A shot of carbon dioxide rewires how cement sets

One September day, it started to snow inside MIT’s Pierce Laboratory. 

Researchers depressurized a tank of liquid carbon dioxide (CO₂), instantly freezing it and releasing solid flakes. These were blended into cement paste and pressed into discs roughly the size of a dime, each sealed with a thin layer of vegetable oil to keep water in and air out. The team trained lasers on each, observing for the first time the transient chemical reaction that might explain why CO₂-injected cement paste gains its strength faster.

Injecting CO₂ into cement products like concrete is one way to store it and keep it out of the atmosphere. The process has attracted commercial interest, with a growing number of companies offering CO₂-injected concrete mixes. But until now, the underlying cement chemistry hadn't been directly visualized.

A new open-access paper in the Journal of the American Ceramic Society — led by Associate Professor Admir Masic and first-authored by graduate student Marcin Hajduczek, both of the MIT Concrete Sustainability Hub and MIT Department of Civil and Environmental Engineering — describes the chemical sequence that unfolds after CO₂ meets fresh cement paste. Co-authors include MIT colleagues Santiago El Awad and Franz-Josef Ulm, alongside researchers from IIT Jodhpur and CarbonCure Technologies.

Previous studies had pieced together a story about CO₂ injection’s chemical impacts from theory and indirect evidence; the key reactions simply moved too fast, and vanished too completely, for conventional techniques to catch them in the act. Raman confocal microscopy could — and it works on a simple principle: Illuminate a molecule with a laser, and the scattered light will reveal its identity. The light interacts with each material’s unique chemical bonds, shifting in energy to produce a distinct spectral “fingerprint.” Even the most fleeting and amorphous phases leave a readable trace.

“We’ve used Raman spectroscopy to better understand some of the most interesting materials in history, from the Dead Sea Scrolls to Ancient Roman concrete,” says Masic. “Cement paste may seem less glamorous in comparison, but pointing a laser at CO₂-injected cement paste as it hardens allows us to visualize things that haven’t been seen before.”

What they saw, unfolding during 24 hours of continuous scanning, was a three-act chemical drama.

Act One: Capturing calcium

The moment that CO₂ is added to the fresh cement paste, it goes to work. It dissolves into the pore solution and reacts with calcium released by the dissolving clinker, precipitating as various forms of calcium carbonate. Clinker is produced by heating limestone and aluminosilicate materials in a kiln, forming the primary ingredient ground into a fine powder to make cement. This happens within the first hour, temporarily slowing the normal hydration reaction, which requires calcium to proceed. 

In contrast, when CO₂ is not present, the calcium released by the dissolving clinker remains available locally, supporting the gradual formation of the material’s binding phases as it sets.

Left without calcium, the silicates released by the clinker dissolve into the pore solution and precipitate far from their source, linking together into chains that form an interconnected silica gel network throughout the paste. This amorphous, fleeting gel sets the stage for what follows.

Act Two: The ghostly gel

Once the injected CO₂ is fully mineralized — around four to five hours after mixing — normal hydration resumes. Calcium hydroxide begins to precipitate into the pore space, and when it does, it encounters the silica gel network waiting for it.

The reaction between the two phases begins immediately, producing calcium silicate hydrate (C-S-H), the compound that gives cement its binding ability. What makes this form of C-S-H distinct is where and how it forms: not clustered around clinker particles as in conventional hydration, but distributed throughout the entire matrix, wherever the silica gel had spread.

The CO₂ had temporarily suppressed the paste’s alkalinity, and that lower pH was the only thing keeping the silica-gel intact. As hydration reasserts itself and produces standard hydration products, namely C-S-H and calcium hydroxide, the latter drives pH back up to typical levels in a self-reinforcing loop; the silica-gel reacts with calcium hydroxide through a so-called pozzolanic reaction. Within eight hours, the silica gel is almost entirely gone — the previously well-distributed gel network turns rapidly into additional C-S-H during this critical early window. 

“At first, the fleeting nature of the silica gel looked like a fluke in the Raman data. But it quickly became clear that its sudden disappearance was a consistent, undeniable feature of every CO₂-injected sample,” says Hajduczek.

Act Three: A rewired matrix

With the silica gel consumed, the paste settles into conventional hydration, but what it leaves behind is measurably different. Because the new binder was distributed more evenly throughout the cement matrix, the resulting microstructure is stronger and more uniform at an early age. In the study, paste mixed with CO₂ at 1 percent by cement weight achieved, on average, 13 percent higher compressive strength at 24 hours, compared to reference mixes.

“We’ve been injecting CO₂ into cement products for years without fully understanding what it was doing inside. Now that we can see it and understand the underlying mechanism that leads to improved performance, we can start to control it. And there’s a lot of room to push,” says Masic.

