an output that can be fed into another transistor. Using these relay strains, the researchers can “wire” the transistors together, just like an electronic circuit board.

For example, they could create a bidirectional switch with two transistors that sense OC-12, and then send that information to different relay strains based on a switch input, ultimately feeding into other transistors that further process the signal.

The researchers created their circuits by printing colonies of bacteria onto plates containing agar, a growth medium. Each colony is printed about 5 millimeters from the nearest one. This allows the signals to travel only to the nearest colony, which then relays them to the next one, so information flows only in one direction.

Complex calculations

In this paper, the researchers demonstrated a transistor that can perform several types of logic operations depending on its location in the circuit layout, including “multi-input,” “or,” and “imply” gates. They also combined the transistors to create more complex circuits that can add up two signals, process more signals simultaneously, or function as a demultiplexer — a circuit that takes one incoming signal and sends it to one of several possible destinations, depending on a control signal.

The largest of these circuits, which adds two inputs together, contains 24 bacterial colonies wired together.

“This work shows that we can get toward more complicated functions by linking up simpler functions in individual cells,” Voigt says. “Computationally, there’s nothing that your iPhone can do that these circuits couldn’t do.” 

Circuits made from these cells take about eight hours to perform each calculation, much longer than a computer circuit. But, for biological applications, that is a reasonable amount of time, the researchers say. 

“We’re not trying to replace computers, but rather put computational control into biology. If you have bacteria on the root of a plant, or the plant itself is doing the computing, running a simple calculation overnight is fast enough relative to a growth season,” Voigt says.

If developed for use in agriculture, this type of circuit could be applied to the roots of plants to detect different types of stress. Once a particular input is detected, it would trigger a response such as synthesizing a fungicide.

The research was funded, in part, by the U.S. Defense Advanced Research Projects Agency and by the U.S. Intelligence Advanced Research Projects Activity.

As we move through everyday life, our brains engage in a huge variety of cognitive tasks. For example, during a grocery run, we might have to recall the items for a recipe, remember where the clerk said the flour was located, and count out money to pay.  

Scientists have long theorized that the brain contains modules, or clusters of neurons, that perform the same computation across many different types of tasks. This type of modularity could help explain why our brains are able to take on so many functions, with little difficulty.

In a new study of mice, MIT neuroscientists have found the first evidence for the existence of these flexible modules. They identified neurons in the prefrontal cortex that can be used to store either a sensory input or an action plan in working memory.

“We found that the brain doesn’t dedicate a separate group of neurons for every type of information. Instead, it uses the same populations of neurons to perform the same computation on different kinds of information, which means the same subset of neurons can hold both an action and a sensory stimulus in working memory,” says Yuma Osako, an MIT postdoc and the lead author of the new study.

The discovery supports the theory that reusable circuits allow the brain to mix and match components to generate a rich variety of behavior, the researchers say. 

Mriganka Sur, the Newton Professor of Neuroscience at MIT’s Picower Institute for Learning and Memory, and Timothy Buschman PhD ’08, a professor at the Princeton Neuroscience Institute, are the senior authors of the paper, which appears today in Nature Neuroscience. MIT graduate student Greggory Heller and postdoc Sofie Ahrlund-Richter are also authors of the study.

Cognitive building blocks

Dating back to his time as a graduate student at MIT, Buschman has been interested in understanding how the brain is able to perform so many different kinds of behavior. 

“One of the solutions that’s always been proposed has been this idea of compositionality — that you can take pieces of cognition that perform part of a task and reuse them in another task,” he says. 

In a study published last year, Buschman’s lab at Princeton showed that when animals perform a task such as categorizing objects based on their shape or color, they assemble neural circuits that perform different pieces of the task. Just like “cognitive Legos,” these building blocks can be flexibly combined to generate new behaviors.

Osako, who joined Sur’s lab several years ago, was also interested in studying cognitive flexibility. He and Sur teamed up with Buschman to explore a related question: whether individual neural circuits can be repurposed to perform different functions. 

“Our everyday life requires us to temporarily hold many different kinds of information. One big question is how the brain can represent an unlimited variability of information using only a finite number of neurons,” Osako says.

To get at that question, the researchers trained mice on a task in which they have to determine whether two sensory stimuli (high or low pitched tones) are the same, and respond accordingly. 

The researchers recorded electrical impulses from the brain while the mice performed this task, focusing on the prefrontal cortex, which is involved in executive functions such as planning and decision-making, and the parietal cortex, which processes sensory information and plans movement.

After measuring electrical activity from thousands of neurons, the researchers performed computational analyses that allowed them to identify groups of neurons that encode specific pieces of information.

They focused on two time periods — the time between the first and second tone, when the animals are holding a memory of the first tone, and the time between the second tone and the point where they have to decide on an action. During that second period, the animals are holding their decision and action plan in their working memory.

Within the parietal cortex, the researchers found that neurons appeared to exclusively store memory of the tone. But in the prefrontal cortex, they identified a cluster of neurons that could switch between the two types of memory. During the first period, they stored a memory of the first tone, but during the second, they were responsible for remembering the plan of action.

Re-using these clusters for different purposes allows the animals to flexibly store different types of information, the researchers say.

“When mice do tasks that test whether memory computations can be reused, the answer is they are. There are subspaces of functional activity in the prefrontal cortex that can be the substrate of mixing and matching toward flexible cognition,” Sur says.

Computational flexibility

The new findings offer support for the idea that the same computational circuits can be used for different purposes, Buschman says.

“The main result from this study is that there’s a circuit in the brain that maintains items in working memory, and you can put either sensory or motor information into it, and flexibly reuse it depending on what your current task is,” he says. “This means you do not have to build an entire new circuit for holding information in mind every time you want to learn a new task.”

The researchers now plan to study whether inhibiting these modules during different parts of the task affects the animals’ behavior, which could offer additional evidence that the flexible modules they identified participate in a variety of functions.

The research was funded by the National Institutes of Health, a MURI Grant, the Picower Institute Innovation Fund, the Japan Society for the Promotion of Science Overseas Research Fellowships, and the Uehara Memorial Foundation Postdoctoral Fellowship. 

Chiang Chung "C.C." Mei, professor emeritus in the MIT Department of Civil and Environmental Engineering (CEE), a renowned hydrodynamicist whose work shaped the field’s understanding of ocean waves and their interactions with coastal and offshore structures, passed away peacefully at home in Waltham, Massachusetts, on July 16. He was 91. 

For more than four decades, Mei was a defining presence in CEE. Since joining the MIT faculty as an associate professor in 1965, he became one of the world's foremost authorities on theoretical hydrodynamics, fluid mechanics, and ocean and coastal wave phenomena, retiring in 2010 after 45 years on the faculty. Throughout his career, he earned a reputation among colleagues and students as a generous mentor and thoughtful leader.

An elegant, rigorous scholar

Mei's research advanced the science of ocean wave hydrodynamics, spanning nearly every aspect, including nearshore currents, sediment transport and resuspension, the formation of sand ripples and bars on beaches, wave-induced stresses and seabed deformation, and the removal of contaminants from soils. In later years, he extended his mathematical approach to biofluid dynamics, publishing on flow problems in blood vessels and the inner ear, including a paper on "Streaming and diffusion in the cochlea" that appeared in the Journal of Fluid Mechanics in July 2025.

