All posts by admin

By now, ChatGPT, Claude, and other large language models have accumulated so much human knowledge that they’re far from simple answer-generators; they can also express abstract concepts, such as certain tones, personalities, biases, and moods. However, it’s not obvious exactly how these models represent abstract concepts to begin with from the knowledge they contain.

Now a team from MIT and the University of California San Diego has developed a way to test whether a large language model (LLM) contains hidden biases, personalities, moods, or other abstract concepts. Their method can zero in on connections within a model that encode for a concept of interest. What’s more, the method can then manipulate, or “steer” these connections, to strengthen or weaken the concept in any answer a model is prompted to give.

The team proved their method could quickly root out and steer more than 500 general concepts in some of the largest LLMs used today. For instance, the researchers could home in on a model’s representations for personalities such as “social influencer” and “conspiracy theorist,” and stances such as “fear of marriage” and “fan of Boston.” They could then tune these representations to enhance or minimize the concepts in any answers that a model generates.

In the case of the “conspiracy theorist” concept, the team successfully identified a representation of this concept within one of the largest vision language models available today. When they enhanced the representation, and then prompted the model to explain the origins of the famous “Blue Marble” image of Earth taken from Apollo 17, the model generated an answer with the tone and perspective of a conspiracy theorist.

The team acknowledges there are risks to extracting certain concepts, which they also illustrate (and caution against). Overall, however, they see the new approach as a way to illuminate hidden concepts and potential vulnerabilities in LLMs, that could then be turned up or down to improve a model’s safety or enhance its performance.

“What this really says about LLMs is that they have these concepts in them, but they’re not all actively exposed,” says Adityanarayanan “Adit” Radhakrishnan, assistant professor of mathematics at MIT. “With our method, there’s ways to extract these different concepts and activate them in ways that prompting cannot give you answers to.”

The team published their findings today in a study appearing in the journal Science. The study’s co-authors include Radhakrishnan, Daniel Beaglehole and Mikhail Belkin of UC San Diego, and Enric Boix-Adserà of the University of Pennsylvania.

A fish in a black box

As use of OpenAI’s ChatGPT, Google’s Gemini, Anthropic’s Claude, and other artificial intelligence assistants has exploded, scientists are racing to understand how models represent certain abstract concepts such as “hallucination” and “deception.” In the context of an LLM, a hallucination is a response that is false or contains misleading information, which the model has “hallucinated,” or constructed erroneously as fact.

To find out whether a concept such as “hallucination” is encoded in an LLM, scientists have often taken an approach of “unsupervised learning” — a type of machine learning in which algorithms broadly trawl through unlabeled representations to find patterns that might relate to a concept such as “hallucination.” But to Radhakrishnan, such an approach can be too broad and computationally expensive.

“It’s like going fishing with a big net, trying to catch one species of fish. You’re gonna get a lot of fish that you have to look through to find the right one,” he says. “Instead, we’re going in with bait for the right species of fish.”

He and his colleagues had previously developed the beginnings of a more targeted approach with a type of predictive modeling algorithm known as a recursive feature machine (RFM). An RFM is designed to directly identify features or patterns within data by leveraging a mathematical mechanism that neural networks — a broad category of AI models that includes LLMs — implicitly use to learn features.

Since the algorithm was an effective, efficient approach for capturing features in general, the team wondered whether they could use it to root out representations of concepts, in LLMs, which are by far the most widely used type of neural network and perhaps the least well-understood.

“We wanted to apply our feature learning algorithms to LLMs to, in a targeted way, discover representations of concepts in these large and complex models,” Radhakrishnan says.

Converging on a concept

The team’s new approach identifies any concept of interest within a LLM and “steers” or guides a model’s response based on this concept. The researchers looked for 512 concepts within five classes: fears (such as of marriage, insects, and even buttons); experts (social influencer, medievalist); moods (boastful, detachedly amused); a preference for locations (Boston, Kuala Lumpur); and personas (Ada Lovelace, Neil deGrasse Tyson).

The researchers then searched for representations of each concept in several of today’s large language and vision models. They did so by training RFMs to recognize numerical patterns in an LLM that could represent a particular concept of interest.

