A detailed virtual house will help robots train to become your butler
A new digital training ground that replicates an average home lets AI learn how to do simple chores like slicing apples, making beds, or carrying drinks in a low-stakes environment.
Background: We all want a robot to run around our home and fetch us a beer. But teaching them to do it in the real world is expensive, because they’re still clumsy and make tons of mistakes.
Virtual beer fetching: So researchers have turned to training AI in virtual settings. Typically those spaces are video games like Doom or Grand Theft Auto. But IEEE Spectrum reports a new training ground, called AI2-THOR, lets AI interact with objects like refrigerators and furniture in something like the real world.
Why it matters: Think of all the unbroken glassware. Beyond saving time and money, AI2-THOR could teach AIs skills that are genuinely useful. Exploring the relatively complex, messy settings that humans inhabit could let AI learn more as we do, too.
Deep Dive
Artificial intelligence
Large language models can do jaw-dropping things. But nobody knows exactly why.
And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.
Google DeepMind’s new generative model makes Super Mario–like games from scratch
Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.
What’s next for generative video
OpenAI's Sora has raised the bar for AI moviemaking. Here are four things to bear in mind as we wrap our heads around what's coming.
Stay connected
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.