He authored over 300 publications, was cited more than 14,000 times, and coauthored the landmark books "Theory and Applications of Ocean Surface Waves" and "Homogenization Methods for Multiscale Mechanics." Mei published decades of influential research on wave power extraction, harbor oscillations, waves over muddy seabeds, coastal vegetation, landslide-generated waves, tsunamis, and hydrodynamic resonance. His work combined mathematical elegance with practical engineering application and continues to guide solutions to some of the world's most complex ocean-based environmental challenges, including coastal defenses against storms and oil spill response strategies.

"C.C. tackled deep, diverse, difficult, and important fluid dynamics questions with utmost finesse and elegance," says Lydia Bourouiba, the Japan Steel Industry Professor. "He was an inspiring scholar, an intellectual leader, and a wonderful mentor, whose rigor set the standard we should continue to uphold. We lost a true giant in our field."

His excellence and leadership in research and teaching earned numerous prestigious recognitions, including a Guggenheim Fellowship in 1972, election to the National Academy of Engineering in 1986, fellowship in the American Physical Society, the Theodore von Kármán Medal in 2007, and appointment as a Ford Professor of Engineering at MIT. 

Beyond his scientific achievements, Mei devoted himself to the MIT community, serving as interim head of CEE from 2001 to 2002 and helping guide the department through a period of transition with the same humility and steadiness that characterized his scholarly contributions. In 2015, the department established the C.C. Mei Distinguished Speaker Series in his honor — an idea that grew out of the initiative of Bourouiba to revive CEE's environmental seminar series and honor its strong historical legacy in fluid dynamics. "Discussing the idea with C.C., he thought it was an excellent idea and was so generously supportive. He embodied the excellence I wanted the new series to reflect; naturally, we named it in his honor," she says. The series continues to bring internationally renowned scholars to MIT. 

A mentor whose students became family

Mei's influence was equally profound in the lives of his students. Over more than 45 years at MIT, Mei advised and mentored generations of engineers, many of whom became leaders in academia, industry, and government. Even in his final days, those relationships endured. 

One of his first doctoral students, Professor Emertius Ole Madsen, visited him just hours before his passing. Former student Yile Li SM '01, PhD '06, who continued collaborating with Mei on biofluid dynamics research in his later years, remained in close conversation with him throughout his final days. Li recalls the highlight of his discussions with Mei. "He told me there are three stages of doing research: solving problems using mathematical methods, modeling problems by capturing core physics, and ultimately discovering entirely new problems," Li says. "His own work proved he was a master of all three."

For Mei's family, MIT was never simply his workplace. "CEE was truly the center of my father's life," says his daughter Deborah Mei. "For more than 50 years, it shaped not just his career, but our whole family's world. His students, colleagues, and collaborators weren't separate from our home life — they were part of it, for as long as I can remember."

Colleagues consistently remember not only Mei's intellectual brilliance, but also his extraordinary generosity. Mei was known for the warmth he extended to junior colleagues finding their footing at MIT. 

"He was so respectful and kind to me when I was hired in 1976, and feeling like a fish out of water," says Institute Professor Sallie “Penny” Chisholm. "I will never forget that. A great gentleman, indeed." 

Heidi Nepf, the Donald and Martha Harleman Professor, recalls Mei as "an exceptional scholar and a wonderful colleague."

Rafael L. Bras, professor emeritus, remembers Mei as the model of the gentleman scholar. "He cared deeply about people, loved his profession, and touched countless lives, both directly and indirectly. Everybody loved him."

A full and joyful life

Mei was born on April 4, 1935, in Wuchang, Hubei Province, China, the only son and first child of Ju-Long Mei and Wu Yu-Ling. He earned his BS from National Taiwan University in 1955, his MS from Stanford University in 1958, and his PhD from Caltech in 1963.

Those who knew him describe a man who was passionate, playful, endlessly curious, and quick with both words and affection. The home he shared with his wife, Caroline, became a gathering place for generations of the Mei family, his MIT colleagues and students alike with animated conversation, humor, and his familiar loving banter with his wife and siblings, as those close to him remember it.

To generations of students and colleagues, Mei was known as much for his patience, kindness, intellectual curiosity, and quiet encouragement as for his scientific accomplishments. He was always willing to discuss an idea, help a student work through a difficult problem, or offer thoughtful guidance to a young colleague beginning an academic career. His legacy lives on not only in the theories that continue to shape coastal and ocean engineering, but also in the worldwide community of scholars he mentored, inspired, and welcomed over more than half a century at MIT.

Mei is survived by his wife, Caroline (Schmitt) Mei of Waltham; his daughter, Deborah Yupin Mei, and her husband, Juan Ignacio Garcia De Motiloa Ubis of Singapore; his sisters Helen Chiang-Hua Mei Chao of Potomac, Maryland; Teresa Chiang-Ming Mei Wu of Bethesda, Maryland; Heidi Chiang-Kuo Mei Hsia and her husband, Jack, of Potomac, Maryland; and Christine Chiang Ying Mei and her husband, Paul Tung, of Rancho Palos Verdes, California; his grandchildren, Juan Ignacio Jr. and Lauren; and many nieces, nephews, grand-nieces, and grand-nephews.

Gifts may be made in Mei's memory to the Chiang and Caroline Mei Fund 

Every year, lung cancer kills more than 100,000 people in the United States. Smoking is the leading risk factor for lung cancer, but other environmental exposures can also contribute to the disease.

In an advance that could help prevent some of those lung cancer deaths, MIT researchers have shown that blocking an enzyme involved in lung inflammation appears to reduce the risk of developing tumors. 

The researchers found that this enzyme, caspase-1, is active in developing tumors in mice. When they treated the mice with a small-molecule drug that inhibits caspase-1, the mice were much less likely to develop lung tumors.

That drug has already gone into clinical trials for other diseases, and the researchers now hope to test it as a preventative drug in people with elevated risk for lung cancer. 

“If you look at global cancer deaths, lung cancer causes most of them, and most of that is driven by tobacco smoking. Additionally, people who are ‘never smokers’ are showing up with lung cancer. You can imagine a future where you get a test and if you’re deemed high-risk, you go on a preventative medicine. This concept is called cancer interception, and it could help millions of people,” says Sangeeta Bhatia, the John and Dorothy Wilson Professor of Health Sciences and Technology and of Electrical Engineering and Computer Science at MIT, and a member of MIT’s Koch Institute for Integrative Cancer Research and the Institute for Medical Engineering and Science (IMES).

Bhatia is the senior author of the new study, which appears today in Science Advances. Cathy Wang PhD ’26 is the lead author of the paper. 

Blocking inflammation

Preventing lung cancer in patients who are at high risk could significantly reduce the death toll of the disease. In 2017, a clinical trial run by Novartis yielded a tantalizing hint that targeting lung inflammation could prevent some lung cancer cases. That trial, known as CANTOS, was designed to examine whether an anti-inflammatory drug — an antibody that blocks the cytokine IL-1 beta — could reduce the risk of strokes and heart attacks. Unexpectedly, the researchers found that this treatment led to lower rates of lung cancer in a subset of people.

Later trials showed that the antibody had little effect in patients who had established lung cancer, but researchers are still exploring the possibility of using it to prevent progression of lung cancer in high-risk patients. A recent study by the Swanton lab at the Francis Crick Institute identified a set of proteins, across several biological pathways and cell types, that could be used to predict which patients would respond to treatment with an IL-1 beta antibody. 