A standard large language model is, broadly, a neural network that takes a natural language prompt, such as “Why is the sky blue?” and divides the prompt into individual words, each of which is encoded mathematically as a list, or vector, of numbers. The model takes these vectors through a series of computational layers, creating matrices of many numbers that, throughout each layer, are used to identify other words that are most likely to be used to respond to the original prompt. Eventually, the layers converge on a set of numbers that is decoded back into text, in the form of a natural language response.

The team’s approach trains RFMs to recognize numerical patterns in an LLM that could be associated with a specific concept. As an example, to see whether an LLM contains any representation of a “conspiracy theorist,” the researchers would first train the algorithm to identify patterns among LLM representations of 100 prompts that are clearly related to conspiracies, and 100 other prompts that are not. In this way, the algorithm would learn patterns associated with the conspiracy theorist concept. Then, the researchers can mathematically modulate the activity of the conspiracy theorist concept by perturbing LLM representations with these identified patterns. 

The method can be applied to search for and manipulate any general concept in an LLM. Among many examples, the researchers identified representations and manipulated an LLM to give answers in the tone and perspective of a “conspiracy theorist.” They also identified and enhanced the concept of “anti-refusal,” and showed that whereas normally, a model would be programmed to refuse certain prompts, it instead answered, for instance giving instructions on how to rob a bank.

Radhakrishnan says the approach can be used to quickly search for and minimize vulnerabilities in LLMs. It can also be used to enhance certain traits, personalities, moods, or preferences, such as emphasizing the concept of “brevity” or “reasoning” in any response an LLM generates. The team has made the method’s underlying code publicly available.

“LLMs clearly have a lot of these abstract concepts stored within them, in some representation,” Radhakrishnan says. “There are ways where, if we understand these representations well enough, we can build highly specialized LLMs that are still safe to use but really effective at certain tasks.”

This work was supported, in part, by the National Science Foundation, the Simons Foundation, the TILOS institute, and the U.S. Office of Naval Research. 

Large language models (LLMs) have been championed as tools that could democratize access to information worldwide, offering knowledge in a user-friendly interface regardless of a person’s background or location. However, new research from MIT’s Center for Constructive Communication (CCC) suggests these artificial intelligence systems may actually perform worse for the very users who could most benefit from them.

A study conducted by researchers at CCC, which is based at the MIT Media Lab, found that state-of-the-art AI chatbots — including OpenAI’s GPT-4, Anthropic’s Claude 3 Opus, and Meta’s Llama 3 — sometimes provide less-accurate and less-truthful responses to users who have lower English proficiency, less formal education, or who originate from outside the United States. The models also refuse to answer questions at higher rates for these users, and in some cases, respond with condescending or patronizing language.

“We were motivated by the prospect of LLMs helping to address inequitable information accessibility worldwide,” says lead author Elinor Poole-Dayan SM ’25, a technical associate in the MIT Sloan School of Management who led the research as a CCC affiliate and master’s student in media arts and sciences. “But that vision cannot become a reality without ensuring that model biases and harmful tendencies are safely mitigated for all users, regardless of language, nationality, or other demographics.”

A paper describing the work, “LLM Targeted Underperformance Disproportionately Impacts Vulnerable Users,” was presented at the AAAI Conference on Artificial Intelligence in January.

Systematic underperformance across multiple dimensions

For this research, the team tested how the three LLMs responded to questions from two datasets: TruthfulQA and SciQ. TruthfulQA is designed to measure a model’s truthfulness (by relying on common misconceptions and literal truths about the real world), while SciQ contains science exam questions testing factual accuracy. The researchers prepended short user biographies to each question, varying three traits: education level, English proficiency, and country of origin.

Across all three models and both datasets, the researchers found significant drops in accuracy when questions came from users described as having less formal education or being non-native English speakers. The effects were most pronounced for users at the intersection of these categories: those with less formal education who were also non-native English speakers saw the largest declines in response quality.

The research also examined how country of origin affected model performance. Testing users from the United States, Iran, and China with equivalent educational backgrounds, the researchers found that Claude 3 Opus in particular performed significantly worse for users from Iran on both datasets.

“We see the largest drop in accuracy for the user who is both a non-native English speaker and less educated,” says Jad Kabbara, a research scientist at CCC and a co-author on the paper. “These results show that the negative effects of model behavior with respect to these user traits compound in concerning ways, thus suggesting that such models deployed at scale risk spreading harmful behavior or misinformation downstream to those who are least able to identify it.”

