A researcher has identified 230 biological anomalies in 221 scientific publications: impossible experiments that slipped through the peer-review process and were subsequently cited by other studies. With artificial intelligence, the risk of these errors multiplying is increasing. The solution is to change the rules of research, by certifying the original data before it is turned into articles

Chip manufacturers fund the companies that buy their products, fuelling investment and sales. But if revenues prove insufficient to sustain growth, public procurement and incentives could shift part of the risk onto taxpayers

If the results obtained in animals are confirmed in humans, the most significant benefit could be for precisely those patients for whom the prevention of new lesions is no longer sufficient. The study published in Science Translational Medicine

Out of 722 manuscripts published in a GitHub repository, Sam Altman’s company withdrew three due to a sign error. As long as OpenAI continues to produce theorems faster than the world can read them, the word ‘proven’ should be taken with a pinch of salt, even when it is a computer that certifies it.

Seven million studies published in a single year, findings that appear to contradict one another and crucial insights that remain buried. To access scientific literature, we rely on search engines, platforms and artificial intelligence systems that sort through pages and decide which results to show us. This process of selection is inevitable, but the way in which it is carried out often remains opaque

The prize goes to Deisseroth, Hegemann and Nagel for their discoveries concerning light-activated ion channels. The medical applications could prove significant, and some have already begun to emerge

A twenty-year study analyses the reintroduction of wolves into the wildlife of the large nature reserve, refuting the ‘trophic cascade’ theory: ecosystems do not automatically restore their lost state 

A generic ‘International Culture Award’ does not turn its recipient into Benedetto Croce. If the same plaque had been presented at a conference centre on the Tiburtina, no one could have told the story of the ‘journey from prison to the Campidoglio’. Public prestige becomes the packaging for a private promotional campaign

A study in Nature reveals important information about the spread of viruses, but cuts to the WHO’s budget are having an impact on precisely those areas that are essential for continuing this work 

The research points to a plausible way of making articles much more dynamic than the PDFs we still use today, by finally enabling text, code and data to work together