IL-1 beta requires protease cleavage to be converted to its mature, active form. Thus, Bhatia and her team wondered if enzymes called proteases, which cleave other proteins, might be involved in driving the inflammatory pathway that includes IL-1 beta.

For several years, Bhatia’s lab has been developing tools to track and visualize proteases, since the activity of these enzymes can contribute to cancer development. Proteases can help tumor cells escape their original locations by cutting through proteins of the extracellular matrix, and they also play essential roles in guiding inflammatory cell migration, which can influence tumor growth and immune system targeting.

By coming up with ways to detect these enzymes, Bhatia’s lab has created diagnostic nanosensors for cancer and other diseases. The sensors consist of nanoparticles decorated with peptides that can be cleaved by certain proteases, revealing when proteases are active in a particular tissue or disease state.

In addition to their role in cancer, proteases are known to be involved in the regulation of inflammation. In their new study, Bhatia and her colleagues adapted their nanosensors to identify proteases that may participate in IL-1 beta-mediated inflammatory pathways. 

“We know that proteases are very important in inflammation, and we wanted to pinpoint which ones might be the most active during early lung cancer development,” Wang says.

For this study, the researchers used a mouse model developed by Tyler Jacks, the David H. Koch Professor of Biology at MIT and a member of the Koch Institute. This model, known as KPS, is engineered to turn on cancer-causing mutations in the p53 and Kras genes. The mice also express a peptide called SIINFEKL, which helps to activate T cells and stimulate inflammation in the lung.

The researchers designed their experiments to allow them to model increased cancer risk, beginning before tumor formation was detectable. Five weeks after they induced the cancer-causing mutations, the researchers injected some of the mice with an antibody that blocks IL-1 beta, while others were untreated. Three weeks later, the researchers used their nanosensors to detect proteases that were active in the lungs. 

Those experiments showed that in untreated mice, which all developed lung tumors, caspase-1 was very active. However, in the treated mice, which had fewer tumors, caspase-1 activity was significantly reduced. The researchers also found that in untreated mice, the active caspase-1 was found primarily in lung tumors, not in nearby healthy tissue.

Working with Lecia Sequist, a professor of medicine at Havard Medical School and physician at Mass General Brigham, the researchers also analyzed a small number of human lung fluid samples. In these samples, they also found higher levels of caspase-1 activity from patients with lung cancer compared to healthy donors, despite a common smoking history.

A repurposed drug

The observation that caspase-1 activity is interrelated with the IL-1 beta inflammation pathway was not completely surprising, given IL-1 beta itself required protease cleavage to be converted to its mature, active form. The MIT team then investigated whether inhibitors of caspase-1 might also provide the same protective effects as inhibitors of IL-1 beta, or even improve them. 

Before tumors developed, the researchers began treating the at-risk KPS mice with either a caspase-1 inhibitor, an IL-1 beta antibody, or both. In mice that received both drugs, nearly 20 percent never developed tumors at all. In the mice that received either the caspase-1 inhibitor or the IL-1 beta antibody alone, tumors were much smaller and less numerous than in untreated mice.

Unlike antibodies, which need to be given intravenously, caspase-1 inhibitors can be taken orally, which could make them more appealing as a preventative treatment. Another opportunity provided by these drugs is that they have previously been tested in clinical trials for treatment of rheumatoid arthritis and other diseases.

“What’s so attractive about using this caspase-1 inhibitor is that it has actually been tested in humans. It’s already been through safety studies, and we think it could potentially be repurposed for cancer prevention,” Bhatia says.

The researchers hope to test the drug in a clinical trial, potentially using the biomarkers that were identified by the Swanton team to identify subjects who are likely responsive to IL-1 beta antibody treatment. 

The authors of the study also include MIT researchers Qian Zhong, Shih-Ting Wang, Carmen Martin-Alonso, Sofia Neaher, Sahil Patel, Tiziana Parisi, Jesse Kirkpatrick, and Tyler Jacks. 

The study was funded by Johnson & Johnson, Upstage Lung Cancer through the Koch Institute Frontier Research Program, the Virginia and D.K. Ludwig Fund for Cancer Research, the Koch Institute’s Marble Center for Cancer Nanomedicine, the Koch Institute Support (core) Grant from the National Cancer Institute, and a core center grant from the National Institute of Environmental Health Sciences. 

Epithelial cells are the tiny shields that line and protect our body. In a developing embryo, epithelial cells grow, divide, and move into positions to form the outer layers of our skin and the surfaces of our organs and blood vessels. When we scrape our skin, suffer an internal tear, or undergo surgery, epithelial cells will migrate to the site of injury to heal a wound. And when epithelial cells go haywire, they can turn malignant and spread through the body as cancer. 

MIT engineers have now discovered that as they migrate, epithelial cells can synchronize and collectively pulse. In a study appearing today in the journal Newton, the researchers report observing groups of epithelial cells repeatedly moving in, then out, like a circle of dancers coming together and pulling apart. 

The team measured this collective rhythmic pulsing in different types of epithelial cells, including healthy cells, cells from benign tumors, and cancerous cells. 

Surprisingly, they discovered that malignant epithelial cells were more persistent in their synchronization, pulsing together for twice as long as healthier cells. It’s unclear why the cells sync up in this way. But the researchers suspect that this cellular dance can serve as a clinical signal.

“More aggressive cancer cells tend to have a steadier and more persistent rhythm as compared to healthy ones,” says study author Ming Guo, professor of mechanical engineering at MIT. “We think this coordination could serve as an early warning sign of how likely a tumor is to spread. The same coordinated waves may help shape embryos during development and close wounds upon injury.”

The study includes first author and former MIT graduate student Wenhui Tang SM ’20, PhD ’24; Mehrana Nejad and L. Mahadevan of Harvard University; and Adrian Pegoraro of the Metrology Research Centre of the National Research Council Canada. 

Cells got rhythm

When studying how epithelial cells organize and develop into whole organs and tissues, scientists have focused mainly on how the cells coordinate in space. Where cells move, where they are in relation to the growing tissue, and where they end up, are questions of spatial coordination that scientists including Guo have looked to investigate. How the movement of cells relate over time is less well-understood. 

Guo’s group at MIT studies cell interactions to identify patterns that relate to healthy versus diseased states. As part of this work, the team takes microscopic snapshots of cells that they grow in the lab, to identify interesting behaviors among cells. Recently, Tang, then a member of Guo’s lab, was looking at a series of movies of epithelial cells when she started to see a rhythm, or pattern over time.

“I was studying collective cell migration, and I observed cells were swelling, then squeezing together, then swelling, again and again, forming local patterns,” Tang recalls. “That’s when I realized there might be something more interesting happening with these cells over time.”

Taking a pulse

In their new study, the researchers focused on the timing of cellular movements. They started by studying healthy, live epithelial cells that they cultured in the lab. They stained the cells with fluorescent dye to illuminate each cell’s nucleus. This way, they could easily identify one cell from another. They kept the cells in dishes with nutrients to help them naturally grow, divide, and move about. 

“We’re looking at their natural migration process, related to how they would migrate during different processes in the body, such as when forming skin and organs, and healing wounds,” Guo explains.

Using a confocal microscope, the team took snapshots of the cells every few minutes, for up to 30 hours. When they strung the images together as a sort of movie, a distinct pattern emerged. 

“If you just stare at any one location, you can see those dots are coming together, and then going further away, then coming together again, and going further away, like waves,” Tang says. 

