In 2020, a machine-learning algorithm helped researchers to develop a potent antibiotic that works against many pathogens (see Nature https://doi.org/ggm2p4; Artificial intelligence (AI) is also being used to aid vaccine development, drug design, materials discovery, space technology and ship design. Within a few years, numerous inventions could involve AI. This is creating one of the biggest threats patent systems have faced. Patent law is based on the assumption that inventors are human; it currently struggles to deal with an inventor that is a machine.

A soft robot with no motors, batteries or computers can roll over a range of surfaces and escape simple mazes by harvesting heat energy and turning it into motion. Jie Yin at North Carolina State University and his colleagues created a spiral-shaped device from a narrow rectangle of rubber-like material impregnated with liquid crystals. When placed on a surface heated to at least 55 C, the areas of the robot touching the surface warm up and expand, while others remain static. This causes a twisting motion that rolls the device along at speeds up to 3.8 millimetres per second. Although the robot has no computational ability, it can achieve relatively complex tasks such as navigating mazes.

Earlier this month, DeepMind presented a new "generalist" AI model called Gato. The model can play the video game Atari, caption images, chat, and stack blocks with a real robot arm, the Alphabet-owned AI lab announced. All in all, Gato can do hundreds of different tasks. But while Gato is undeniably fascinating, in the week since its release some researchers have got a bit carried away. One of DeepMind's top researchers and a coauthor of the Gato paper, Nando de Freitas, couldn't contain his excitement.

The UK's data watchdog has fined a facial recognition company £7.5m for collecting images of people from social media platforms and the web to add to a global database. The Information Commissioner's Office (ICO) also ordered US-based Clearview AI to delete the data of UK residents from its systems. Clearview AI has collected more than 20bn images of people's faces from Facebook, other social media companies and from scouring the web. John Edwards, the UK information commissioner, said Clearview's business model was unacceptable. "Clearview AI Inc has collected multiple images of people all over the world, including in the UK, from a variety of websites and social media platforms, creating a database with more than 20bn images," he said. "The company not only enables identification of those people, but effectively monitors their behaviour and offers it as a commercial service.

Explains Nikola Konstantinov of Switzerland's ETH Zürich, "Fairness in AI is about ensuring that AI models don't discriminate when they're making decisions, particularly with respect to protected attributes like race, gender, or country of origin." As artificial intelligence (AI) becomes more widely used to make decisions that affect our lives, making certain it is fair is a growing concern. Algorithms can incorporate bias from several sources, from the people involved in different stages of their development to modelling choices that introduce or amplify unfairness. A machine learning system used by Amazon to pre-screen job applicants was found to display bias against women, for example, while an AI system used to analyze brain scans failed to perform equally well across people of different races. "Fairness in AI is about ensuring that AI models don't discriminate when they're making decisions, particularly with respect to protected attributes like race, gender, or country of origin," says Nikola Konstantinov, a post-doctoral fellow at the ETH AI Center of ETH Zürich, in Switzerland. Researchers typically use mathematical tools to measure the fairness of machine learning systems based on a specific definition of fairness.

Although workflows are very popular, prior to the CWL standards, all workflow systems were incompatible with each other. This means that users who do not use the CWL standards are required to express their computational workflows in a different way each time they use another workflow system, leading to local success but global unportability. The success of workflows is now their biggest drawback. Users are locked into a particular vendor, project, and often a specific hardware setup, hampering sharing and reuse. Even non-academics suffer from this situation, as the lack of standards, or their adoption, hinders effective collaboration on computational methods within and between companies.

If you own a Tesla, or a loved one does, or you're thinking about buying one, or you share public roads with Tesla cars, you might want to watch the new documentary "Elon Musk's Crash Course." Premiering Friday on FX and Hulu, the 75-minute fright show spotlights the persistent dangers of Tesla's automated driving technologies, the company's lax safety culture, Musk's P.T. Barnum-style marketing hype and the weak-kneed safety regulators who seem not to care. Get Screen Gab for weekly recommendations, analysis, interviews and irreverent discussion of the TV and streaming movies everyone's talking about. You may occasionally receive promotional content from the Los Angeles Times. The central through line is the story of Joshua Brown, a rabid Tesla fan and derring-do techno-geek beheaded when his Autopilot-engaged Tesla drove itself at full speed on a Florida highway underneath the trailer of a semi-truck in 2016.

The technology consists of sensors that use WiFi signals to help the robot map where it's going. Most systems rely on optical light sensors such as cameras and LiDARs. In this case, the so-called "WiFi sensors" use radio frequency signals rather than light or visual cues to see, so they can work in conditions where cameras and LiDARs struggle -- in low light, changing light, and repetitive environments such as long corridors and warehouses. And by using WiFi, the technology could offer an economical alternative to expensive and power hungry LiDARs, the researchers noted. A team of researchers from the Wireless Communication Sensing and Networking Group, led by UC San Diego electrical and computer engineering professor Dinesh Bharadia, will present their work at the 2022 International Conference on Robotics and Automation (ICRA), which will take place from May 23 to 27 in Philadelphia.

Federal authorities are investigating whether a Tesla involved in a crash that left three people dead and three others injured last week in Newport Beach had its Autopilot system activated at the time of the wreck. A special crash investigation team was sent for the May 12 incident on Pacific Coast Highway, the National Highway Traffic Safety Administration said Wednesday. In that crash, Newport Beach police were called around 12:45 a.m. to the 3000 block of Pacific Coast Highway, where they found a 2022 Tesla Model S sedan had crashed into a curb and hit construction equipment. Three people were found dead in the Tesla; they were identified last week as Crystal McCallum, 34, of Texas; Andrew James Chaves, 32, of Arizona; and Wayne Walter Swanson Jr., 40, of Newport Beach, according to the Orange County Sheriff's Department. Three construction workers suffered non-life-threatening injuries, police said, adding that the department's Major Accident Investigation Team had been brought in.

A machine learning model can solve physics problems by simplifying them to be more symmetric. "There are many, many cases in the history of science where people thought things were more complicated than they actually were because they hadn't found the most simple description of it," says Max Tegmark at the Massachusetts Institute of Technology (MIT).