Try taking a picture of each of North America's roughly 11,000 tree species, and you'll have a mere fraction of the millions of photos within nature image datasets. These massive collections of snapshots ' ranging from butterflies to humpback whales ' are a great research tool for ecologists because they provide evidence of organisms' unique behaviors, rare conditions, migration patterns, and responses to pollution and other forms of climate change. While comprehensive, nature image datasets aren't yet as useful as they could be. It's time-consuming to search these databases and retrieve the images most relevant to your hypothesis. You'd be better off with an automated research assistant ' or perhaps artificial intelligence systems called multimodal vision language models (VLMs). They're trained on both text and images, making it easier for them to pinpoint finer details, like the specific trees in the background of a photo. But just how well can VLMs assist nature researchers with image retrieval' A team from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL), University College London, iNaturalist, and elsewhere designed a performance test to find out. Each VLM's task: locate and reorganize the most relevant results within the team's 'INQUIRE' dataset, composed of 5 million wildlife pictures and 250 search prompts from ecologists and other biodiversity experts. Looking for that special frog...

I research the intersection of artificial intelligence, natural language processing and human reasoning as the director of the Advancing Human and Machine Reasoning lab at the University of South Florida. I am also commercializing this research in an AI startup that provides a vulnerability scanner for language models. At the heart of commercially available generative AI products like ChatGPT are large language models, or LLMs, which are trained on vast amounts of text and produce convincing humanlike language. Their size is generally measured in parameters, which are the numerical values a model derives from its training data. The larger models like those from the major AI companies have hundreds of billions of parameters. First, organizations with the most computational resources experiment with and train increasingly larger and more powerful language models. Those yield new large language model capabilities, benchmarks, training sets and training or prompting tricks. In turn, those are used to make smaller language models ' in the range of 3 billion parameters or less ' which can be run on more affordable computer setups, require less energy and memory to train, and can be fine-tuned with less data....

Craig Wright, the computer scientist ruled to have lied 'extensively and repeatedly' about being the inventor of Bitcoin, has been given a one-year prison sentence by a UK judge after being found in contempt of court. The sentence is suspended for two years, meaning that Wright will only face prison if he reoffends during that period. At a hearing Thursday in the UK High Court, Justice James Edward Mellor ruled that Wright'in bringing a $1.15 trillion lawsuit in October against Bitcoin developers and payments firm Square'had violated an earlier court order. The order required that Wright refrain from claiming publicly to be Satoshi Nakamoto, the creator of Bitcoin, and taking legal action on that basis, among other things. The contempt of court issue was raised by the Crypto Open Patent Alliance (COPA), a nonprofit consortium of crypto firms, which in February took Wright to trial in the hope of securing a formal declaration that he is not Satoshi. The aim was to prevent Wright from carrying forward multiple separate lawsuits against Bitcoin developers and other parties, through which he was trying to assert intellectual property rights over Bitcoin'and to ward off any future lawfare....

Ancient human relatives ran on two legs, like modern humans, but at a much slower pace, suggest 3D computer simulations of Australopithecus afarensis1 ' a small hominin that lived more than three million years ago. The analysis offers a detailed snapshot of the hominin's running speed and the muscular adaptations that enabled modern humans to run long distances, says Herman Pontzer, an evolutionary anthropologist at Duke University in Durham, North Carolina. 'It's a very thorough approach,' he says. The findings were published this week in Current Biology. A. afarensis walked upright on two legs, making its fossils a favourite for researchers looking to unpick how bipedalism evolved in the human lineage. But few studies have explored the hominin's running ability because it requires more than studying fossilized footprints and bones, says study co-author Karl Bates, an evolutionary biomechanics researcher at the University of Liverpool, UK. Bates and his colleagues created a 3D digital model of the 'Lucy' skeleton ' a near-complete 3.2-million-year-old A. afarensis specimen discovered in Ethiopia half a century ago. They used the muscular features of modern apes and the surface area of Lucy's bones to estimate the ancient hominin's muscle mass. The researchers then used a simulator to make their Lucy model 'run' and compared its performance with that of a digital model of a modern human....