They observed that a single pulse occurred over about an hour. This pulsing persisted in healthy cells, as a slow and steady rhythm over the 30-hour period. 

Curious as to whether other types of epithelial cells would sync up in similar fashion, the team tried the same experiment with several different lines of human breast cancer epithelial cells. They studied the movement of cells from benign tumors and cells of increasing malignancy. They observed similar pockets of synchronized pulsing in every cell type, especially in the most cancerous cells. 

“We found the really dangerous cancer cells team up over time, and do this persistent oscillation, twice as long as healthy cells,” Guo says. “This is unexpected. We see they really team up, synchronize, and oscillate together, which potentially facilitates their invasion.”

The researchers also observed a correlation between cell synchronization, and cell density: In each dish of cells, regardless of type, the cells continued to grow, divide, and pulse. As their numbers grew, more cells pulsed together, and their synchronization increased, up to a point. Once the cells reached a certain density, their pulsing began to die down.

“There’s a peak of synchrony before it decreases as cell density continues to increase,” Tang says.

This connection is especially interesting in the context of certain conditions such as asthma. Epithelial cells line the inside of many organs and tissues, including the airways. In healthy people, these cells pack together and “jam” up to form a solid, stable lining that protects the airways. In asthmatic airways, however, epithelial cells are less able to jam together. This results in airways that are loose and fragile, easily irritated, and difficult to heal. 

Guo and Tang suspect that, as there appears to be a connection between cell density and cell synchronization, there may be a way to target asthma treatments, by watching how potential drugs affect asthma cell synchronization. A similar approach could be taken for the screening of cancer drugs.

“More malignant cells would be better synchronized. After treating them with a drug, if their synchronization is disrupted, then it might be an efficient drug where we can consider the next step,” Guo envisions.

This research was supported, in part, by the National Institutes of Health.

Within the brain, neurons compute by generating electrical impulses. These signals travel throughout neurons, which are in turn connected in vast networks that control brain functions such as sensory perception, memory formation, and control of movement. 

In an advance that could help neuroscientists map those neural networks, leading to a better understanding of how neural activity underlies behavior and other brain functions, MIT engineers have invented a new microscope that can image electrical activity in neurons distributed across the brain of an entire organism, the experimental model Danio rerio (zebrafish).

Using a microscope that they adapted for fast, high-volumetric rate imaging, the researchers were able to track electrical activity across the brain on the scale of milliseconds. This method revealed patterns of neural activity from neurons throughout the brain that were activated in response to ultraviolet light.

“All of the parts of the brain are connected together, so if you want to truly understand the brain, you have to understand how all the neurons work together as an emergent whole,” says Ed Boyden, the Y. Eva Tan Professor in Neurotechnology at MIT; a professor of biological engineering, media arts and sciences, and brain and cognitive sciences; and a member of MIT’s McGovern Institute for Brain Research, Yang Tan Collective, and the Koch Institute for Integrative Cancer Research.

Boyden is the senior author of the study, which appears today in Nature Methods. Former J. Douglas Tan Postdoctoral Fellow Zeguan Wang PhD ’24 and former MIT research scientist Jie Zhang are the lead authors of the paper. Other authors include former MIT postdoc Panagiotis Symvoulidis, Picower Institute research scientist Wei Guo, graduate students Davy Deng and Lige Zhang, Koch Institute research scientist Adam Amsterdam, Picower Institute research scientist Takato Honda, Boston College undergraduate Steven Roche, and Matthew Wilson, the Sherman Fairchild Professor of Neuroscience at MIT and a member of the Picower Institute.

High-speed imaging

One technique often used to measure neuron activity in the brain is calcium imaging. Calcium flows into neurons after they fire an electrical impulse, so measuring calcium levels in the cells can serve as a proxy for neural activity. However, this type of imaging isn’t fast enough to capture single spikes of activity.

“Calcium imaging inherently is very slow, so you’re talking about imaging activity on the order of seconds or even minutes. Typically that is too slow for us to be able to see a lot of these high-speed neural activities,” Zhang says. “Neurons compute using electrical activity, so with voltage imaging, you can get direct observation of that.” 

To enable direct imaging of voltage, researchers have developed proteins called genetically encoded voltage indicators — fluorescent proteins that can be genetically expressed in neurons. When a neuron fires an impulse, the protein fluoresces, which can be detected with a fluorescence microscope.

In previous work, researchers have used these proteins to image small populations of neurons, usually focusing on one localized part of the brain. Until now, there hasn’t been a way to image a large volume, such as the entire brain, with the millisecond-scale resolution needed to see electrical impulses from individual neurons.

To achieve that, the MIT team decided to modify a commonly used microscope known as a light sheet microscope. This type of microscope uses a sheet of laser light to illuminate a thin slice of a sample. By imaging many layers in sequence, this technique can generate 3D images of a large volume. However, with previous microscopes, the scanning of an entire volume would take too long to be able to capture neuronal impulses across the volume at single cell resolution. 

“Different groups of neurons that are distributed across the brain coordinate together at millisecond timescales to generate a lot of behaviors and brain computations,” Wang says. “To understand the principles, we need the technology to observe their activity at the same time, across the whole brain, so we are not missing any important participant neurons.”

To make the imaging process fast enough to image millisecond-scale activity, the researchers increased the image acquisition speed of the microscope’s camera, and they also boosted the scanning speed of the microscope using a technique called remote refocusing.

Using this approach, the researchers showed that they could scan the entire zebrafish brain 200 times per second, or once every five milliseconds. 

Mapping brain activity

To test the new microscope, the researchers engineered neurons in larval zebrafish to express a voltage indicator called Positron2-Kv. Although they had hoped that the indicator would end up in every neuron, it produced signals in neurons distributed throughout the brain, with about one quarter of the neurons exhibiting acceptable signals. This was enough, however, to observe patterns of activity across the brain. The researchers imaged the brain as the fish were resting, and they were able to observe single voltage spikes from neurons, as well as rapid bursts of spikes.

Additionally, this technique revealed patterns in how the brain is activated following a stimulus such as ultraviolet light. Immediately following the stimulus, activity was seen in the optic tectum, which receives and processes visual input from the retina. This activity propagated from one side of a part of the brain called the tectum to the other. Stimulus-independent activity also occurred in sequences across sets of neurons in the cerebellum and hindbrain.

The researchers now hope to increase the percentage of neurons that they can image across the brain, as well as the microscope’s speed and resolution. They are also working on expanding the use of this technique to other experimental models, including mice.

This approach, they say, could offer neuroscientists a new way to generate hypotheses about what happens in the brain when it engages in specific behaviors, or about how brain activity is linked to states of mind such as daydreaming.

“A big question is simply to understand how neurons work together as a network. And this might be the first time that you could do that, because you can image the voltage of neurons distributed throughout the network,” Boyden says.

The research was funded by the National Institutes of Health, the BRAIN Initiative, the Picower Institute Innovation Fund, K. Lisa Yang, Ashar Aziz, the K. Lisa Yang and Hock E. Tan Center for Molecular Therapeutics in Neuroscience at MIT, the Hock E. Tan and K. Lisa Yang Center for Autism Research, the Alana Down Syndrome Center, John Doerr, Jed McCaleb, James Fickel, and the Howard Hughes Medical Institute.

