MIT researchers found that purchased AI tools succeed roughly twice as often as comparable in-house builds, Forbes’ Dennis Vorobyov highlights. “[S]o spend the engineering effort where you are genuinely differentiated,” writes Vorobyov. “Budget wise, none of this requires a moonshot, just clarity and sequencing.”
MIT’s Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training has released a report on AI and education and recommendations for the Institute moving forward, writes Forbes’ Ron Schmelzer. “President Sally Kornbluth describes the moment as a ‘watershed’ for MIT and higher education,” writes Schmelzer. “The recommendations include redesigned assessments, renewed attention to hands-on learning, explicit AI rules for courses and new communities of practice for faculty.”
Prof. Emeritus Rodney Brooks speaks to Scientific American’s Mary Randolph about humanoid robots interacting with the real world. “When you see a performance of an [artificial intelligence] system or a robot on one thing,” says Brooks, “that fools us into thinking that it has the same general competence as a human. And that’s a mistake people make.”
A new study by MIT researchers found that it is impossible to draw a direct link between AI diffusion model outputs and an artist’s body of work, writes Gizmodo’s Webb Wright. “If a [diffusion] model generates something, you want to be able to say, ‘Oh, this part of the training data was responsible,'” says Zheng Dai, PhD ’24, lead author on the study. “It’s important for us to understand how these models work to properly study them or regulate them.”
In a message to campus, President Sally Kornbluth shared a new report by MIT’s Ad Hoc Committee on AI Use that found that “even as AI accelerates and expands discoveries, it is also driving dramatic changes in campus culture, upending foundational experiences such as study groups, office hours and undergraduate research,” writes the Washington Post’s Susan Svrluga. “For the sake of our students, for the future of MIT and for the shape of our sector and our society,” says Kornbluth, “it’s imperative that we get this right.”
Forbes’ Satya Krishnaswamy highlights a study co-authored by Lawrence Schmidt, a research affiliate at Sloan, that found AI can contribute to growth in certain roles where workers focus on higher-value tasks. “Schmidt’s research found that increased AI use was linked to both employment and sales growth,” writes Krishnaswamy.
Writing for Forbes, Arvid Ernst Gollwitzer, a research scholar at the Broad Institute of MIT and Harvard, spotlights FINGERS-7B, an AI foundation model for Alzheimer’s prevention his research team developed at MIT. “By analyzing signals spanning everything from genetic risk factors to long-term environmental exposures, the model detects patterns that no single-domain approach could find,” writes Gollwitzer. “These patterns, which we call multi-omic precision signatures, represent a new kind of biomarker for gauging early disease risk.”
MIT researchers have found that AI search features are unevenly distributed worldwide, and can sometimes share misleading information about images, writes Forbes’ Hastimal Jangid. “These incidents are important reminders that systems are only as good as the data they're trained on and that the long tail of data documenting the physical world is still being filled in,” writes Jangid.
In a Financial Times opinion piece, Prof. Carlo Ratti discusses alternatives to implementing traditional air conditioning systems, which move rather than reduce heat, that could be used in Europe as temperatures continue to rise. “The best answer, then, is a mixed system, co-ordinated in real time, with sensors to track temperature, occupancy and demand, and algorithms deciding when to produce, store and distribute cooling in step with the electricity grid,” writes Ratti.
Forbes’ Bryan Robinson highlights Prof. Paul Osterman’s new book, “Disposable Workers: The Transformation of Employment,” and identifies what constitutes disposable work. “We’re not becoming a gig economy; we’re becoming a disposable one,” says Osterman. “Marginal workers are employees who have no career prospects at their organizations.”
Prof. David Gifford and Zheng Dai PhD ’24 discovered that AI diffusion models can replicate the style of an artist even when their work was omitted from datasets, leading to ‘attribution decay,’ writes Fast Company’s Jesus Diaz. “This trend, what the researchers call ‘attribution decay,’ means that artificial intelligence can produce an image that resembles a particular artist’s work, while having no provable causal link to that artist’s actual contribution to the training data,” explains Diaz.
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.”
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.”
The Register’s Thomas Claburn highlights a new study by MIT researchers that found large AI models can still reproduce an artist’s style, even when the artist’s work is removed from its training data, making attribution more difficult. “[G]iven the contemporary adoption of these models for creative and commercial purposes, attributability also carries ethical, policy, financial, and legal implications,” the researchers note.
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.”