The findings also refine a leading explanation for CO₂-injected cement paste’s higher early age strength: the calcium carbonate crystals, previously suspected to seed C-S-H growth, turn out to be passive bystanders embedded in the silica gel template rather than reacting to form C-S-H. 

Where the chemistry goes next

Knowing the mechanism gives researchers a more specific set of questions to pursue. The silica gel template explains the distribution of the new C-S-H, but directly measuring its mechanical properties remains a next step.

On the practical side, dosage matters: Flood the system with too much CO₂ and calcium gets locked into carbonate before the gel can form and react. If the paste used here forms abundant C-S-H, it could theoretically offset up to 40 percent of the carbon emissions from cement production, excluding emissions associated with the fossil fuels used in the process. In practice, however, the achievable offset is likely to be only a fraction of that value, although still potentially significant.

But even with these open questions, the ghostly gel has been caught. And now that researchers know what to look for, the chemistry that unfolds in those first eight hours is no longer invisible.

de MIT News https://ift.tt/aR52rAK

New imaging system sees through murky waters

For remotely operated underwater vehicles, cloudy and turbulent waters are often a no-go. When vehicles settle on the seafloor or dig through a sandbed, they can kick up clouds of sediment that make it tough for onboard cameras to see through. Often, the only thing to do is to wait until the marine dust settles before a vehicle can safely proceed. 

But a new underwater mapping technique developed by engineers at MIT and the Woods Hole Oceanographic Institution (WHOI) may allow vehicles to see through murky, low-visibility waters. 

The method fuses visual images from optical cameras with acoustic data from sonar sensors. The combination enables a vehicle to quickly map the general shape of its surroundings using sonar, even in low-visibility waters. A vehicle can move toward certain shapes in the sonar-mapped environment, coming close enough for optical cameras to visually resolve specific objects in detail. 

The technique is akin to pairing a dolphin’s echolocation with a sea turtle’s close-range vision to see and navigate through murky water, in real-time. 

The researchers tested the method in tank experiments where they could control the water’s degree of visibility. Even in the cloudiest conditions, the system was able to see through the sediment to map the tank’s environment and visualize centimeter-scale details of objects in the tank. 

The team is further improving the technique, which they’ve named Sonar-MASt3R. They envision that the mapping method could safely guide underwater vehicles through murky environments for a range of applications, including scientific exploration, underwater construction and maintenance, and deep-sea recovery. 

“We hope that this work enables us to do more operations in those challenging, low-visibility environments, and helps provide more coverage in areas that are difficult to operate in today,” says Amy Phung, a graduate student in MIT’s Department of Aeronautics and Astronautics, who led the work. 

Phung presented a paper detailing Sonar-MASt3R this week at the IEEE International Conference on Robotics and Automation (ICRA). The paper’s co-author is Richard Camilli, senior scientist of applied ocean physics and engineering at WHOI. 

The best of both

To see underwater, scientists have generally taken an either/or approach, using either optical cameras or sonar sensors to guide the way. Optical cameras can provide detailed visual imagery of a scene, but only in waters that are relatively clear and well-lit. In contrast, sonar sensors perform just as well in clear and murky water; by emitting acoustic waves and measuring the time and angle at which they return, sonar sensors can determine the exact shape, distance, and depth of objects in the environment, though a sonar map lacks any visual detail. 

To get the best of both modes, scientists have looked to combine the two in a new approach known as “opti-acoustic fusion.” In a handful of prior works, research groups have merged sonar and optical data in mapping techniques that are mostly geared toward object recognition and reconstructing workplace environments. Most techniques require time to sync and process the data and therefore do not work in real-time, while only a few can map an environment in 3D. None have been applied to high-resolution mapping underwater in murky, turbid conditions. 

Phung, who is a student in the MIT-WHOI Joint Program, and Camilli, her advisor, aimed to develop an opti-acoustic fusion technique that would generate detailed 3D maps of underwater environments in real time and in low-visibility conditions. The team was motivated, in part, by challenges in safely recovering unexploded underwater mines.

“There can be old explosives in areas that make it unsafe for ships to be in, and the ability to get rid of those safely is best done by robotics,” Camilli says. “But a lot of these explosives are set in surf zone environments where visibility adds to the challenge of doing this safely. That’s one of many applications that our technique can be used for.”

Cloudy, with a chance of mapping

The new method, Sonar-MASt3R, builds on an existing technique, MASt3R, that was developed by researchers in France. MASt3R is an image matching algorithm that is trained to take in visual images of the same scene and quickly estimate the relative depth of each pixel in the scene. In this way, MASt3R can generate a 3D map of the environment in real-time, based on a camera’s 2D images. 

“The downside is that there is no sense of scale,” Phung says. “It will say ‘this pixel is five units closer than this pixel,’ but it can’t say whether that’s 5 meters or 5 feet.”

