Research

Prof. Joseph Doyle and his colleagues are studying whether type 2 diabetes could be treated or improved by nutrition, reports Donna Eastlake for Food Navigator.

Senior lecturer Renée Richardson Gosline speaks with Boston Globe reporter Maddie Khaw about new marketing strategies and how they are being used to promote the latest water bottle trends.  “If you can get the right influencer to talk about your stuff, you’ve got more targeted communications,” says Gosline. “What people buy has started to replace a sense of social connection and identity that is missing for many people.”

Fast Company reporter Shalene Gupta spotlights new research by Prof. David Autor that finds “about 60% of jobs in 2018 did not exist 1940. Since 1940, the bulk of new jobs has shifted from middle-class production and clerical jobs to high-paid professional jobs and low-paid service jobs.” Additionally, the researchers uncovered evidence that “automation eroded twice as many jobs from 1980 to 2018 as it had from 1940 to 1980. While augmentation did add some jobs to the economy, it was not as many as the ones lost by automation.”

Prof. David Autor speaks with New York Times reporter Steve Lohr about his hope that AI can be harnessed to become “worker complementary technology,” enabling individuals to take on more highly skilled work and find better paying jobs. “I do think there is value in imagining a positive outcome, encouraging debate and preparing for a better future,” Autor explains. “This technology is a tool, and how we decide to use it is up to us.”

Researchers at MIT have found that large language models mimic intelligence using linear functions, reports Kyle Wiggers for TechCrunch. “Even though these models are really complicated, nonlinear functions that are trained on lots of data and are very hard to understand, there are sometimes really simple mechanisms working inside them,” writes Wiggers. 

Forbes reporter Oludolapo Makinde spotlights research by Prof. Daron Acemoglu and Prof. Simon Johnson that explores the impact of AI on the workforce. “Instead of aiming to create artificial superintelligence or AI systems that outperform humans, [Acemoglu and Johnson] propose shifting the focus to supporting workers,” writes Makinde.

MIT researchers have developed a new tool to quantify how climate change will impact the number of “outdoor days” where people can comfortably spend time outside in specific locations around the world, reports Tim DeChant for TechCrunch. “The MIT tool is a relatable application of a field of study known as climate scenario analysis, a branch of strategic planning that seeks to understand how climate change will impact various regions and demographics,” writes De Chant.

MIT researchers have discovered a protein found in human sweat that holds antimicrobial properties and can “inhibit the growth of the bacteria that causes Lyme disease,” reports Matt Fortin for NENC. The team believes this “type of protein could be put into a topical cream to make something called ‘Lyme Block’ – like sunblock, but for preventing Lyme.”  "Ideally what we would love to do is give people more control over their own risk," says Principal Research Scientist Michal Tal. "And really try to develop this into a possible preventative that you could put on repellant or sunblock to protect against other elements of the outdoors that you could also protect yourself against Lyme."

A new report by researchers from MIT and Boston Consulting Group (BCG) has uncovered “how AI-based machine learning and predictive analytics are super-powering key performance indictors  (KPIs),” reports Sheryl Estrada for Fortune. “I definitely see marketing, manufacturing, supply chain, and financial folks using these value-added formats to upgrade their existing KPIs and imagine new ones,” says visiting scholar Michael Schrage.

Astronomers from MIT and other institutions have found that periodic eruptions from a supermassive black hole located in a galaxy about 800 million light-years from Earth could be caused by a, “second, smaller black hole slamming into a disk of gas and dust, or ‘accretion disk,’ surrounding the supermassive black hole, causing it to repeatedly ‘hiccup’ out matter,” writes Rob Lea for Space.com. 

Research Scientist Robert Ajemian, graduate student Greta Tuckute and MIT Museum Exhibit Content and Experience Developer Lindsay Bartholomew appear on The Economist’s Babbage podcast to discuss the development of generative AI. “The way that current AI works, whether it is object recognition or large language models, it’s trained on tons and tons and tons of data and what it’s essentially doing is comparing something it’s seen before to something it’s seeing now,” says Ajemian.  

FutureTech researcher Tamay Besiroglu speaks with New Scientist reporter Chris Stokel-Walker about the rapid rate at which large language models (LLMs) are improving. “While Besiroglu believes that this increase in LLM performance is partly due to more efficient software coding, the researchers were unable to pinpoint precisely how those efficiencies were gained – in part because AI algorithms are often impenetrable black boxes,” writes Stokel-Walker. “He also points out that hardware improvements still play a big role in increased performance.”

Financial Times reporters Andrew Hill and Emma Jacobs spotlight how Prof. Carlo Ratti and his colleagues have found that “people are traveling less far from their homes, consistent with more remote working.” The findings demonstrate a more flexible approach to commuting. “That’s good news [for cities], because the problem is not a problem of capacity, it’s a problem of peaks,” Ratti explains.

Prof. Katharina Ribbeck speaks with Christopher Intagliata of Scientific American’s “Science Quickly” podcast about her research exploring how mucus can treat and prevent disease. “The basic building blocks of mucus that give mucus its gooey nature are these threadlike molecules—they look like tiny bottlebrushes—that display lots and lots of sugar molecules on their backbone,” explains Ribbeck. “And these sugar molecules—we call them glycans—interact with molecules from the immune system and microbes directly. And the exact configuration and density of these sugar molecules is really important for health.”

MIT researchers have developed a new technique that uses a large language model to allow robots to self-correct after making a mistake, reports Brian Heater for TechCrunch. “Researchers behind the study note that while imitation learning (learning to do a task through observation) is popular in the world of home robotics, it often can’t account for the countless small environmental variations that can interfere with regular operation, thus requiring a system to restart from square one,” writes Heater. “The new research addresses this, in part, by breaking demonstrations into smaller subsets."