Refusals and condescending language

Perhaps most striking were the differences in how often the models refused to answer questions altogether. For example, Claude 3 Opus refused to answer nearly 11 percent of questions for less educated, non-native English-speaking users — compared to just 3.6 percent for the control condition with no user biography.

When the researchers manually analyzed these refusals, they found that Claude responded with condescending, patronizing, or mocking language 43.7 percent of the time for less-educated users, compared to less than 1 percent for highly educated users. In some cases, the model mimicked broken English or adopted an exaggerated dialect.

The model also refused to provide information on certain topics specifically for less-educated users from Iran or Russia, including questions about nuclear power, anatomy, and historical events — even though it answered the same questions correctly for other users.

“This is another indicator suggesting that the alignment process might incentivize models to withhold information from certain users to avoid potentially misinforming them, although the model clearly knows the correct answer and provides it to other users,” says Kabbara.

Echoes of human bias

The findings mirror documented patterns of human sociocognitive bias. Research in the social sciences has shown that native English speakers often perceive non-native speakers as less educated, intelligent, and competent, regardless of their actual expertise. Similar biased perceptions have been documented among teachers evaluating non-native English-speaking students.

“The value of large language models is evident in their extraordinary uptake by individuals and the massive investment flowing into the technology,” says Deb Roy, professor of media arts and sciences, CCC director, and a co-author on the paper. “This study is a reminder of how important it is to continually assess systematic biases that can quietly slip into these systems, creating unfair harms for certain groups without any of us being fully aware.”

The implications are particularly concerning given that personalization features — like ChatGPT’s Memory, which tracks user information across conversations — are becoming increasingly common. Such features risk differentially treating already-marginalized groups.

“LLMs have been marketed as tools that will foster more equitable access to information and revolutionize personalized learning,” says Poole-Dayan. “But our findings suggest they may actually exacerbate existing inequities by systematically providing misinformation or refusing to answer queries to certain users. The people who may rely on these tools the most could receive subpar, false, or even harmful information.”

Originally from the small Balkan country of Montenegro, Strahinja (Strajo) Janjusevic says his life has unfolded in unexpected ways, for which he is deeply grateful. After graduating from high school, he was selected to represent his country in the United States, studying cyber operations and computer science at the U.S. Naval Academy in Annapolis, Maryland. He has since continued his cybersecurity studies and is currently a second-year master’s student in the Technology and Policy Program (TPP), hosted by the MIT Institute for Data, Systems, and Society (IDSS). His research with the MIT Laboratory for Information and Decision Systems (LIDS) and the MIT Maritime Consortium team aims to improve the cybersecurity of critical maritime infrastructure using artificial intelligence, considering both the technology and policy frameworks of solutions.

“My current research focuses on applying AI techniques to cybersecurity problems and examining the policy implications of these advancements, especially in the context of maritime cybersecurity,” says Janjusevic. “Representing my country at the highest levels of education and industry has given me a unique perspective on cybersecurity challenges.”

Janjusevic’s pathway from Montenegro to Maryland was created by a program that allows selected students from allied countries to attend the U.S. Naval Academy. Janjusevic graduated with a dual bachelor’s degree in cyber operations and computer science. His undergraduate experience provided opportunities to collaborate with the U.S. military and the National Security Agency, exposing him to high-level cybersecurity operations and fueling his interest in tackling complex cybersecurity challenges. During his undergraduate studies, he also interned with Microsoft, developing tools for cloud incident response, and with NASA, visualizing solar data for research applications.

Following his graduation, he realized that he still needed more knowledge, particularly in the area of AI and cybersecurity. TPP appealed to him immediately because of its dual emphasis on rigorous engineering innovation and the policy analysis needed to deploy it effectively. Janjusevic’s experiences at TPP have been a big change from his time at the U.S. Naval Academy, with a different pace and environment. He has especially appreciated being able to broaden his understanding about a variety of research domains and apply the discipline and knowledge he earned during his time at the academy.

“My TPP experience has been amazing,” says Janjusevic. “The cohort is really small, so it feels like a family, and everyone is working on diverse, high-impact problems.”