Luckily, sonar provides absolute measurements of scale. The timing of sonar reflections can be translated directly into a specific depth and distance of objects that the signals bounced off, as well as their shape and contour. 

In their new work, Phung and Camilli used sonar data to correct MASt3R’s scaling and generate precise 3D maps of underwater environments. Even in murky water, the method’s sonar-corrected map would enable a vehicle to know the precise location of objects, and therefore how far to safely move in for a closer inspection, which the vehicle could then do using conventional optical cameras.

The team tested Sonar-MASt3R in experiments with a tank that they filled with water, sediment, and a variety of objects such as a small boulder, a coffee mug, and a packing crate. Inside the tank, they also set up a robotic arm, onto which they mounted an underwater camera, and a sonar sensor. 

For each experimental run, they first carried out a sweep trajectory, in which the robotic arm slowly swept from one side of the tank to the other to capture sonar and visual data. With this first sweep, Sonar-MASt3R quickly creates a coarse sonar-based map of the shapes and contours of the tank and its objects. The coarse map is then used to record close-up camera images of the objects, which are used to improve the map resolution. A “keyframe” approach quickly compares each new image frame to the last keyframe. If a frame provides new information not contained in the last keyframe, the image is added as a new keyframe to the map. If it is similar, it is immediately discarded. In this way, the approach can quickly fill in the map with relevant visual detail, in real-time. 

The researchers tested their new approach underwater, testing eight different levels of turbidity, which they created by stirring up the tank’s sediment. Compared with other opti-acoustic fusion approaches, Sonar-MASt3R generated more accurate 3D maps and resolved smaller, centimeter-scale details, and in cloudier conditions. In the cloudiest condition, which the robotic arm’s cameras could not see through, its sonar sensors were able to generate a rough map of the tank’s hidden objects. This initial map enabled the arm to move safely through the murk and closer to specific objects, which its underwater camera could then visualize in more detail. 

“An analogy would be if you were to go into a china shop in the dark, and try to pick your way around to find a specific coffee mug without knocking things over,” Camilli offers. “This would allow you to do that.”

The team plans to test the approach in natural underwater conditions, where they suspect that the mapping task should be more straightforward. 

“In a tank, it’s like an echo chamber,” Camilli says. “It’s like trying to do this in a funhouse mirror setting where you get all these distortions and reverberations and ghost images that really complicates the processing. If you put it in the real world, it should be easier.”

Then, they say, Sonar-MASt3R could help scientists safely explore in cloudy, turbid, and murky underwater regions.

“The real value in this effort is so we can use this technology in mission scenarios that are untractable right now,” Phung says. “And there are plenty of untractable missions because we don’t have the observational or perception capabilities.”

This research was supported, in part, by NASA, and the National Science Foundation.

de MIT News https://ift.tt/YwPuAvR

To study how chips really work, MIT researchers built their own operating system

A new kernel, or core program within an operating system, gives researchers a cleaner view of what’s happening inside a processor. Called Fractal and developed at MIT, the kernel has already surfaced previously unknown behavior in Apple’s M1.

When security researchers want to understand what a modern processor is really doing with the kind of detail that determines whether attacks like Spectre and Meltdown are possible, they usually run their experiments on top of an operating system that was never built for the job. They open up macOS or Linux, patch the kernel by hand, and hope the modifications hold. The approach is unstable, hard to reproduce, and on Apple’s platforms, slated for deprecation.

A team at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) decided to build something different. Fractal, an operating system kernel written from the ground up, treats the hardware itself as the object of study. Its first major use, a deep look at branch predictors — a CPU’s way of guessing what code to run next, before it knows for certain, so it doesn’t have to waste time waiting to find out — inside Apple’s M1 processor, has already turned up findings that prior work missed, including the first evidence that a class of speculative attack known as “Phantom” affects Apple Silicon.

“We’re using hardware in ways it wasn’t designed for,” says Joseph Ravichandran, the MIT PhD student in electrical engineering and computer science (EECS) who led the project. “It’s not even obvious that this is a possible thing you could do with the hardware. But we found a way to pull all these different primitives off. It’s like a microscope. If you’ve got a hand magnifying glass, you can see a little bit. But if you had an electron microscope, now we’re really talking. That’s what Fractal is. The electron microscope of operating systems.”

A clean room for chip research

The core problem Fractal solves is one that researchers have worked around for years. Modern processors keep state in many internal structures: branch predictors, caches, translation lookaside buffers, and more. To study how those structures behave across the boundary between user code and kernel code, two domains the chip is supposed to keep isolated, researchers need to run nearly identical experiments on each side of that boundary. On a general-purpose operating system, that is very difficult. The system itself manages privilege levels, address spaces, and scheduling, and it injects its own activity into every measurement.

