// NATURE NEWS — SPAZIO & SCIENZA
Daily briefing: AI agents sniff out decades-old errors in scientific literature
Search author on:
PubMed
Google Scholar
Hello Nature readers, would you like to get this Briefing in your inbox free every day? Sign up here.
A laser beam tracks the motions of the tiny magnet suspended in a vacuum chamber as it wobbles in response to magnetic fields.Credit: Wei Ji, Peking University
A new magnetometer — a device used to measure magnetic fields — is sensitive enough to detect the faint signals of the brain’s electrical activity. The technique uses laser beams to track the motions of a tiny levitating magnet suspended in a vacuum chamber. This method is simpler than existing approaches, which require temperatures close to absolute zero or near-perfect shielding from external magnetic interference. The device could be used to detect magnetic fields produced by the brain or heart, and similar levitation techniques could be applied to search for dark matter, researchers say.
Scientists using artificial-intelligence agents to audit the scientific literature have uncovered decades-old errors, from typos to incorrect values in the boiling-point measurements of molecules. The advantage of using AI to fact-check is the speed at which it can scan scientific databases and literature compared with humans, says computer scientist James Zou. But others argue that AI is still prone to mistakes, and its use should be focused on identifying ‘objective’ errors, such as those in calculations, and not assessing how data are interpreted or the novelty of the research.
The number of people living in eastern China who have been evacuated from their homes in the wake of heavy rainfall, flooding and the risk of landslides caused by tropical storm Dolphin. (BBC | 5 min read)
A mutation in a specific gene called FNIP1 is linked to a host of health benefits such as decreased levels of belly fat and blood sugar and a reduced risk of conditions including type 2 diabetes and heart disease. The gene plays a part in cellular metabolism and in how cells detect and respond to certain nutrients. The mutations appear to be rare, but the findings reveal a possible target for drugs that could one day reproduce some of the mutations’ protective effects against cardiometabolic diseases.
The citations included in patents — documents filed to gain legal protection for potentially money-making inventions — can provide a glimpse of how academic research flows into commercial innovation. Although many papers have obvious importance, some studies in this list are rarely mentioned in academic journals, and even their authors don’t know why they are cited in so many patents. Nonetheless, the analysis does reveal insights about the influence of academia on industry: not least, that the majority of papers most cited in patents are the product of publicly funded research. (Nature | 15 min read) (Source: Nature analysis/relianceonscience.org)
After suspecting that his students cheated on an at-home mid-term exam, economist Robert Serrano held the course’s final examination in person — the class’s average score plummeted. The shock “was not that cheating seems to have occurred on the take-home exam, but how prevalent it appeared to be”, he writes. Universities should establish dedicated committees to ensure academic integrity is upheld, and work as a collective to establish best practices for the appropriate use of AI, to ensure that such tools don’t undermine learning, he argues.
Geopolitical unrest, rising costs of living, research funding cuts and a widespread sense of decreasing trust in science can make research an uncertain career path for people in Generation Z — those born after 1997. But there are ways to stay motivated and cultivate resilience, four early-career researchers tell Nature. One is to find a research environment that suits your needs, and a support network for the difficult times. Another is to advocate for change from within. “We, as scientists, must suppo