Mitigating the risks of emerging technologies

Janjusevic’s thesis brings together disciplines of cybersecurity, AI and deep learning, and control theory and physics, focusing on securing maritime cyber-physical systems — in particular, large legacy ships. The hacking of these ships’ networks can result in substantial damage to national security, as well as serious economic effects.

“Strajo is working to outsmart maritime GPS spoofing,” says Saurabh Amin, the Edmund K. Turner Professor in Civil Engineering. “Such attacks have already lured vessels off course in contested waters. His approach layers physics-based trajectory models with deep learning, catching threats that no single method can detect alone. His expertise has been very helpful in advancing our work on threat modeling and attack detection.”

The research utilizes advanced threat modeling and vessel dynamics to train AI systems to distinguish between legitimate maneuvers and spoofed signals. It involves building a framework that employs an internal LSTM (long short-term memory) autoencoder to analyze signal integrity, while simultaneously using a physics-based forecaster to predict the vessel’s movement based on environmental factors like wind and the sea state. By comparing these predictions against reported GPS positions, the system can effectively distinguish between natural sensor noise and malicious spoofing attacks. This hybrid framework is designed to empower, not replace, human operators, providing verified navigation data that allows watch standers to distinguish technical glitches from strategic cyberattacks.

Janjusevic has been able to enhance his academic research with industry experience. In summer 2025, he interned with the Network Detection team at the AI cybersecurity company Vectra AI. There, he investigated potential threats new technologies can bring, particularly AI agents and the model context protocol (MCP) — the emerging standard for AI agent communication. His research demonstrated how this technology could be repurposed for autonomous hacking operations and advanced command and control. This work on the security risks of agentic AI was recently presented in the preprint, “Hiding in the AI Traffic: Abusing MCP for LLM-Powered Agentic Red Teaming.”

“I was able to gain practical insights and hands-on experience into how a data science team uses AI models to detect anomalies in a network,” says Janjusevic. “This work within industry directly informed the anomaly detection models in my research.”

International policy perspective

“Strajo brings not just a high level of intelligence and energy to his work on cyber-physical security for merchant vessels, but also a strong instinct from his Navy training that resonates deeply with the research effort and grounds it in actionable policy,” says Fotini Christia, the Ford International Professor of the Social Sciences, director of IDSS, and a leader of the MIT Maritime Consortium.

Janjusevic participates in the cybersecurity efforts of the Maritime Consortium, a collaboration between academia, industry, and regulatory agencies focused on developing technological solutions, industry standards, and policies. The consortium includes cooperation with some international members, including from Singapore and South Korea.

“In AI cybersecurity, the policy element is really important, as the field is so fast-moving and the consequences of hacking can be so dangerous,” says Janjusevic. “I think there’s still a lot of need for policy work in this space.”

Janjusevic is also currently helping to organize two upcoming major conferences: the Harvard European Conference in February, which will convene officials and diplomats from across the globe, and the Technology and National Security Conference in April, a collaboration of Harvard and MIT that brings together top leaders from government, industry, and academia to tackle critical challenges in national security.

“I’m striving to find a position where I can influence and advance the cybersecurity field with AI, while at the same time leading collaboration and innovation between the United States and Montenegro,” says Janjusevic. “My goal is to be a bridge between Europe and the U.S. in this space of national security, AI, and cybersecurity, bringing my knowledge to both sides.”

Generative artificial intelligence models have been used to create enormous libraries of theoretical materials that could help solve all kinds of problems. Now, scientists just have to figure out how to make them.

In many cases, materials synthesis is not as simple as following a recipe in the kitchen. Factors like the temperature and length of processing can yield huge changes in a material’s properties that make or break its performance. That has limited researchers’ ability to test millions of promising model-generated materials.

Now, MIT researchers have created an AI model that guides scientists through the process of making materials by suggesting promising synthesis routes. In a new paper, they showed the model delivers state-of-the-art accuracy in predicting effective synthesis pathways for a class of materials called zeolites, which could be used to improve catalysis, absorption, and ion exchange processes. Following its suggestions, the team synthesized a new zeolite material that showed improved thermal stability.

The researchers believe their new model could break the biggest bottleneck in the materials discovery process.

“To use an analogy, we know what kind of cake we want to make, but right now we don’t know how to bake the cake,” says lead author Elton Pan, a PhD candidate in MIT’s Department of Materials Science and Engineering (DMSE). “Materials synthesis is currently done through domain expertise and trial and error.”