Fractal inverts the model. It boots directly on bare metal, with no other software running, and exposes primitives that let a single experiment switch privilege levels at runtime while executing the same instructions in the same address space. The team calls the underlying technique multi-privilege concurrency, and it relies on a new construct they introduced: the outer kernel thread, which sits inside a user process’s memory but executes with kernel privileges.

The result is an experimental setup with almost no background noise. Where measurements taken under macOS or Linux are blurred by interrupts, scheduler activity, and address-space management, Fractal produces flat baselines and clean signals.

What Fractal found on the M1

Apple’s M1 implements an ARM specification called CSV2, which is supposed to prevent code running in one privilege level from steering speculation in another. Using Fractal, the MIT team confirmed that the protection works for the execute stage of indirect branch prediction: a user-mode program cannot make the kernel speculatively execute a chosen target through the indirect branch predictor.

But the team also found something the chip’s designers may not have intended. The CPU still fetches the target into the instruction cache before the protection kicks in. That fetch is observable through a side channel, which means user code can still influence what the kernel pulls into its caches across the privilege boundary. The same pattern appeared between processes assigned different address space identifiers.

The team also produced the first evidence that Apple Silicon exhibits Phantom speculation, a class of misprediction previously demonstrated only on AMD and Intel processors. In Phantom, ordinary instructions, including a no-op, can be misinterpreted by the CPU as branches, triggering speculative behavior the program never asked for. On the M1, Fractal showed that Phantom fetches succeed across both privilege levels and address spaces, though the execute phase remains blocked.

A separate Fractal experiment overturned a finding from earlier work on the M1’s conditional branch predictor, which had reported that cross-privilege training worked on Apple’s performance cores, but not its efficiency cores. The Fractal team showed that the conditional branch predictor has no privilege isolation at all, on either core type, and that the earlier result was likely an artifact of macOS quietly migrating threads between cores during system calls.

“For us, it is a true independent variable,” Ravichandran says. “You change the privilege level, nothing else changes. The only thing that could explain whether the attack succeeds or not is the privilege level.”

A tool, not a one-off

Fractal supports x86_64, ARM64, and RISC-V, and consists of more than 31,000 lines of code. The team designed it as infrastructure rather than as a single experiment, with familiar POSIX system calls, a C library, and ports of standard tools like vim, GCC, and the dash shell, so that researchers can move existing experiment code over with minimal friction.

The MIT team disclosed its M1 findings to Apple’s product security team. In an unusual reversal, Apple’s engineers also examined Fractal. 

The longer-term ambition is bigger than any single result. Ravichandran wants Fractal to become to microarchitecture research what tools like QEMU and FFmpeg are to their fields: shared infrastructure that the whole community builds on. 

“My hope is that our results as a community get significantly more reliable, significantly more accurate,” says Ravichadran. “With this reduced noise, this clarity, and this guarantee that you’re running on the right core, on the right system.”

“Fractal is a strong architecture contribution because it turns an often ad hoc microarchitectural reverse-engineering workflow into reusable research infrastructure,” says University of Southern California assistant professor Mengyuan Li, who wasn’t involved in the paper. “By reducing software noise and giving researchers tighter control across privilege boundaries, it makes difficult hardware experiments much easier to interpret.”

Ravichandran worked with Mengjia Yan, an MIT associate professor of EECS and CSAIL principal investigator, on the paper. Their work was supported, in part, by the National Science Foundation, the U.S. Air Force Office of Scientific Research, and ACE, which is part of a program sponsored by the U.S. Defense Advanced Research Projects Agency. They presented their work at the IEEE Symposium on Security and Privacy in San Francisco, California.

de MIT News https://ift.tt/9vBqKk7

Pablo Jarillo-Herrero wins Kavli Prize in Nanoscience

MIT professor of physics Pablo Jarillo-Herrero is among 10 researchers worldwide to receive this year’s prestigious Kavli Prize. 

Jarillo-Herrero is co-recipient of the 2026 Kavli Prize in Nanoscience “for foundational work that established the field of twistronics.” His co-recipients are professors Eva Y. Andrei at Rutgers University and Allan MacDonald from the University of Texas at Austin.

These three physicists are being honored for the theoretical foundation and experimental validation of a new field of “twistronics,” where superconductivity, magnetism, and other properties can be obtained by rotating two-dimensional materials such as graphene to a “magic angle.”

A partnership among the Norwegian Academy of Science and Letters, the Norwegian Ministry of Education and Research, and the Kavli Foundation, the Kavli Prizes are awarded every two years to “honor scientists for breakthroughs in astrophysics, nanoscience and neuroscience that transform our understanding of the big, the small and the complex.” The laureates in each field will share $1 million.

“Pablo’s groundbreaking research has once again been given well-deserved recognition,” says Nergis Mavalvala, dean of the MIT School of Science and the Curtis and Kathleen Marble Professor of Astrophysics. “Pablo and his co-recipients have pioneered twistronics, very fundamental scientific research that has opened up a new field with myriad possibilities for novel quantum materials.”