The moment a nuclear reactor begins operation, a complex chain of events is initiated within the fuel: Heavy atoms split into fission products, knocking other atoms out of place and creating defects that can change how the fuel swells, transfers heat, and reacts chemically over time. 

Understanding those processes is key to understanding how safe and efficient a nuclear reactor will be. But even for some of the most-studied fuel types, the mechanisms controlling those processes are unclear.

Such is the case with a particular kind of metallic fuel, uranium alloyed with 10 percent zirconium by weight, also known as U-10Zr. This fuel was extensively tested in historic sodium-cooled fast reactors such as the Experimental Breeder Reactor-II (EBR-II) in Idaho and the Fast Flux Testing Facility (FFTF) in Washington state, helping establish the foundation for metallic fuel development in the U.S. Today, U-10Zr is again attracting attention for use in next-generation advanced reactors.

But most studies of U-10Zr took place decades ago, leaving unanswered questions about exactly how the fuel changes when it undergoes nuclear fission in a reactor and how it interacts with the protective fuel cladding surrounding it.

Now, together with Idaho National Laboratory (INL), MIT researchers have led one of the most detailed three-dimensional studies of irradiated U-10Zr to date. The researchers used a technique known as high energy synchrotron X-ray computed tomography at Brookhaven National Laboratory (BNL) in New York to analyze the pore networks and chemical changes that formed under irradiation during use inside the FFTF reactor, providing new insights into how the material swells, transfers heat, and interacts with the fuel cladding. 

The findings could help keep some nuclear reactors running for longer, while also informing the next generation of nuclear reactor fuel systems.

“This study helps us model the pore distribution in the fuel more accurately,” says senior author Ericmoore Jossou, MIT’s John Clark Hardwick (1986) Professor of Nuclear Science and Engineering. “It also helps us design for the safe operation of metallic fuels in reactors by giving us a better understanding of the role of pores and their importance.”

Joining Jossou on the paper are first author and MIT postdoc Anthony Harrup; Riley Moeykens ’25, SM ’25; BNL researchers Michael Drakopoulos and Nghia Vo; and INL researchers Jana Howard, Colby Jensen, and Tiankai Yao.

Understanding nuclear fuel

A class of nuclear reactors known as sodium-cooled fast reactors generate energy from rods of metallic fuels that are sealed inside metal tubes called cladding. In each rod, heat generally moves outward from the center to the edge and then to the cladding, where liquid sodium carries heat away to be harvested into power.

“As you operate the reactor, the contact between the fuel and the cladding material creates chemical interactions that can be problematic,” explains Jossou. “There is a migration of materials from the fuel to the cladding, like fission gases and rare earth elements called lanthanides, which can react with the cladding, cause embrittlement, and damage the fuel system.”

Studies of previously irradiated fuel and its cladding have captured mostly two-dimensional snapshots, preventing scientists from seeing the full scale of the pore networks that influence heat transfer and transport materials like lanthanides. Previous studies also mainly focused on specific sections of the fuel system, such as the fuel center or the fuel cladding interface.

For their study, the MIT researchers used fuel samples from the Fast Flux Testing Facility reactor, a sodium-cooled fast neutron reactor located in Washington state that operated from 1982 to 1992.

The Idaho National Lab managed the samples and prepared the samples. The team studied the prepared samples using high-energy synchrotron X-ray tomography at the Brookhaven National Laboratory. The synchrotron generated high-energy X-rays that allowed the researchers to reconstruct the fuel’s internal pore networks in three dimensions, revealing how porosity, chemistry, and fuel-cladding interactions evolve across the fuel radius.

The researchers found porosity increased modestly from the center of the fuel toward the fuel edge, but pore density jumped by over two orders of magnitude at the fuel’s edge by the cladding. The researchers also characterized the size and shape of pores, finding small pores at the center that turn into larger pore networks pointing outward toward the edge.

“The pores are currently modeled as spheres; however, in reality they are more complex, especially when many pores merged together,” Harrup says. “That’s true from the center all the way to the cladding. It explains why the cladding reacts the way it does, and why we see cladding chemicals in the fuel.”

The pore networks toward the edge allow fission products and lanthanides to move but slow down heat transport, impacting the fuel’s performance and lifetime. The researchers also mapped their microstructural findings with changes in the chemistry of the fuel in different areas.

“With this study, we’ve conducted an in-depth analysis enabled by advanced computational imaging methods that has never been done before, with correlations between local chemical environments and the formation of pores,” Harrup says. “It turns out that whether the environment is uranium rich or zirconium rich impacts the morphology and the channels of the pores. That has never been reported before.”

“The ability to directly visualize pore connectivity and fuel cladding interaction in three dimensions gives us important insight for improving fuel performance for advanced metallic fuel for sodium fast reactors,” says Tiankai Yao of INL.

Informing reactor designs

The experimental findings differed from some models of how pores form and how the fuel system swells, which could improve simulations to help keep reactors running for longer. They also give a more nuanced picture of how pores influence reactor performance and safety.

“This helps optimize the current metallic fuel proposed for sodium fast reactors,” Jossou says. “Now, together with INL, we better understand how pores are influencing the thermal performance of metallic fuel in reactors. At high temperature, the pores are not all bad, because we found they act as pathways for liquid sodium metal to flow through the fuel and sustain thermal conductivity. Connected pores could also serve as releasing channels for fission gases which reduce the internal fuel matrix stress.”

The findings could also be used to design better fuel systems for next generation of sodium fast reactors.

“This excellent piece of work generated by Professor Jossou’s group in collaboration with INL and BNL has elegantly combined the strength of attenuation-based X-ray tomography and focused ion beam lift-outs and produced valuable insights to the location-specific 3D porosity distribution in neutron-irradiated U-10Zr fuel,” says Dong Liu, a professor at Oxford University who was not associated with this work. “What is also impressive is that they correlated 3D porosity to the thermal properties of the fuels: The total volume fraction is not the only parameter that is important, the 3D topology also matters. This is extremely informative for the study of other types of porous nuclear materials.”

The work was supported by the U.S. Department of Energy Office of Nuclear Energy and utilized resources at BNL and INL. The sample preparation was carried out at INL, which is part of the Nuclear Science User Facilities, through a Rapid Turnaround Award.

A new short film spotlights the life and career of MIT Institute Professor and School of Engineering Dean Paula Hammond ’84, PhD ’93.

The documentary, “Full Circle: Paula Hammond at MIT,” traces Hammond’s path from childhood in Detroit, Michigan, to her arrival at MIT at 16 years old, to her evolution into a pioneering researcher in nanotechnology and ovarian cancer, as well as a leader at the Institute and around the globe.

The film is one of the debut offerings within “MIT Stories,” a new documentary series on MIT Learn that spotlights the innovators and changemakers whose work extends far beyond campus walls. Produced through intimate storytelling by MIT Open Learning’s Emmy Award-winning video team, the series aims to explore the passions that spark global impact and the human stories behind innovation.

“Everything Paula Hammond does is grounded in a deeply personal sense of purpose,” says Lana Scott, assistant media development director at MIT Open Learning who produced the film with Nick Vandenberg. “As a pioneering researcher and the first woman to lead MIT’s School of Engineering, she didn’t just break barriers, she changed what leadership can look like in a field that hasn’t always made space for people like her. Her story blends curiosity, care, and conviction, turning complex science into something human, relatable, and genuinely cinematic.”

The film’s original score was composed by Vandenberg, who was inspired by a musician Hammond has long cherished.