The paper describing the work appears today in Nature Computational Science. Joining Pan on the paper are Soonhyoung Kwon ’20, PhD ’24; DMSE postdoc Sulin Liu; chemical engineering PhD student Mingrou Xie; DMSE postdoc Alexander J. Hoffman; Research Assistant Yifei Duan SM ’25; DMSE visiting student Thorben Prein; DMSE PhD candidate Killian Sheriff; MIT Robert T. Haslam Professor in Chemical Engineering Yuriy Roman-Leshkov; Valencia Polytechnic University Professor Manuel Moliner; MIT Paul M. Cook Career Development Professor Rafael Gómez-Bombarelli; and MIT Jerry McAfee Professor in Engineering Elsa Olivetti.

Learning to bake

Massive investments in generative AI have led companies like Google and Meta to create huge databases filled with material recipes that, at least theoretically, have properties like high thermal stability and selective absorption of gases. But making those materials can require weeks or months of careful experiments that test specific reaction temperatures, times, precursor ratios, and other factors.

“People rely on their chemical intuition to guide the process,” Pan says. “Humans are linear. If there are five parameters, we might keep four of them constant and vary one of them linearly. But machines are much better at reasoning in a high-dimensional space.”

The synthesis process of materials discovery now often takes the most time in a material’s journey from hypothesis to use.

To help scientists navigate that process, the MIT researchers trained a generative AI model on over 23,000 material synthesis recipes described over 50 years of scientific papers. The researchers iteratively added random “noise” to the recipes during training, and the model learned to de-noise and sample from the random noise to find promising synthesis routes.

The result is DiffSyn, which uses an approach in AI known as diffusion.

“Diffusion models are basically a generative AI model like ChatGPT, but more like the DALL-E image generation model,” Pan says. “During inference, it converts noise into meaningful structure by subtracting a little bit of noise at each step. In this case, the ‘structure’ is the synthesis route for a desired material.”

When a scientist using DiffSyn enters a desired material structure, the model offers some promising combinations of reaction temperatures, reaction times, precursor ratios, and more.

“It basically tells you how to bake your cake,” Pan says. “You have a cake in mind, you feed it into the model, the model spits out the synthesis recipes. The scientist can pick whichever synthesis path they want, and there are simple ways to quantify the most promising synthesis path from what we provide, which we show in our paper.”

To test their system, the researchers used DiffSyn to suggest novel synthesis paths for a zeolite, a material class that is complex and takes time to form into a testable material.

“Zeolites have a very high-dimensional synthesis space,” Pan says. “Zeolites also tend to take days or weeks to crystallize, so the impact [of finding the best synthesis pathway faster] is much higher than other materials that crystallize in hours.”

The researchers were able to make the new zeolite material using synthesis pathways suggested by DiffSyn. Subsequent testing revealed the material had a promising morphology for catalytic applications.

“Scientists have been trying out different synthesis recipes one by one,” Pan says. “That makes them very time-consuming. This model can sample 1,000 of them in under a minute. It gives you a very good initial guess on synthesis recipes for completely new materials.”

Accounting for complexity

Previously, researchers have built machine-learning models that mapped a material to a single recipe. Those approaches do not take into account that there are different ways to make the same material.

DiffSyn is trained to map material structures to many different possible synthesis paths. Pan says that is better aligned with experimental reality.

“This is a paradigm shift away from one-to-one mapping between structure and synthesis to one-to-many mapping,” Pan says. “That’s a big reason why we achieved strong gains on the benchmarks.”

Moving forward, the researchers believe the approach should work to train other models that guide the synthesis of materials outside of zeolites, including metal-organic frameworks, inorganic solids, and other materials that have more than one possible synthesis pathway.

“This approach could be extended to other materials,” Pan says. “Now, the bottleneck is finding high-quality data for different material classes. But zeolites are complicated, so I can imagine they are close to the upper-bound of difficulty. Eventually, the goal would be interfacing these intelligent systems with autonomous real-world experiments, and agentic reasoning on experimental feedback to dramatically accelerate the process of materials design.”

The work was supported by MIT International Science and Technology Initiatives (MISTI), the National Science Foundation, Generalitat Vaslenciana, the Office of Naval Research, ExxonMobil, and the Agency for Science, Technology and Research in Singapore.