In 2009, using scanning tunneling microscopy and spectroscopy on graphene, most commonly found as a single layer of carbon atoms arranged in hexagons resembling a honeycomb structure, Andrei and her research group demonstrated that small variations in twist angle profoundly modified the electronic structure. This demonstration — that geometric control, rather than chemical composition, could modify a material’s electronic structure — represented a fundamental advance in materials design and arguably launched the field now known as “twistronics.”

In 2011, MacDonald quantitatively explained the emergence of this electronic structure by geometries at discrete magic angles. This framework has since become the theoretical foundation of what are known as moiré materials, and has guided subsequent experimental and theoretical developments across a wide range of twisted and layered systems. 

In 2018, Jarillo-Herrero’s group observed correlated insulating phases and superconductivity in magic-angle twisted bilayer graphene devices. The resulting platform, “combining atomic-scale structural simplicity with electronic tunability, has enabled systematic investigations has had broad and lasting impact across nanoscience and quantum material research,” according to the Kavli Prize citation.

“It was a big surprise, because the technique we used, though conceptually straightforward, was hard to pull off in the lab,” said Jarillo-Herrero recently. He is also the Cecil and Ida Green Professor of Physics at MIT and a member of the Research Laboratory of Electronics. 

“I’m humbled and incredibly honored to be sharing this award with [Andrei and MacDonald],” Jarillo-Herrero noted in an essay describing his journey to the Kavli Prize. “I want to also emphasize that this award honors fundamental physics research in nanoscience. It is incredibly important for society to continue to support fundamental research: Although it often doesn’t have a direct near-term application, in the long run it happens to be the most transformative and impactful in society.”

“Pablo’s research has helped spark a revolution in condensed matter physics and nanoscience, inspiring physicists worldwide to explore superconductivity and other emergent phenomena in engineered quantum materials. This work could potentially lead to the creation of superconductors at room temperature, which would would have an enormous technological impact,” says Deepto Chakrabarty, physics department head and William A. M. Burden Professor in Astrophysics.

Jarillo-Herrero's win brings the number of all-time MIT faculty recipients of the Kavli Prize to nine. Prior winners include Nancy Kanwisher in neuroscience (2024), Bob Langer in nanoscience (2024), Sara Seager in astrophysics (2024), Rainer Weiss in astrophysics (2016), Alan Guth in astrophysics (2014), Mildred Dresselhaus in nanoscience (2012), Ann Graybiel in neuroscience (2012), and Jane Luu in astrophysics (2012).

de MIT News https://ift.tt/lkzDWJT

Augmented reality system could make medical ultrasounds easier to interpret

Interpreting medical ultrasound images is a difficult task, requiring a technician to look at 2D images and mentally arrange them into a 3D representation of what the tissue looks like. 

To make that job easier, MIT researchers developed a new approach to ultrasound imaging that allows the user to visualize a 3D augmented-reality image of the object being scanned. Using a virtual-reality headset, they can see a precise 3D digital representation of what the object actually looks like, making it easier to identify and analyze.

This technique could help speed up the training process for ultrasound technicians and other health care providers who use ultrasound. It could also be deployed for use in hospitals, for tasks such as using ultrasound to place a needle in the right location for a biopsy.

“For training, this could make ultrasound more intuitive and more understandable. On the clinical side, it could be less time-consuming, more accurate, and also give health care providers more peace of mind. They wouldn’t have to wonder if they missed anything,” says Canan Dagdeviren, an associate professor of media arts and sciences at MIT and the senior author of the study.

MIT graduate students Jason Hou and Shrihari Viswanath are the lead authors of the paper, which appears today in Nature Communications Engineering. Other authors of the paper include Bowen Wu ’24 and two MIT Summer Research Program students, Cinay Dilibal, a senior at Dartmouth College, and Tanisha Shende, a senior at Oberlin College.

3D representations

Ultrasound imaging works by bouncing high-frequency sound waves off tissues in the body, which are then reflected back to an ultrasound transducer. The transducer converts these sound waves to electrical signals, which are used to create a 2D image of the tissue. Ultrasound technicians are trained to convert these images into a 3D mental representation of the tissue.

“It's a difficult skill to master, and there are long learning curves,” says Hou. “The hardest thing is this mental tomography bottleneck where you’re trained to reconstruct the 2D slices in your 3D mental space. That is a cognitive burden that can lead to inaccuracies in scanning.”

To reduce that cognitive load, the MIT team thought it could be helpful to combine two technologies: 3D ultrasound imaging and augmented reality (AR). 

Three-dimensional ultrasound imaging is occasionally used in fields such as fetal imaging and echocardiography, which is used to image the heart, but most 3D ultrasound imaging systems are expensive and not widely available. For this study, the MIT team used a real-time 3D system they developed recently for use in breast-cancer detection.

