There’s a new snake-loving website that wants to crowdsource reptile and amphibian identification, and it hopes your backyard photos can help its AI bot learn how to do it. The Verge has a rundown of What’s the Herp? and its related site HerpMapper, the products of developer Don Becker and a team of IT professionals, biologists, and herpetologists who hope to remove some of the stigma surrounding snakes and their assorted relatives.
“People kill things because they’re afraid of them,” Becker told The Verge. “The more you learn about something, the more you appreciate it.”
The AI bot is named Fitch, whose namesake, the late Henry S. Fitch, is known as the “father of snake biology.” The team is hoping the public will join in to assist Fitch in his endeavors.
https://twitter.com/FieldEcology/status/946454793232535553
You can create an account on the What the Herp site and upload your own photos, or help out by flipping through the photos already on the site and drawing boxes around the critters. Fitch is easily confused by background environments in the photos, and annotating them helps in his training. “It’s not the brightest thing in the world right now,” Becker admitted. “But it’s getting there.”
The ultimate goal is to have a way for people to identify snakes and other creatures in their backyard. “By giving people a way to identify what it is that they’re looking at, that can help dissuade people from killing it,” said Becker. They want to have at least 250 images of each species of reptile and amphibian on file, and there’s also a “Most Wanted List” to help reach that goal.
You can also quiz Fitch by sending a photo to their Twitter account, and he’ll reply with his best guess as to what the creature is.
Their companion site, HerpMapper, is a non-profit designed to collect and share data about amphibians and reptiles from around the world. Members can submit images or audio files, along with the location using GPS data.
Detailed location info is not available to the public, to help conserve sensitive or endangered species. Only partners of the site, such as state biologists, university researchers, or conservation organizations can see specific location data.
Becker admits that it’s no small challenge to try and get people to like snakes. “The biggest issue with conservation of herps is that we work with one of the most detested groups on the planet — you’re up there with spiders,” he said.
San Francisco startup Dimensional advertised for people to walk robot dogs through city streets. The New York Times reported that the job could include carrying a roughly 33-pound robot back to the office if it overheated. A lift home is an awkward requirement for a labor-saving machine.
Keeping the hardware running is only one challenge. Anthropic's robotics tests found that language models mostly failed when controlling joints directly, but did better when directing trained controllers. Giving an AI agent a body doesn't teach it to handle ordinary household objects.
Soon, a mattress protector may be more than a bed cover. Australian medtech company Sleeptite and researchers at RMIT University have received a US patent for FlexiSense, a smart bedding technology that tracks body position and posture to help caregivers identify when older residents may be at risk of falling.
The technology is being developed for Sleeptite’s in-bed monitoring system, REMi, and was recently on display at The MedTech Conference in Boston. A peer-reviewed study in Flexible and Printed Electronics says the sensors can also be made at scale without sacrificing performance or durability.
AI companies spend enormous amounts of time convincing us their technology is safe and secure. But what happens when those same companies become targets of increasingly sophisticated AI-powered attacks? Anthropic, the company behind Claude, is now facing that question. According to a Washington Post report, cybersecurity startup Cathedral discovered vulnerabilities that let its researchers access sensitive information in Anthropic's internal systems.
The researchers used AI agents to identify potential entry points in less than 12 hours. Interestingly, they didn't need to crack Claude itself. Instead, they exploited weaknesses in third-party business applications that Anthropic uses for everyday operations. Anthropic's security testing program disclosed the findings, which allows outside researchers to identify and report vulnerabilities.