“Before our second interview, Paula and I spoke about our shared love of jazz, including artists like Charlie Parker and Miles Davis,” says Vandenberg, a videographer and senior editor at MIT Open Learning. “She mentioned Ramsey Lewis as a particular favorite of hers. So, as a little Easter egg for her, I wrote and recorded a composition with upright bass, drums, and organ based loosely on the sound of his early trio recordings.”

Video by Lana Scott and Nick Vandenberg / MIT Open Learning | 8 minutes, 40 seconds

The National Science Foundation (NSF) has selected MIT to establish and lead a new Materials Research Science and Engineering Center (MRSEC) focused on materials technologies for medical imaging, sustainable metals production, and next-generation semiconductors, according to an NSF announcement released July 30.

Expected to provide $18 million in research funding over six years, the award brings together 16 research groups from nine departments across four institutions, including five MIT departments, three collaborating universities, and a teaching hospital. The award is pending MIT’s negotiation of a formal research agreement with the NSF.

The MIT Materials Research Science and Engineering Center will be directed by Associate Professor Rafael Jaramillo of the Department of Materials Science and Engineering (DMSE), with Professor Caroline Ross of DMSE serving as associate director. The center will be housed administratively within the MIT Materials Research Laboratory.

The center will have two main research thrusts. One will engineer specialized materials to advance X-ray detectors used in medical imaging, potentially leading to better cancer diagnosis, lower radiation exposure, and improved industrial and security imaging. The other will explore high-temperature sulfur-based molten materials to transform how metals and semiconductors are made, opening a path to more efficient metal production, improved access to critical materials, and new thin-film semiconductor technologies. 

The expected funding will also support a new shared laboratory for testing magnetic materials and materials under extreme conditions, managed by MIT.nano. This facility will be available to academic and industry users, expanding the nationwide portfolio of NSF-supported research facilities. 

“The long-term goal is for the broader materials and engineering community to see the disruptive potential of bringing researchers together across disciplines to solve complex challenges,” says Jaramillo, the Stavros V. Salapatas Career Development Professor of Materials Science and Engineering. “And that includes specifically in medical diagnostics and metals production, where entirely new things will be possible that aren’t considered possible today.”

A legacy of collaboration

The selection of MIT’s MRSEC is part of a $108 million NSF investment in six research centers that will explore a range of topics, including artificial intelligence-driven experimental laboratories and hybrid quantum materials that combine light and matter. NSF’s MRSEC program brings together interdisciplinary teams of researchers to push the boundaries of materials science and engineering and tackle complex scientific challenges.

The MIT center builds on nearly 60 years of interdisciplinary materials research at the Institute, extending a legacy that began with U.S. Department of Defense-supported laboratories in the 1960s and continued through NSF-funded centers in subsequent decades. Past MRSEC investments helped build research communities that enabled MIT centers of excellence such as the MIT Microphotonics Center and the Microsystems Technology Laboratories.

“We were inspired to continue that legacy of collaborative research in materials science,” Jaramillo says. “It’s mainly the mode of working — the mode of working in a very intentional way as a team across disciplinary boundaries and having this program that brings people together.”

MIT departments involved in the MRSEC include DMSE; Chemistry; Chemical Engineering; Earth, Atmospheric and Planetary Sciences (EAPS); and Physics. Collaborating institutions identified in the MRSEC proposal are Yale University, the University of California at Santa Barbara, and the Department of Radiology at Massachusetts General Hospital and Harvard Medical School. 

The first research group will focus on re-engineering scintillators — materials that convert X-rays into visible light — at the nanoscale, with the goal of improving resolution, speed, and energy sensitivity.

“My vision for that is really Marin and JJ’s vision. So I'm basically cheerleading for them,” Jaramillo says, referring to optical materials experts Professor Marin Soljačić of Physics and Professor Juejun Hu of DMSE, who are expected to lead the effort.

The second group is closer to Jaramillo’s own research in semiconductor and advanced electronic materials. It seeks to develop a deeper understanding of high-temperature sulfur-based liquids to improve the yield and efficiency of producing critical metals such as copper. Expertise in these types of materials has become increasingly rare in U.S. academia, Jaramillo says, and one goal of the center is to rebuild that capability at MIT. “I’m very excited about that being a new intellectual center of gravity.”

Telling stories about materials

Beyond research, the center is also expected to develop outreach activities highlighting the importance of materials science in society, particularly in the Boston region, where Jaramillo said industries need more workers with backgrounds in materials processing.

“For example, our community colleges don’t offer it,” Jaramillo says. “If you were looking at a community college in Michigan, everyone would know what materials science is.”

One initiative, DISASTER! — “with all caps and an exclamation mark,” Jaramillo says — will tell stories of real-world catastrophes and the materials failures that contributed to them.

A major part of materials science over the last century has been understanding why things fail, Jaramillo says. “It’s also a tremendous foot in the door for introducing the field. Because frankly, ‘if it bleeds, it leads.’ If you have giant disasters, then suddenly people are like, ‘Why did the bridge fail?’”

The program will encourage MIT undergraduates to research and tell these stories, illustrating how forensic materials science has helped prevent future failures.

Among the examples Jaramillo cited are the rivets used to assemble the RMS Titanic, whose impurities made the rivets more brittle in the freezing North Atlantic, and the crashes of the world’s first commercial jetliner, the de Havilland Comet, which revealed the dangers of metal fatigue.

“There are so many other stories that need to be told around how a material failed,” Jaramillo said. “It really cost people money and time and lives. And then through forensic materials science, we understood why it failed and we avoided future failures.”

The MRSEC team is planning to stage public outreach events at the MIT Future Fest.

Looking ahead six years, Jaramillo hopes the center will have become a self-sustaining hub for materials research. 

“I hope that we will have rebuilt the muscle memory to come together in an interdisciplinary way around materials science, and that it should have a bit of a self-sustaining element to it. I hope that we then compete successfully for the next center, and lay the groundwork for the next 60 years.”

MIT Research Administration Services supported the MRSEC proposal development through its Research Development team, which specializes in providing substantive assistance for large and complex research proposals, and in supporting early-career faculty.

MIT faculty expected to be involved in the MRSEC are Rafael Jaramillo, Caroline Ross, Juejun Hu, and Antoine Allanore of DMSE; Moungi Bawendi of Chemistry; Martin Bazant of Chemical Engineering; Nicole Nie and Shuhei Ono of EAPS; and Marin Soljačić, Riccardo Comin, Nuh Gedik, and Long Ju of Physics.

Astronomers at MIT and elsewhere have spotted an extremely bright red spot in the early universe. The object resembles an enormous star, spanning the size of our solar system. But it also is putting out 100 billion times more energy than any known star can physically produce. In fact, such energies are closer to what a black hole might generate.

The curious combination suggests that the red spot is an entirely new type of astrophysical source. The astronomers are calling it a “black hole star.”

In a paper appearing today in the journal Nature, the team presents their analysis of the new object, which they discovered using NASA’s James Webb Space Telescope (JWST). The telescope spotted the bright red dot in the very early universe, just a few hundred million years after the Big Bang.

The scientists conclude that the most likely explanation for the strange red dot is that it is a mashup of a black hole and a star — a combination that has never been observed until now. The object is likely a hugely dense cloud of gas, powered not by standard nuclear fusion, but by a central black hole.