Their new system includes an ultrasound probe, slightly smaller than a deck of cards, that transmits information using a chirped data acquisition system (cDAQ). The probe contains an ultrasound array arranged in the shape of an empty square, a configuration that allows the array to take 3D images of the tissue below.

Because this system has fewer ultrasound elements than a typical 3D ultrasound system, it requires less power and is less expensive to build.

The data collected by the ultrasound probe can then be compressed and streamed into a 3D computer graphics engine called Unreal Engine, which converts the voxel data from the ultrasound image into a direct 3D representation of the object, with no loss of information. Wearing an AR/VR headset, the user can see this 3D rendering representing the internal structure, superimposed over the object’s actual location — like X-ray vision. By tilting their head or approaching from a different direction, the user can see different views of the object, making it easier to identify.

Easier to use

The researchers tested their new technology, which they call AR-VIU (augmented real-time volumetric imaging in ultrasound), with a group of 18 participants. Nine of the subjects were experts in ultrasound technology (including sonographers and physicians), and nine had never used ultrasound before.

Each user performed identification tasks using four different ultrasound technologies. In one condition, they viewed 2D images on a regular screen, which is the way that most ultrasounds are now performed. They also viewed 3D images on a regular screen, as well as two augmented reality conditions: one 2D and one 3D (AR-VIU).

In one round of experiments, users were asked to identify an object embedded in gelatin — such as a spring, a ball, or a screw — inside an opaque container that was scanned with ultrasound. In a second set, they were asked to use a pen to mark the location of “tissue phantom” — a gel-like material engineered to mimic human tissue. This simulates the task of locating the right spot for a needle during a biopsy.

The researchers found that the AR-VIU system significantly improved all users’ ability to identify and locate objects. The effect was especially strong for novices, who performed nearly as well as experts when using AR-VIU. When using the traditional 2D imaging system, experts performed much better than novices.

“Overlaying images with the anatomy and providing 3D visual context makes ultrasound significantly easier for novices to understand,” Viswanath says.

In interviews after the experiments, most of the novices reported that they preferred the AR-VIU approach, with many saying that it made the tasks easier.

“The 3D system imposes less brain drain, it’s more intuitive, and it’s easier to understand what is happening in the targeted region,” Dagdeviren says.

Many of the experts said they preferred the traditional 2D imaging because that is what they were accustomed to and had been trained to use. However, those experts also said they could see the benefits of the AR-VIU system in some situations, such as placing a needle for a biopsy or visualizing the movement of the heart wall during echocardiography.

The researchers are now working on further improving the resolution of the imaging and doing additional tests to demonstrate the accuracy of the AR-VIU technology.

The research was funded by the MIT Media Lab Consortium, the National Science Foundation, an MIT HEALS graduate fellowship, and an MIT-Tata graduate fellowship.

de MIT News https://ift.tt/wnOElxF

Startup’s nuclear-inspired cooling system could make data centers more sustainable

The rise of artificial intelligence is riding on the back of an enormous data center expansion. Data centers are projected to account for anywhere from 9 to 17 percent of total electricity usage in the U.S. by the end of the decade. Today, around a third of data center electricity is devoted to cooling the chips that run AI models.

That’s the process Ferveret is working to make more efficient. The startup, founded by Reza Azizian, a former MIT postdoc in nuclear engineering, and Matteo Bucci, MIT’s Esther and Harold E. Edgerton Associate Professor in the Department of Nuclear Science and Engineering, is adapting an approach from nuclear reactors to cool chips using no water and significantly less electricity.

The company’s cooling system submerges computer servers in a specialized liquid that absorbs heat much more efficiently than air from a fan. What makes the solution different from other liquid cooling systems are the bubbles: Ferveret’s Adaptive Phase Cooling (APC) solution produces much smaller bubbles at the surface of the server, which detach more frequently, accelerating the heat transfer process.

Ferveret is already testing its solutions with companies including CleanSpark, the data center developer and operator, as well as FuriosaAI, an AI accelerator company, and Switch, one of the largest data center operators in the U.S.

In a recent study in collaboration with the Samueli Computer Science Department at the University of California at Los Angeles, Ferveret found its APC solution led to a 15 percent improvement in computational power efficiency compared to state-of-the-art liquid cooling solutions. By combining those savings with Ferveret’s power control system to optimize operating conditions, the company says it allows data centers to get 35 percent more tokens — small pieces of text or data — from their AI models with the same amount of power.

“Our goal is to make data centers as sustainable as possible and help them use every single watt of power to generate tokens, which are the most useful outputs,” Azizian says. “Our system enables the operation of more powerful chips, it helps data centers waste a lot less energy, and it accomplishes all that with zero water consumption.”

