Google's AI Can Make Its Own AI Now


Artificial intelligence is advanced enough to do some pretty complicated things: read lips, mimic sounds, analyze photographs of food, and even design beer. Unfortunately, even people who have plenty of coding knowledge might not know how to create the kind of algorithm that can perform these tasks. Google wants to bring the ability to harness artificial intelligence to more people, though, and according to WIRED, it's doing that by teaching machine-learning software to make more machine-learning software.

The project is called AutoML, and it's designed to come up with better machine-learning software than humans can. As algorithms become more important in scientific research, healthcare, and other fields outside the direct scope of robotics and math, the number of people who could benefit from using AI has outstripped the number of people who actually know how to set up a useful machine-learning program. Though computers can do a lot, according to Google, human experts are still needed to do things like preprocess the data, set parameters, and analyze the results. These are tasks that even developers may not have experience in.

The idea behind AutoML is that people who aren't hyper-specialists in the machine-learning field will be able to use AutoML to create their own machine-learning algorithms, without having to do as much legwork. It can also limit the amount of menial labor developers have to do, since the software can do the work of training the resulting neural networks, which often involves a lot of trial and error, as WIRED writes.

Aside from giving robots the ability to turn around and make new robots—somewhere, a novelist is plotting out a dystopian sci-fi story around that idea—it could make machine learning more accessible for people who don't work at Google, too. Companies and academic researchers are already trying to deploy AI to calculate calories based on food photos, find the best way to teach kids, and identify health risks in medical patients. Making it easier to create sophisticated machine-learning programs could lead to even more uses.

[h/t WIRED]

This Amateur Rocketeer Builds Functioning, Miniature Replicas of SpaceX Rockets

Jeff J Mitchell, Getty Images
Jeff J Mitchell, Getty Images

Amateur rocketry is a hobby that predates NASA. Hobbyists have successfully made it to space using rockets built without the massive budgets and resources available to larger organizations. And some of these rockets do more than reach incredible heights: As Motherboard reports, Joe Barnard, a 25-year-old rocketeer from Nashville, Tennessee, is working on making model rockets capable of propulsive landings, the same trick that makes some SpaceX rockets reusable.

Most rocket boosters that propel loads past the Earth's atmosphere are designed to go only one way. In 2015, Elon Musk's space exploration company SpaceX made history when it successfully maneuvered the boosters used to launch its Falcon 9 rocket back onto the landing pad. SpaceX says its latest version of the rocket can be re-flown up to 100 times, saving the company millions of dollars per launch.

Joe Barnard is bringing this same level of innovation to the amateur rocketry world. He first became interested in aerospace engineering after watching early SpaceX videos, and instead of earning a degree in the field, he taught himself the basics. He's since made rocketry into a career, founding Barnard Propulsion Systems (BPS), a small business that sells supplies to other hobbyists, and working on rockets of his own.

Like the rockets at SpaceX, Barnard's creations use thrust vectoring—the technology that makes it possible to navigate and stabilize a rocket after launch—only on a much smaller scale. He's built miniature models of SpaceX's Falcon 9 rockets, and as is the case at SpaceX, his launches don't always run smoothly.

Barnard is still perfecting propulsive landings in amateur rockets, but for now he says each failure is a learning experience. You can watch the progress of his experiments on his YouTube channel.

[h/t Motherboard]

This High-Tech Skin Turns Almost Any Object Into a Robot


Instead of trading in their low-tech toys for fancy new gadgets, kids of the future may have the option to turn Teddy into a cyborg. As Gizmodo reports, researchers at Yale University have invented a robot "skin" that can give mobility to just about any inanimate object that can bend.

The researchers describe how this technology, called OmniSkins, works in a paper published in Science Robotics. The material, made from elastic sheets that contain sensors and actuators, can be wrapped around any malleable surface and controlled remotely or with on-board light sensors, essentially turning the item into a makeshift robot. The actuators in OmniSkins manipulate the object, while sensors can gauge exactly how much pressure needs to be applied to perform an action.

The project was largely funded by NASA, and it's easy to see how the tech might be used on one of the agency's missions in the future. On board a spacecraft, every cubic inch of space is precious, and the ability to repurpose something into a robotic arm or rover in a pinch would be invaluable to astronauts.

But the technology has potential applications on Earth as well, and not just to allow stuffed animals to walk on their own. Add a few of the skins to a shirt and it could automatically adjust your posture throughout the day, or you could use them to create a gripper to handle hard-to-reach objects.

You can see OmniSkins in action in the video below.

[h/t Gizmodo]