What if ultrasound imaging is no longer confined to hospitals? Patients with chronic conditions, such as hypertension and heart failure, could be monitored continuously in real-time at home or on the move, giving health care practitioners ongoing clinical insights instead of the occasional snapshots — a scan here and a check-up there. This shift from reactive, hospital-based care to preventative, community and home-based care could enable earlier detection and timely intervention, and truly personalized care.

Bringing this vision to reality, the Singapore-MIT Alliance for Research and Technology (SMART), MIT’s research enterprise in Singapore, has launched a new collaborative research project: Wearable Imaging for Transforming Elderly Care (WITEC). 

WITEC marks a pioneering effort in wearable technology, medical imaging, research, and materials science. It will be dedicated to foundational research and development of the world’s first wearable ultrasound imaging system capable of 48-hour intermittent cardiovascular imaging for continuous and real-time monitoring and diagnosis of chronic conditions such as hypertension and heart failure. 

This multi-million dollar, multi-year research program, supported by the National Research Foundation (NRF) Singapore under its Campus for Research Excellence and Technological Enterprise program, brings together top researchers and expertise from MIT, Nanyang Technological University (NTU Singapore), and the National University of Singapore (NUS). Tan Tock Seng Hospital (TTSH) is WITEC’s clinical collaborator and will conduct patient trials to validate long-term heart imaging for chronic cardiovascular disease management.

“Addressing society’s most pressing challenges requires innovative, interdisciplinary thinking. Building on SMART’s long legacy in Singapore as a hub for research and innovation, WITEC will harness interdisciplinary expertise — from MIT and leading institutions in Singapore — to advance transformative research that creates real-world impact and benefits Singapore, the U.S., and societies all over. This is the kind of collaborative research that not only pushes the boundaries of knowledge, but also redefines what is possible for the future of health care,” says Bruce Tidor, chief executive officer and interim director of SMART, who is also an MIT professor of biological engineering and electrical engineering and computer science.

Industry-leading precision equipment and capabilities

To support this work, WITEC’s laboratory is equipped with advanced tools, including Southeast Asia’s first sub-micrometer 3D printer and the latest Verasonics Vantage NXT 256 ultrasonic imaging system, which is the first unit of its kind in Singapore.

Unlike conventional 3D printers that operate at millimeter or micrometer scales, WITEC’s 3D printer can achieve sub‑micrometer resolution, allowing components to be fabricated at the level of single cells or tissue structures. With this capability, WITEC researchers can prototype bioadhesive materials and device interfaces with unprecedented accuracy — essential to ensuring skin‑safe adhesion and stable, long‑term imaging quality.

Complementing this is the latest Verasonics ultrasonic imaging system. Equipped with a new transducer adapter and supporting a significantly larger number of probe control channels than existing systems, it gives researchers the freedom to test highly customized imaging methods. This allows more complex beamforming, higher‑resolution image capture, and integration with AI‑based diagnostic models — opening the door to long‑duration, real‑time cardiovascular imaging not possible with standard hospital equipment.

Together, these technologies allow WITEC to accelerate the design, prototyping, and testing of its wearable ultrasound imaging system, and to demonstrate imaging quality on phantoms and healthy subjects.

Transforming chronic disease care through wearable innovation 

Chronic diseases are rising rapidly in Singapore and globally, especially among the aging population and individuals with multiple long-term conditions. This trend highlights the urgent need for effective home-based care and easy-to-use monitoring tools that go beyond basic wellness tracking.

Current consumer wearables, such as smartwatches and fitness bands, offer limited physiological data like heart rate or step count. While useful for general health, they lack the depth needed to support chronic disease management. Traditional ultrasound systems, although clinically powerful, are bulky, operator-dependent, can only be deployed episodically within the hospitals, and are limited to snapshots in time, making them unsuitable for long-term, everyday use.

WITEC aims to bridge this gap with its wearable ultrasound imaging system that uses bioadhesive technology to enable up to 48 hours of uninterrupted imaging. Combined with AI-enhanced diagnostics, the innovation is aimed at supporting early detection, home-based pre-diagnosis, and continuous monitoring of chronic diseases.