From nuclear reactors to AI

Azizian was a postdoc at MIT in 2013 when he met Bucci, who was then a research scientist. They worked on heat transfer in nuclear reactors before Azizian went into industry, where he shifted his focus to cooling chips. Azizian first worked on Microsoft’s HoloLens augmented reality headset and then joined Nvidia, which produces the graphical processing units companies use to train and run the latest AI models. Meanwhile, Bucci continued conducting research at MIT, becoming an assistant professor in 2016.

Azizian walked into his first data center in 2017, where he was struck by the massive, noisy fans that filled the building as they cooled.

“I thought, ‘Holy crap, this is not how you cool facilities,’” Azizian recalls, noting air cooling can still take up 40 percent of the power going into a data center. “It was not an efficient way of doing things, but since it wasn’t hurting the performance, no one cared that the cooling technology was 50 years old.”

Azizian began talking with Bucci about applying their knowledge around optimizing heat transfer in nuclear reactors to data centers. Scientists have spent decades finding better ways to move heat in nuclear reactors.

“Heat transfer determines how much energy you can extract from the reactor core, which translates directly to revenue,” Azizian explains.

The founders started Ferveret in 2021. A lot has changed since Azizian walked into his first data center. Chip companies have packed more and more components onto their chips as the explosion in artificial intelligence has put a premium on squeezing as much computing capacity as possible out of limited power supplies.

That has driven data center operators to use liquid to cool chips — often through a technique known as immersion cooling that submerges chips in liquid. The most effective form of immersion cooling brings the liquid to a boil.

“Liquid is a better heat transfer medium than air. That’s why when you stick your hand into room temperature water it still feels cold,” Bucci explains. “When liquid is boiling, it becomes even better at removing heat because the phase change requires a lot of energy, which is the energy you remove from the chip. That lets you transfer large quantities of heat with minimal temperature differences between the chips and the liquid.”

Unfortunately, boiling liquid adds complexity to the system because it forces operators to capture and reliquefy the bubbles while controlling for pressure, temperature, and fluid inventory.

Ferveret’s system is adapted from a process in nuclear reactors called subcooled boiling. It uses a liquid with a low boiling point and none of the toxic PFAS “forever chemicals” that other approaches rely on. At the surface of the chip, Ferveret’s liquid produces smaller bubbles than other immersion cooling approaches. Those bubbles detach more frequently and quickly recondense in the surrounding liquid, accelerating the bubble-rewetting cycle at the surface of the chip to hasten heat transfer.

Ferveret delivers its APC system in small boxes, each of which houses one server. The founders say their modular systems make it easier to deploy the system and simplify maintenance.

“The physics enable us to get to form factors that weren’t possible in the past,” Azizian says. “Most immersion cooling solutions are large tanks that people submerge the servers in. We have a smaller, modular rack-mounted solution that makes it adaptable to the current infrastructure, so it’s easier for people to deploy our technology.”

Ferveret also offers control software that adjusts the power going to each server in real-time to further improve efficiency.

“We deliver full-stack systems that include the cooling box, the rack, the cooling distribution units, and sensors that measure the temperature and pressure,” Bucci says. “Our software monitors those sensors and optimizes the operating condition inside each box to ensure that energy consumption is minimized in the system.”

AI with fewer resources

In addition to helping data centers to run more efficiently, Ferveret is also improving sustainability by making it easier to operate data centers in remote regions with more renewable energy.

“The sun shines in places where you don’t have much water, so the advantage of us being water-free is we allow you to build data centers where you have solar energy but nothing to cool the data center down,” Bucci says. “This technology can help deploy data centers in regions where normally you wouldn’t have the resources to do so, including Africa, the Middle East, and of course parts of America. It’s a huge unlock.”

Ferveret is in talks with the large cloud computing companies known as hyperscalers, and is currently part of Nvidia’s Inception program for startups. The company plans to announce expanded partnerships later this year. From there, the founders plan to quickly scale their technology to help the AI industry continue to grow without further straining the planet.

“The computing industry is facing a huge challenge in the form of access to power, and they have a problem with access to water in many regions,” Azizian says. “That will only become more limiting as the industry grows. The main goal for these data center operators would be to get more tokens from the power they have. We’ve shown we can do that.”

de MIT News https://ift.tt/KDPLFmN

The consequences of relying on AI for accurate news

It’s no secret that the last few years have seen a massive explosion in the use of artificial intelligence for general information-gathering. An even more recent trend, though, is how large language models (LLMs) like ChatGPT, Claude, and Gemini are increasingly being used for verifying and consuming news; reports from the Pew Research Center over the last year found that one-in-five U.S. teens regularly use LLMs to get their news, while one-in-four young adults have reported using them for that purpose at least once. 

A new open-access study from the MIT Media Lab should give some of those users pause: Researchers found that, over the course of a month, participants who relied on AI systems to verify facts actually got worse at detecting misinformation on their own when their chatbots were taken away.

