Prof. Regina Barzilay speaks with The Economist about how AI can help advance medicine in areas such as uncovering new drugs. With AI, “the type of questions that we will be asking will be very different from what we’re asking today,” says Barzilay.
Writing for Nature, graduate student Jelle van der Hilst offers advice on determining whether the data resulting from an experiment is meaningful and useful. “Although in research it is crucial that you don’t fully trust your data until it has been triple-proven and peer-reviewed,” writes van der Hilst, “we do have to gain some operational confidence in our methods and results. Otherwise, crippled by self-doubt, we’d never bring any new research into the world.”
MIT researchers have developed a new implant that in the future could be used to deliver insulin to patients for up to a month, potentially enabling patients to control diabetes without injections, reports Tony Ho Tran for the Daily Beast. In the future, the researchers hope to “develop a device for humans that would be roughly the size of a stick of gum,” writes Tran. “The implant could also be used to deliver things like drugs or proteins to help treat other diseases in humans as well.”
USA Today reporter Zoe Wells spotlights the Mars MOXIE device developed by MIT researchers, which “has already made 122 grams of oxygen, comparable to 10 hours of breathable air for a small dog. MOXIE produced 12 grams of oxygen per hour at 98% purity, which exceeded NASA's original expectations.”
A new study by MIT researchers demonstrates how “machine-learning systems designed to spot someone breaking a policy rule—a dress code, for example—will be harsher or more lenient depending on minuscule-seeming differences in how humans annotated data that were used to train the system,” reports Ananya for Scientific American. “This is an important warning for a field where datasets are often used without close examination of labeling practices, and [it] underscores the need for caution in automated decision systems—particularly in contexts where compliance with societal rules is essential,” says Prof. Marzyeh Ghassemi.
MIT researchers have devised a new carbon capture system that “relies on the reaction between CO2 in flue gas and a fine mist of electrically charged particles,” reports Karmela Padavic-Callaghan for New Scientist. The researchers believe their design could be “95 per cent efficient at capturing carbon dioxide, while measuring less than 4 meters long,” Padavic-Callaghan notes. “Their analysis also showed that the approach would reduce capital costs of adding carbon capture to power plants by about 2.6 times compared with current technology.”
Popular Science reporter Andrew Paul writes that MIT researchers have developed a new long-range, low-power underwater communication system. Installing underwater communication networks “could help continuously measure a variety of oceanic datasets such as pressure, CO2, and temperature to refine climate change modeling,” writes Paul, “as well as analyze the efficacy of certain carbon capture technologies.”
Politico reporter Joanne Kenen spotlights Prof. Adam Berinsky’s new book, “Political Rumors: Why We Accept Misinformation and How to Fight it.” The book “examines attitudes toward both politics and health, both of which are undermined by distrust and misinformation in ways that cause harm to both individuals and society.”
Forbes reporter Rob Toews spotlights Prof. Daniela Rus, director of CSAIL, and research affiliate Ramin Hasani and their work with liquid neural networks. “The ‘liquid’ in the name refers to the fact that the model’s weights are probabilistic rather than constant, allowing them to vary fluidly depending on the inputs the model is exposed to,” writes Toews.
Prof. Arnaud Costinot and Prof. Iván Werning speak with TechCrunch reporter Brian Heater about their research examining the potential impact of a robot tax on automation and jobs. “The potential wages people can earn may become more unequal with new technologies and the idea is that the tax can mitigate these effects,” Costinot and Werning explain. “In a sense, one can think of this as pre-distribution, affecting earnings before taxes, instead of redistribution.”
Research fellow Michael Schrage speaks with Fortune reporter Sheryl Estrada about generative AI’s role in the digital economy. “If you truly understand and structure your use cases for generative AI correctly, there’s much less risk associated with the investment,” says Schrage.
A new working paper, co-authored by Prof. Jonathan Gruber, explores the impact of the New Co-operative Medical Scheme (NCMS), “a health-insurance plan for rural Chinese that was launched in 2003 and folded into a more comprehensive program in 2013,” reports The Economist. “Though it is perhaps best known for being stingy, the NCMS saved millions of lives,” writes The Economist.
Researchers at MIT and elsewhere have developed a medical device that uses AI to evade scar tissue build up, reports Andrew Paul for Popular Science. “The technology’s secret weapon is its conductive, porous membrane capable of detecting when it is becoming blocked by scar tissue,” writes Paul.
A study co-authored by Prof. S.P. Kothari has found that, at an aggregate level, repurchasing shares neither creates nor destroys much wealth, reports Jamie McGeever for Reuters. The study concludes that “buybacks return several hundred billion dollars of capital to shareholders every year and are a mainstream financial avenue open to companies ‘that for the most part do not harm the overall market,” reports Jamie McGeever for Reuters.
In a new MIT study, researchers found that people were less likely to order a menu item when it was specifically labeled as "vegan" compared to when it was not, reports Meryl Davids Landau for National Geographic. “The research is not trying to tell anyone they need to strictly transition into these diets in order to make an impact,” says graduate student Alex Berke. “This is about people eating more sustainably, more often, and what can we do to guide people towards those practices.”