Beyond improving patient outcomes, this innovation could help ease labor shortages by freeing up ultrasound operators, nurses, and doctors to focus on more complex care, while reducing demand for hospital beds and resources. By shifting monitoring to homes and communities, WITEC’s technology will enable patient self-management and timely intervention, potentially lowering health-care costs and alleviating the increasing financial and manpower pressures of an aging population.

Driving innovation through interdisciplinary collaboration

WITEC is led by the following co-lead principal investigators: Xuanhe Zhao, professor of mechanical engineering and professor of civil and environmental engineering at MIT; Joseph Sung, senior vice president of health and life sciences at NTU Singapore and dean of the Lee Kong Chian School of Medicine (LKCMedicine); Cher Heng Tan, assistant dean of clinical research at LKCMedicine; Chwee Teck Lim, NUS Society Professor of Biomedical Engineering at NUS and director of the Institute for Health Innovation and Technology at NUS; and Xiaodong Chen, distinguished university professor at the School of Materials Science and Engineering within NTU. 

“We’re extremely proud to bring together an exceptional team of researchers from Singapore and the U.S. to pioneer core technologies that will make wearable ultrasound imaging a reality. This endeavor combines deep expertise in materials science, data science, AI diagnostics, biomedical engineering, and clinical medicine. Our phased approach will accelerate translation into a fully wearable platform that reshapes how chronic diseases are monitored, diagnosed and managed,” says Zhao, who serves as a co-lead PI of WITEC.

Research roadmap with broad impact across health care, science, industry, and economy

Bringing together leading experts across interdisciplinary fields, WITEC will advance foundational work in soft materials, transducers, microelectronics, data science and AI diagnostics, clinical medicine, and biomedical engineering. As a deep-tech R&D group, its breakthroughs will have the potential to drive innovation in health-care technology and manufacturing, diagnostics, wearable ultrasonic imaging, metamaterials, diagnostics, and AI-powered health analytics. WITEC’s work is also expected to accelerate growth in high-value jobs across research, engineering, clinical validation, and health-care services, and attract strategic investments that foster biomedical innovation and industry partnerships in Singapore, the United States, and beyond.

“Chronic diseases present significant challenges for patients, families, and health-care systems, and with aging populations such as Singapore, those challenges will only grow without new solutions. Our research into a wearable ultrasound imaging system aims to transform daily care for those living with cardiovascular and other chronic conditions — providing clinicians with richer, continuous insights to guide treatment, while giving patients greater confidence and control over their own health. WITEC’s pioneering work marks an important step toward shifting care from episodic, hospital-based interventions to more proactive, everyday management in the community,” says Sung, who serves as co‑lead PI of WITEC.

Led by Violet Hoon, senior consultant at TTSH, clinical trials are expected to commence this year to validate long-term heart monitoring in the management of chronic cardiovascular disease. Over the next three years, WITEC aims to develop a fully integrated platform capable of 48-hour intermittent imaging through innovations in bioadhesive couplants, nanostructured metamaterials, and ultrasonic transducers.

As MIT’s research enterprise in Singapore, SMART is committed to advancing breakthrough technologies that address pressing global challenges. WITEC adds to SMART’s existing research endeavors that foster a rich exchange of ideas through collaboration with leading researchers and academics from the United States, Singapore, and around the world in key areas such as antimicrobial resistance, cell therapy development, precision agriculture, AI, and 3D-sensing technologies.

To what extent can an artificial system be rational?

A new MIT course, 6.S044/24.S00 (AI and Rationality), doesn’t seek to answer this question. Instead, it challenges students to explore this and other philosophical problems through the lens of AI research. For the next generation of scholars, concepts of rationality and agency could prove integral in AI decision-making, especially when influenced by how humans understand their own cognitive limits and their constrained, subjective views of what is or isn’t rational.

This inquiry is rooted in a deep relationship between computer science and philosophy, which have long collaborated in formalizing what it is to form rational beliefs, learn from experience, and make rational decisions in pursuit of one’s goals.

“You’d imagine computer science and philosophy are pretty far apart, but they’ve always intersected. The technical parts of philosophy really overlap with AI, especially early AI,” says course instructor Leslie Kaelbling, the Panasonic Professor of Computer Science and Engineering at MIT, calling to mind Alan Turing, who was both a computer scientist and a philosopher. Kaelbling herself holds an undergraduate degree in philosophy from Stanford University, noting that computer science wasn’t available as a major at the time.