“Our picture of this object is evolving very rapidly,” says lead author Rohan Naidu, a NASA Hubble Fellow and Pappalardo Fellow at MIT’s Kavli Institute for Astrophysics and Space Research (MKI). “We think there is a central black hole that is 100,000 times as massive as the sun. And around this black hole, there would be this very extended envelope of gas that looks like a star the size of the solar system. It’s huge.”

If the bright red dot is indeed a black hole star, it would help to solve the identity of other mysterious “little red dots” that have appeared in nearly every deep space image JWST has taken to date.

“These little red dots seem to be everywhere in the early universe but essentially disappear by the present day,” Naidu says. “What exactly these objects are has been one of the most debated topics of the JWST era.”

The study’s MIT co-authors are MKI Director Robert Simcoe, the Bruno B. Rossi Professor of Experimental Physics; and Wendy Sun ’26, along with collaborators from multiple other institutions.

A singular source

Naidu and his colleagues didn’t intend to find a black hole star. They were looking for the most distant, earliest galaxies, as part of a survey that they named “Mirage or Miracle” (MoM). The team used the JWST to look into deep space, back when the universe was a few hundred million years old. Their goal was to look for galaxies that actually formed at those early times.

“There’s been this puzzle of many bright galaxies showing up at extremely early times,” Naidu says. “What we found was that what looks like an extremely bright early galaxy, aka a ‘miracle,’ in some cases actually could be a ‘mirage.’”

As they looked through JWST’s images for intriguing sources to target with their survey, they noticed a feature that stood out from the rest: a dot that was very red, and very bright.

“When we see something very red in the universe, we often assume that it is surrounded by dust, like soot or ash,” Simcoe explains. “The same way that the wildfire smoke from Canada recently made the sky in Boston look bright red, astronomical objects can also appear redder than their intrinsic color when you see them through a veil of dust.”

But there were other signatures in the light that didn’t quite match up with what physicists expect from dust. The team also observed another strange pattern: The dot’s light was extremely bright, except below certain wavelengths, where the light completely disappeared. 

This spectral drop-off is known as a “Balmer break” — a signature traditionally associated with dense gas soaking up photons in the atmospheres of stars that are a few hundred millions of years old. Vega, one of the brightest stars in the night sky shows exactly this pattern. 

“The break we observed in this object is the deepest break we have ever observed in any object, ruling out ‘ordinary’ stars as the source,” Naidu says. “But it made us wonder if we were seeing a new kind of ‘stellar atmosphere,’ but on a spectacular scale.” 

What’s more, the red dot’s light contained almost no signature of metals or any elements other than hydrogen and helium. “It was truly singular in so many ways,” Naidu says.

Pure light

To puzzle out what the source of the red dot could be, the team ran simulations of different scenarios to see what combination of astrophysical features could produce the red dot’s distinctive color.

“We started to ask: Could you make something that red using just hydrogen, without any dust?” Simcoe says. “To our surprise, it turns out you can, if you have an extremely dense screen of hydrogen, so dense that it looks more like the surface of an enormous star than a wispy interstellar nebula.”

Their simulations pointed to the red dot possibly being some powerful enshrouded energy source, surrounded by an extremely dense cocoon of hydrogen. If this were the case, it would explain the light-blocking Balmer break and the lack of anything other than hydrogen and helium that the astronomers observed. But it still wouldn’t explain the object’s extreme brightness.

“You have something that looks a bit like a star but is 100 billion times brighter,” Naidu says. “That means you can’t be powering this by nuclear fusion, which is the energy source that sits at the heart of all the stars we have.”

Black holes, however, routinely produce energy at the scales the team observed. Naidu and his colleagues incorporated an active, accreting black hole into their simulations of the hydrogen-cocooned star and varied the black hole’s mass, along with other parameters. They then compared the resulting brightness of the simulated “black hole star” with the brightness that JWST observed from the red dot.

From these simulations, they found the closest match, and concluded that the most likely scenario to explain the red dot, is a black hole star. Specifically, the object likely contains a central black hole that is about 100,000 times as massive as the sun. This powerful core is surrounded by a dense, star-like cocoon of hydrogen that is roughly the size of the solar system.

The team has named the object MoM-BH*-1, after the survey that detected it, as well as the moniker “black hole star – one,” which implies that the object is the first of others. The researchers suspect that black hole stars could explain many of the other little red dots that appear in JWST images. Those objects are not as bright as MoM-BH*-1. 

“Every little red dot is consistent with being a black hole star, embedded in a generic early galaxy,” Naidu says. “But what is special about MoM-BH*-1 is, the black hole star is essentially completely outshining its surrounding host galaxy, such that we’re seeing pure black hole star light.”

This research was supported, in part, by the MIT Department of Physics, NASA, and the Space Telescope Science Institute.

It is a transformation for the ages: The Metropolitan Storage Warehouse in Cambridge, Massachusetts, is opening as the new home of MIT’s School of Architecture and Planning, after a makeover turning the century-old storage facility into a light-infused center for teaching, research, and public engagement.

The massive structure is a unique addition to daily life at the Institute. A hulking brick building and local landmark over 500 feet long and five stories high, the Met Warehouse now stands as a remarkable feat of architecture, engineering, and “adaptive reuse.” It includes four segments of glass walls, double-height studio spaces, copious common areas, and building-long walkways overlooking the work areas on all five floors — a 21st-century variation on the Infinite Corridor in MIT’s main group buildings. 

Designed by the architecture studio Diller Scofidio + Renfro (DS+R), the Met Warehouse is intended to serve as a new campus hub. Beyond work studios, offices, and classrooms, there is an auditorium, galleries, and common spaces where MIT scholars and students can learn and design together, and the public can engage in lectures, exhibitions, and other programming.

“Walking through the Met Warehouse, everywhere you look you see the artful melding of the original architecture with the new design. It’s a perfect expression of the historical importance of architecture at MIT and of the creative promise of this new hub,” says MIT President Sally Kornbluth. “The new Met Warehouse will create a central home for design at MIT, and together with the new Linde Music Building, the presence of the Met will create a magnetic new west campus district for arts and design.”

Faculty, staff, and students have started moving into the Met Warehouse this month. The School of Architecture and Planning will stage a ceremonial procession into the building on Sept. 8, with a formal dedication event on Oct. 1, and a day welcoming the general public on Oct. 3 as part of MIT Future Fest. 

The Met Warehouse’s conversion began in the late 2010s, championed by Hashim Sarkis, the dean of MIT’s School of Architecture and Planning, and his collaborators. They envisioned a new and dedicated space for architecture, design, and planning at MIT — while reusing an existing structure for that purpose.

“I think it sends a very good message that this vanguard school of architecture, at the Massachusetts Institute of Technology, is moving into a historic building and adapting it for the future,” says Sarkis, the Elizabeth and James Killian 1926 Professor. “This is a big statement on the part of MIT.” 

Sarkis adds: “We’re expecting the Met to facilitate a very vibrant in-person culture. The vitality of interpersonal connection will be highlighted in the building. The faculty and the students wanted more research space, more space for exhibitions and galleries, and more spaces that enable what we do best, which is to work together. Design is about collaboration, and planning is about community.”

From fortress to studio 

First opened in 1894 and completed in 1923, the building known as the Metropolitan Storage Warehouse long stood as a forbidding, fortress-like facility, with some tiny window slits. Only a few people had reason to venture inside. Visible from across the river in Boston, the Met Warehouse was a landmark, an advertisement of services, and a curiosity. It had about 1,500 storage spaces inside, and few other uses. 

