Imagine robots learning complex new skills not in weeks or months, but in mere minutes. Sounds like science fiction, right? Well, an Nvidia-backed startup called General Robotics is making it a reality with their game-changing AI platform, GRID. They’re claiming that GRID slashes robot onboarding time by a jaw-dropping 99.7% and cuts the time it takes to transfer skills between different robot types by nearly 98%. Even wilder? General Robotics says 80% of GRID’s *own code* was written by AI. Talk about eating your own dog food!
So, how do they pull off this magic trick? General Robotics calls their secret sauce “Auto-Engineering.” Instead of tedious manual programming, GRID can apparently watch a short video, reconstruct the motion in a simulation, generate its own training data, and then teach a robot that new skill. It’s like a robot watching a YouTube tutorial and instantly becoming proficient. What’s really cool is how it learns recursively: a robot’s first new task might take a few hours, but subsequent skills on the same platform can deploy in as little as 10 to 15 minutes. It even finds and fixes issues with existing AI models and hardware calibrations on its own!
This isn’t just some lab experiment; it’s already making waves in the real world. General Robotics says GRID is currently deployed with a top U.S. automotive manufacturer, a major global port operator, and even government agencies. If these claims hold up, this technology could drastically reduce the massive costs and complexities that have historically bottlenecked the widespread deployment of advanced robotics. With such impressive backing and real-world traction, GRID could be the platform that finally unlocks the true potential of physical AI, making robots not just possible, but practical for countless new applications.