Using publicly available AI foundation models such as OpenAI’s ChatGPT and Meta’s Segment Anything Model (SAM), Duke engineers built an AI microscope platform that can analyze materials as accurately as a trained graduate student in a fraction of the time. The findings point to a new era of autonomous research, where AI systems work alongside humans to design experiments, run instruments and interpret data.

Researchers at the University of Kansas have conducted a study based on their building of the Kansas Energy Transition Atlas. The tool gathers wind energy regulations for all of the state's 105 counties, showing what kind of regulations each has, as well as information on transmission lines, turbine height and much more. The study shows how many counties have regs that enable wind development, how many seek to block them and show that agricultural reliance factors into allowing wind in a county, while voting patterns did not factor in to types of regs. The site not only shows how a state with massive wind energy potential handles the issue, but can help guide the nation in energy transition policies, the researchers argue.

New research suggests that even warnings about near-term job automation do little to shake public confidence. Two political scientists used a survey to see how people would respond to forecasts of the arrival of “transformative AI,” from as early as 2026 to as distant as 2060. They found that shorter timelines made respondents slightly more anxious about losing their jobs, but did not meaningfully alter their views on when job losses would occur or their support for government responses such as retraining workers or providing a universal basic income.

Older adults face a high risk of isolation, especially after they retire. A recent study from the University of Georgia suggests smartphones, computers and tablets may help older adults stay connected.

Almost every big life change, from starting a new school to changing jobs or moving towns, elicits the same advice: get involved and make friends. While there’s some truth to that guidance, it takes more than simply participating in activities to grow a social network, according to research by an international team.

A multinational collaboration at Eitri medical innovation center in Bergen, Norway, has used machine learning models to identify patient groups at risk of being mistreated at the ER.