GoKawiilSign up to see the future, today Can’t-miss innovations from the bleeding edge of science and tech Email address Sign Up Thank you!
The adage, “data is the new currency,” has taken on a whole new meaning in a society obsessed with AI. The prevailing thinking is that with more data, AI models will continue to improve dramatically, turning the training input into a highly valuable resource or capability.
Some startups are taking this notion to the extreme. As the BBC reports, an AI company called Micro AGI is sending college-educated Silicon Valley types, loaded with cameras, to scrub filthy New York City apartments for free as part of an initiative called Shift. The hope is that the resulting footage will be worth more to the robotics and AI industries in the form of training data than what residents would pay for the service.
The collection of often-sensitive training footage has developed into an entire cottage industry, a data marketplace that has businesses stumbling over themselves.
Shift’s bizarre initiative highlights just how desperate startups have become to stay relevant in a rapidly changing business landscape. Instead of cushy, office-bound jobs, entrepreneurs are cleaning toilets and doing the dishes in the midst of a dire job market.
The BBC‘s Archie Mitchell had “two mid-twenties college graduates who have bounced around the start-up world and were looking to work” show up at his apartment in New York’s Upper East Side. The pair said they were cleaning around five apartments a day, five days a week.
Cameras attached to the front of their baseball caps captured everything they did, data that could one day teach humanoid or other specialized robots how to do the job instead of them.
It’s a reminder of how the AI and robotics industries are essentially looking to put workers out of a job. But that’s a reality that Shift isn’t eager to focus on. Instead, founder Bercan Kilic told the BBC the venture is trying to “advance humanity.”
The main technical challenge? Lighting, oddly.
“In the real world, every object is different, the lighting is different and nothing is the same as it was a couple of hours earlier,” Kilic told the BBC. “Models need to learn how their hands, cameras and environments work together.”
... continue reading