PLUS: The latest AI and tech news.
By Jennifer Conrad | 10.04.21 Facial-recognition company Clearview AI stoked controversy by scraping websites, including Facebook, Instagram, and Twitter, for billions of photos that feed into its software. The database has been used by police and others to help identify people in photos. As Will Knight reports, the company believes artificial intelligence can help make its tech even more powerful—but it may also make it more dangerous and error-prone. Clearview's CEO says the company has now collected more than 10 billion images of faces from across the web—more than three times as many as has been previously reported. The allure of such a tool is obvious, but so is the potential for it to be misused. Clearview already faces numerous lawsuits for alleged privacy breaches, and Facebook and Twitter have demanded that the company stop scraping their sites. A Not-So-Clear View? New technologies may spark further debate: Clearview is developing machine-learning tools that "deblur" images and automatically remove a person's mask and generate the lower half of their face—but the technique is essentially a best guess based on statistical patterns found in other images. Read about the features that could make Clearview AI even more controversial. | Court documents suggest the FBI has been using geofence search warrants at a scale not publicly seen before, collecting account information and location data on hundreds of devices inside the US Capitol during the deadly invasion by a right-wing mob on January 6. The investigators were then apparently able to build a large, searchable database in their hunt for the rioters. Forty-five criminal cases cite Google geolocation data to place suspects inside the Capitol that day, a WIRED investigation found. As Mark Harris writes, those cases include at least six where the identity of the suspect appears to have been unknown to the FBI prior to the geofence warrant. Geofence warrants are essentially a fishing expedition: Investigators know roughly where and when a crime was committed, and want to find out who might have been nearby at the time. Google requires law enforcement to go through a three-step process, and only in the final step will Google unmask the real name, email, phone number, and other information of an account holder. A Slippery Slope? Harris' research suggests that investigators had access to personal information about a much larger number of people around the Capitol, which could be combined with surveillance video and other sources to identify suspects. Some say the measure was necessary given the extraordinary circumstances. Others fear the tactic will accelerate its use in routine investigations. Read WIRED's investigation into the geofence warrant. | The electric currents produced by microorganisms that live in soil may be a way to determine how healthy the soil is, writes Jennifer Clare Ball. In August, researchers at Washington State University published a proof-of-concept study demonstrating that they could use electrodes to measure electric currents produced by microbes. Detecting a current meant that the soil was healthy, because those microbes were conducting metabolic activity—doing things like recycling nutrients and creating compounds that help crops weather environmental stressors. Soil microbes include anything from bacteria to symbiotic fungi. The tiny organisms decompose material, store carbon, transform nitrogen, and enhance the availability of important nutrients. To do these jobs, microbes need nutrients to live and reproduce, and their process of consuming energy produces activity that can be read by a sensor. "The electrochemical sensor is going to be the voice of the soil," says Haluk Beyenal, one of the study's authors. Ground Truth Industrial agriculture—with its reliance on pesticides and focus on growing the same crops year after year—is endangering these microbes. Ultimately that can lead to a loss of biodiversity, and economic losses for farmers. Read about what Beyenal calls his "sci-fi vision." | |
No comments:
Post a Comment