This phenomenon, which is often referred to as the “AI dependency paradox,” has been observed in a wide range of knowledge domains, like the 2025 study that found that doctors who used AI got worse at detecting cancer on their own. The dynamic mirrors broader tech trends around so-called “deskilling” (or “cognitive offloading”) that have been well-documented for decades, from calculators weakening our math skills to Global Positioning System (GPS) technologies impacting our natural sense of direction.

In the new Media Lab study, which tracked 67 people over four weeks as they evaluated news headline-image pairs, participants were 21 percent more accurate in detecting fake news when assisted by an AI chatbot during a session — confirming previous research out of the MIT Sloan School of Management demonstrating that AI can be an effective tool in reducing people’s beliefs in false information.

However, the study showed that a new wrinkle emerged when the AI was no longer present: By week four, participants’ unassisted performance on new news items declined by 15 percentage points compared to before the study started. (Roughly a quarter of all participants actually reported feeling that they were getting better at detection, even as their performance declined.)

Dunning-Kruger creeps in

“Users get excited about these ‘magical’ LLMs, but forget that they’re just statistical models that predict the next ‘token’ in a sequence [of letters/words],” says MIT media arts and sciences (MAS) PhD student Anku Rani, co-lead author of a new paper about the research, alongside fellow MAS PhD student Valdemar Danry. “Many impressive behaviors emerge from scaling this, but it comes with real limitations, both in what the model can reliably generate and in its broader impact on the people using it.”

Qualitative analysis identified distinct behavioral patterns, with the team labeling one-fifth of all participants as "Dependency Developers” who gradually shifted from active self-reliance to passive acceptance of AI guidance.

In the post-experiment survey, one respondent explicitly acknowledged this transition, noting their passive role in the process. “While [the chatbots] did emphasize that you must check across multiple sources to make sure a story is true, they didn’t teach me much about exploring the context of the images themselves,” the participant said.

The research team said that these AI models are particularly vulnerable to mistakes in the midst of emotionally charged breaking news, as exhibited by the widespread misinformation that accompanied President Trump’s recent assassination attempt and major events during the Iranian war. (The authors also point out that the original human-created news content that’s used to train the AI models is increasingly unreliable and/or biased, further exacerbating the problem.)

The paper, which Danry and Rani presented at the 2026 CHI Conference on Human Factors in Computing Systems, was co-authored by Assistant Professor Paul Pu Liang, Senior Research Scientist Andrew Lippman, and senior author Pattie Maes, the Germeshausen Professor of Media Arts and Sciences. 

The solution: Being a coach, not a crutch

The researchers say that the results of their project suggest that the specific way in which an AI interacts with a user determines whether its impact will be “as a coach, versus as a crutch.” The study found a clear distinction between conversational strategies that simply help in the moment and those that actually support active learning and skill development.

For the latter, the Media Lab team uncovered several strategies associated with stronger independent detection later on, even if the strategies initially slowed down performance during the interaction. This included the Socratic method of the AI asking guided questions, as well as so-called “deep probing,” where the system provides gently persuasive statements if the user appears to be veering away from the correct response.

“AIs that ‘tell’ by providing direct answers are more likely to foster reliance, while those that ‘ask’ via Socratic questioning are better at engaging someone to actually learn how to discern the truth on their own,” says Danry. “But it’s very much a trade-off between speed and effort.”

Rani noted a few key limitations to the one-month study, from the small dataset of roughly 50 validated news items to the demographic focus on the United States and the United Kingdom. In the future, she says that the team hopes to do similar experiments with more geographically diverse cohorts, including low-resource communities, and is also eager to explore whether other multi-modal interaction strategies — like interacting with culturally adaptive digital twins instead of text-based chatbots — help people improve their abilities to detect misinformation. 

At a higher level, the researchers hope that the project will be something that educators can examine as they develop teaching plans that incorporate AI tools into their school curricula.

“It’s especially important to raise awareness in our schools and academic communities about the shortcomings of using AI as learning tools,” says Maes. “People need to know that if they ‘delegate’ their thinking, they’re not going to get better at that particular brand of problem-solving. Ultimately, the ability to question and analyze information is important for everyone, because it empowers us to solve problems and form our own independent opinions about the world.”

Danry adds that the rapidly-evolving field of machine learning and deep learning will require continuous education on the benefits and drawbacks of LLMs.

“There’s a lot of work to do in making sure that we don’t just fully offload critical tasks that we want to be able to keep on doing to these models,” he says. “We need to develop a new kind of AI literacy.”

The research project was supported, in part, by the Media Lab Consortium, an MIT Tata Center Technology and Design Fellowship, and a Google PhD Fellowship in Human–Computer Interaction.

de MIT News https://ift.tt/WrMRSGb