We’ve observed agents discovering progressively more complex tool suse while playing a simple game of hide-and-seek. Through training in our new simulated hide-and-seek environment, agents build a series of six distinct strategies and counterstrategies, some of which we did not know our environment supported. The self-supervised emergent complexity in this simple environment further suggests that multi-agent co-adaptation may one day produce extremely complex and intelligent behavior.
Still using elbow grease to clean the sinks, tubs and other grimy surfaces around your house? Save your elbows, and some time. If you’ve got a power drill, the RevoClean® 4-in-1 Drill Brush Cleaning Kit will instantly turn it into a professional scrubber that can tackle any stain on any surface. Attach the 4″ nylon […]
Need data storage? Join the club. It may still seem like the wild west out there, and for many companies, it’s a tough choice between security and accessibility. Luckily, there’s a platform that gives you a lot of both: Polar Backup Cloud Storage. Whether you’re a busy private citizen or managing valuable company data, Polar […]
There are a lot of different language apps out there because nobody learns anything the same exact way – especially not something as complex as a new language. For some people, the best way is to dive in and start talking, but that’s easier said than done if you’re not around those natives you aspire […]