Assil Halimi PhD ‘25 speaks to TechCrunch’s Tim De Chant about his company Apollo Atomics, which is shrinking the steam generator component of a reactor to streamline nuclear power. “We don’t want to be incremental in nuclear and just improve a small thing,” says Halimi. “Our target is to beat natural gas.”
MIT researchers have engineered three strains of bacteria to function as living transistors, carrying signals across circuits, writes Gizmodo’s Gayoung Lee. “For instance, the circuit could sit near the roots of plants to detect different stresses or autonomously respond to the presence of pests or other environmental threats,” writes Lee.
MIT researchers have found that lower birth rates are associated with higher growth in Gross Domestic Product globally, writes The Atlantic’s Idrees Kahloon. “[P]erhaps more important than the observed correlation is the authors’ proposed explanation: that when prime-age workers become scarce, companies adapt by developing productivity-boosting technology,” notes Kahloon.
In a STAT opinion piece, Arya Rao, a candidate in the Harvard/MIT MD-PhD program, and Marc Succi write about how AI impacts clinical reasoning. “[C]linical reasoning and model reasoning are not the same thing,” Rao and Succi explain. “AI not only lacks this hidden framework and data repository for learning, but it arrives at conclusions using a fundamentally different method.”
Prof. Robert Langer, cofounder of Moderna, speaks to The Boston Globe’s Marin Wolf and Jonathan Saltzman about the company’s successful results of a personalized messenger RNA vaccine in a late-stage trial to treat melanoma. “So far, the only proof is with melanoma, but it’s a stunning example of what can be done,” says Langer. “I don’t see any reason why it wouldn’t work with other types of cancer.”
Prof. Erin Kara speaks to NPR’s Ari Daniel about how the fastest star ever observed in the galaxy, located next to the Sagittarius A* black hole, experiences how space and time are distorted by the black hole. “This is a really exciting result," says Kara. "It kind of sets the stage in the next decade to make these kind of unprecedented measurements of black holes.”
MIT researchers have discovered that when individual images are removed from large AI diffusion models, the models are still able to produce like images, writes Digital Trends’ Varun Mirchandani. “The more interesting takeaway is that this connection becomes increasingly difficult to trace as datasets grow,” writes Mirchandani. “An AI-generated image may draw on patterns learned from an enormous pool of material without having a clear, identifiable source image behind it.”
Prof. Danielle Wood speaks to the Scientific American’s Tom Metcalfe about how the current models of space travel, that have created trash and debris in Earth’s atmosphere and orbit, are unsustainable. “Sustainability in space means that we are able to leave to future generations something like the natural state of the Earth, the moon and the solar system,” says Wood.
Boston.com’s Elizabeth Mehler spotlights an AI-powered flood predictor and disaster response tool built by Arush Shangari, an MIT Beaver Works Research Scholar. “Through MIT Beaver Works and MIT CSAIL, I taught myself artificial intelligence and time-series forecasting,” says Shangari. “There was a lot of that error analysis and just learning in general.”
Writing for WBUR, Prof. Christopher Knittel and Visiting Prof. Gilbert Metcalf discuss implementing a vehicle miles traveled (VMT) tax for electric vehicles, instead of the flat annual fee of $130 recently proposed by Congress. “A flat EV tax does not differentiate between high-mileage and low-mileage drivers — a VMT tax does,” write Knittel and Metcalf. “It links payments to actual road use, which is precisely what proponents of EV fees, including the Bipartisan Policy Center and the Competitive Enterprise Institute, say they want.”
In a study of developers using GitHub Copilot, MIT researchers found AI shifted time away from teamwork and toward coding, writes Forbes’ Sarah Davis. “If AI reduces the need to ask colleagues for help and advice, workplaces become the mechanism for creating human interactions that promote connection and knowledge sharing,” writes Davis.
Prof. Emilio J. Castilla speaks to Forbes’ Lauren Russell about how bias can shape employee performance reviews. “Bias—whether explicit or unconscious—can potentially shape employment decisions at every stage, from who is interviewed and hired to who is promoted and who is rewarded for excellent performance,” says Castilla.
In an opinion piece for Fast Company, Hala Hanna, Executive Director of MIT Solve, outlines a blueprint for solving the world’s hardest problems in the next decade. “The last decade rewarded whoever could invent the best solution. The next one will belong to whoever can build the conditions for good solutions to survive, spread, and rewire the systems around them,” writes Hanna.
Research Scientist Giovanni Lavezzi speaks to Scientific American’s Jonathan O’Callaghan about managing the number of satellites in Earth’s orbit. “So far [the amount] seems manageable” through active coordination, says Lavezzi. “The problem is how much can we push.”
MIT researchers found the more data an AI model is trained on could complicate AI copyright, auditing, and governance, writes Computer World’s Taryn Plumb. “[S]ingle artworks by specific artists, or photographs of certain people, could be entirely removed from datasets, and the model could still reproduce that image or style,” writes Plumb. “Essentially, tangible connections are lost, and linking to specific data points responsible for generated samples is ‘practically impossible,’ or can even vanish, the researchers explained.”