Add BrightSurf on Google Email

Massachusetts Institute of Technology


How adults understand what kids are saying

Researchers used computational models to analyze thousands of hours of transcribed audio recordings of children and adults interacting. The findings suggest that adults' ability to make context-based interpretations provides crucial feedback for babies acquiring language. These interpretations are critical for understanding what small ...

SourceMassachusetts Institute of Technology·JournalNature Human Behaviour·DateOct 26, 2023

LIGO surpasses the quantum limit

Researchers at LIGO have developed a significant advance in quantum squeezing technology, allowing them to measure undulations in space-time across the entire range of gravitational frequencies detected by LIGO. This breakthrough boosts the observatory's ability to study exotic events and detect about 60 percent more mergers than before.

SourceMassachusetts Institute of Technology·JournalPhysical Review X·DateOct 23, 2023

To excel at engineering design, generative AI must learn to innovate, study finds

Researchers at MIT found that similarity-focused generative AI models falter when tasked with designing new products, highlighting the need to prioritize innovation in engineering tasks. By adjusting training objectives and metrics, AI can be an effective 'co-pilot' for engineers, enabling faster creation of innovative products.

SourceMassachusetts Institute of Technology·JournalComputer-Aided Design·DateOct 19, 2023

Germicidal UV lights could be producing indoor air pollutants, study finds

New research from MIT found that germicidal UV lights can produce potentially harmful compounds in indoor spaces. The study suggests that the lights should be used with appropriate ventilation to minimize health risks. Researchers emphasize that the new UV lights are not a replacement for ventilation but rather a complement to it.

SourceMassachusetts Institute of Technology·JournalEnvironmental Science & Technology·DateOct 18, 2023

Is AI in the eye of the beholder?

Researchers discovered that users' prior beliefs about an AI chatbot's motives significantly impact their interactions with the agent. Priming users to believe certain things about the AI's empathy, neutrality, or manipulation influences their perception of its trustworthiness and effectiveness.

SourceMassachusetts Institute of Technology·JournalNature Machine Intelligence·DateOct 2, 2023

Computational model helps with diabetes drug design

Researchers developed a computational model to analyze glucose-responsive insulin (GRI) performance in human patients. The model predicted that differences in sugar receptor behavior between humans and lab animals led to the drug's poor effect in clinical trials. This breakthrough helps researchers design better GRIs, potentially reduc...

SourceMassachusetts Institute of Technology·JournalACS Pharmacology & Translational Science·TypeComputational simulation/modeling·DateSep 20, 2023

How to keep people out of the emergency room

A new study co-authored by MIT economist Jonathan Gruber found that a New York City program helped arrange medical appointments for undocumented immigrants with limited incomes, leading to a substantial drop in emergency room use. The program resulted in a 21 percent decline in ER visits and a 42 percent decrease for individuals with h...

SourceMassachusetts Institute of Technology·JournalAmerican Economic Review·DateSep 19, 2023

AI models are powerful, but are they biologically plausible?

Researchers propose a hypothesis that astrocytes, non-neuronal cells in the brain, can perform core computation as transformers, providing insights into human brain function and machine learning success. This discovery could spark future neuroscience research and help explain transformer performance across complex tasks.

SourceMassachusetts Institute of Technology·JournalProceedings of the National Academy of Sciences·DateAug 15, 2023