MIT acquired the building in 1962, and by 2015 it was no longer used for storage. That raised a question: What comes next? Over time, the idea of moving the School of Architecture and Planning into the Met Warehouse took hold. That left the hard work of designing and transforming the building into a place that people could inhabit, while respecting the historically designated façade’s monolithic qualities. 

To create such a thoroughgoing transformation, MIT engaged DS+R, known for the design of high-profile cultural and institutional projects, including the Broad Museum in Los Angeles; the Institute of Contemporary Art in Boston; the Shed, a nonprofit cultural and performing-arts space in New York City; and, not least, the transformation of a postindustrial rail line into New York City’s High Line. Shawmut Design and Construction managed the renovation, and the entire endeavor was made possible by the generous philanthropic support of MIT alumni, volunteers, and friends. 

Significantly, some of the signature projects of DS+R, including the High Line and the renovation of Alice Tully Hall at Lincoln Center in New York, involved updating and adaptively reusing existing structures. For the Met Warehouse, this meant a revamping of the interior, creating new workspaces, new ways to help people circulate through the massive building, and new ways to bring light inside the structure. In addition to the glass wall segments, the architects expanded the building’s windows, added a connective staircase, and found additional ways to let light and air permeate throughout. 

“Our thinking was always around trying to bring communities on campus together, knowing there would be a convergence of labs, classrooms, resource spaces, and disciplines,” says Elizabeth Diller, founding partner at DS+R. “The big challenge from the start of the project was the building itself. The building is stubborn and big and heavy, and it was conceived to hold furniture and suitcases and pianos, not humans.”

When thinking through the project, Diller adds, “The first thing was assessing the building itself and its potential, and our ability to perforate it [allowing light] and to create new spaces inside of it. … We saw the potential, because of the structure, that it could endure some surgery.”

“The choice by MIT and Hashim Sarkis to adaptively reuse a building as a center for design represents a bold vision,” says Benjamin Gilmartin, partner at DS+R. “It’s a courageous idea: that the future of design and architecture very much lives in the reuse of structures we already have.”

MIT campus leaders say they are delighted with the outcome. 

“The way the building is structured, the architects, Liz Diller, Ben Gilmartin, and their team, have been unbelievably shrewd in understanding our culture and respecting it while transforming the building,” Sarkis says. “That transformation enables the things we want, which include collaborative work, while also combining instruction and research.”

The makeover of the building also represents a collaboration between the City of Cambridge and MIT. Because the Metropolitan Storage Warehouse is a historically listed building, the city had to approve the substantial exterior renovations — such as on the north side, where several glass walls now cascade from the top of the Met to ground level. On the south side, the architects preserved many of the small storage units, redesigning them as offices with an innovative “skin” of new windows.

“That was one of the big decisions, based on light and the sensitivities of the history, that the large studios would be facing the north, and extracted from the north side of the building,” Diller explains. “Which left a lot of peripheral areas to act as small-scale and more intimate spaces, offices, and other types of spaces as needed.”

Indeed, the architects emphasize, the redesign of the Met Warehouse is not simply an overhaul; the plan significantly reflects the longtime interior structure of the building, too. 

“It wasn’t just about converting the shell,” Gilmartin says. “It was about trying to find a balance and determining how much was already there [structurally] that we could use as a fabric.” 

That historical fabric is evident through one of the building’s signature features: The old brick structure in key places is exposed to view, next to many places where the architects made dramatic cuts to create platforms for light-filled studio spaces. Students, designers, and visitors can see both how the old Met Warehouse was built and how the new version of it was created.

“The building itself can be a teaching tool,” Diller says. “When we did those extractions from the building, we left our intervention exposed, so there’s a kind of conversation between a contemporary strategy and the historical building. The traces are all there; they’re all revealed.”

Educators at the Institute view the building in a similar manner as they think about architectural teaching broadly.

“Our move to the Met is an exciting physical transition for the school, and an occasion for us to articulate the shifts in architectural education we have been undertaking,” says Ana Miljački, the Francis White Davis Professor at MIT and head of the Department of Architecture. “Making our home in the building will be part of our rethinking of the discipline, the profession, and our pedagogical tasks.” 

Five stories, five blocks, one vision

As originally constructed, the Met Warehouse had five contiguous segments. Given that it is also five stories high, the building has 25 natural segments, in a sense. A wide range of activity will be housed inside it, including several core parts of the School of Architecture and Planning: the Department of Architecture, the Department of Urban Studies and Planning (DUSP), and the Norman B. Leventhal Center for Advanced Urbanism. (The MIT Media Lab, the Art, Culture, and Technology Program, and the Center for Real Estate, all part of the School of Architecture and Planning, will remain in their existing locations on campus.)

The MIT Morningside Academy for Design (MAD), a campus-wide center promoting interdisciplinary design work, will also be located in the Met Warehouse, helping to further establish the building as the essential hub of design and planning work on campus.

Many MIT scholars say they welcome the opportunity to bring so many related programs into greater proximity with each other, along with all the physical assets the Met Warehouse will provide. 

“At MIT we have fewer boundaries, less conventions, and we bump into each other on campus,” says Jinhua Zhao, the Class of 1941 Professor and head of DUSP. “I have always appreciated this spirit since I first came here as a student and walked along the Infinite Corridor. A lot of places value interdisciplinary research. At MIT, you can’t help it happening. I believe the new Met Warehouse will expand that custom.”

Those who saw the inside of the building in its old days as a storage space, and are moving into it now, are deeply impressed by the complete readaptation of the Met Warehouse and the provision of new “commons” spaces for the campus.

“It’s almost inconceivable that this brick box, which was not designed for human habitation but to store objects, has been opened up, through the work of Diller Scofidio + Renfro,” says John Ochsendorf, the Class of 1942 Professor and director of MAD. “Our hope is you will find vibrant cross-fertilization across disciplines, across the School of Architecture and Planning, but also across all of MIT. That’s really important.”

Indeed, as Ochsendorf and others have noted, the building figures to produce its own urban dynamics within its monumental walls. 

“As you go up into the building, you will find different neighborhoods concerned with different aspects of design,” Ochsendorf says. “These are all areas pushing frontiers in research and education and design of the built environment, which interact with so many of the pressing issues facing humanity. We’re excited to create new neighborhoods of inquiry with the building.”

That is certainly part of the intention, the architects say. 

“There are a lot of opportunities for smaller groupings of people to be organized in ways that are visible and connected to the larger shared spaces but also offer the prospect of retreat in different places to work,” Gilmartin observes. 

“The challenges facing cities cannot be addressed by any one discipline,” says Sarah Williams, director of the Norman B. Leventhal Center for Advanced Urbanism. “Innovation comes from bringing together all the fields that shape — and are shaped by — the built environment. The Met Warehouse gives us a place to work across those boundaries, inspiring new ways to imagine and build the future of our cities.”

Sarkis, for his part, professes some happy relief that the long-held conception of the Met Warehouse is finally becoming reality. The building, he thinks, will influence the flow of people through MIT’s campus, bringing a transformative multiuse space into the daily lives of students, faculty, staff, and the public.

“It is going to be a new center of gravity for the campus,” Sarkis says. 

MIT News will offer a further look at the Met Warehouse’s transformative architecture in concert with the Sept. 8 procession, as well as coverage of events from the formal dedication weekend in October.