Brian Hedden, a professor in the Department of Linguistics and Philosophy, holding an MIT Schwarzman College of Computing shared position with the Department of Electrical Engineering and Computer Science (EECS), who teaches the class with Kaelbling, notes that the two disciplines are more aligned than people might imagine, adding that the “differences are in emphasis and perspective.”

Tools for further theoretical thinking

Offered for the first time in fall 2025, Kaelbling and Hedden created AI and Rationality as part of the Common Ground for Computing Education, a cross-cutting initiative of the MIT Schwarzman College of Computing that brings multiple departments together to develop and teach new courses and launch new programs that blend computing with other disciplines.

With over two dozen students registered, AI and Rationality is one of two Common Ground classes with a foundation in philosophy, the other being 6.C40/24.C40 (Ethics of Computing).

While Ethics of Computing explores concerns about the societal impacts of rapidly advancing technology, AI and Rationality examines the disputed definition of rationality by considering several components: the nature of rational agency, the concept of a fully autonomous and intelligent agent, and the ascription of beliefs and desires onto these systems.

Because AI is extremely broad in its implementation and each use case raises different issues, Kaelbling and Hedden brainstormed topics that could provide fruitful discussion and engagement between the two perspectives of computer science and philosophy.

“It’s important when I work with students studying machine learning or robotics that they step back a bit and examine the assumptions they’re making,” Kaelbling says. “Thinking about things from a philosophical perspective helps people back up and understand better how to situate their work in actual context.”

Both instructors stress that this isn’t a course that provides concrete answers to questions on what it means to engineer a rational agent.

Hedden says, “I see the course as building their foundations. We’re not giving them a body of doctrine to learn and memorize and then apply. We’re equipping them with tools to think about things in a critical way as they go out into their chosen careers, whether they’re in research or industry or government.”

The rapid progress of AI also presents a new set of challenges in academia. Predicting what students may need to know five years from now is something Kaelbling sees as an impossible task. “What we need to do is give them the tools at a higher level — the habits of mind, the ways of thinking — that will help them approach the stuff that we really can’t anticipate right now,” she says.

Blending disciplines and questioning assumptions

So far, the class has drawn students from a wide range of disciplines — from those firmly grounded in computing to others interested in exploring how AI intersects with their own fields of study.

Throughout the semester’s reading and discussions, students grappled with different definitions of rationality and how they pushed back against assumptions in their fields.

On what surprised her about the course, Amanda Paredes Rioboo, a senior in EECS, says, “We’re kind of taught that math and logic are this golden standard or truth. This class showed us a variety of examples that humans act inconsistently with these mathematical and logical frameworks. We opened up this whole can of worms as to whether, is it humans that are irrational? Is it the machine learning systems that we designed that are irrational? Is it math and logic itself?”

Junior Okoroafor, a PhD student in the Department of Brain and Cognitive Sciences, was appreciative of the class’s challenges and the ways in which the definition of a rational agent could change depending on the discipline. “Representing what each field means by rationality in a formal framework, makes it clear exactly which assumptions are to be shared, and which were different, across fields.”

The co-teaching, collaborative structure of the course, as with all Common Ground endeavors, gave students and the instructors opportunities to hear different perspectives in real-time.

For Paredes Rioboo, this is her third Common Ground course. She says, “I really like the interdisciplinary aspect. They’ve always felt like a nice mix of theoretical and applied from the fact that they need to cut across fields.”

According to Okoroafor, Kaelbling and Hedden demonstrated an obvious synergy between fields, saying that it felt as if they were engaging and learning along with the class. How computer science and philosophy can be used to inform each other allowed him to understand their commonality and invaluable perspectives on intersecting issues.

He adds, “philosophy also has a way of surprising you.”

Gradient Labs uses GPT-4.1 and GPT-5.4 mini and nano to power AI agents that automate banking support workflows with low latency and high reliability.

OpenAI raises $122 billion in new funding to expand frontier AI globally, invest in next-generation compute, and meet growing demand for ChatGPT, Codex, and enterprise AI.

AI for Disaster Response in Asia: OpenAI Workshop with Gates Foundation

Learn how STADLER uses ChatGPT to transform knowledge work, saving time and accelerating productivity across 650 employees.