Skip to main content

Researchers are programming robots to learn as human babies do

faulty robots
Image used with permission by copyright holder
Google is betting big on artificial intelligence. Robots have come a long way over the past decade or so, but they’re still not as good at interacting with humans as they could be. In fact, robots still struggle to do some basic tasks.

A team at Carnegie Mellon, however, is trying to fix that. The team, led by assistant professor Abhinav Gupta, is taking a new approach — allowing robots to play with everyday physical objects and explore the world to help them learn — exactly as a human baby would.

“Psychological studies have shown that if people can’t affect what they see, their visual understanding of that scene is limited,” said Lerrel Pinto, a PhD student in the group, in a report by The Verge. “Interaction with the real world exposes a lot of visual dynamics.”

The group first showed off its tech last year, and the demo helped it land a three-year, $1.5 million award from Google, which will be used to expand on the number of robots that are being used in the study. More robots allows the researchers to gather data more quickly, which basically helps the group build increasingly advanced robots.

But the team isn’t just looking toward more robots to help speed up data gathering. It’s also trying to teach robots skills that will, in turn, help the robot learn other skills. The team also uses adversarial learning — which, according to the Verge report, is akin to a parent teaching a child how to catch a ball by pitching increasingly difficult throws. Apparently, taking this approach results in significantly faster learning than alternative methods.

It will certainly be interesting to see what comes of the project, and we’ll likely hear more about it as time goes on. Check out the video below to see the robots in action.

Robot Adversaries for Grasp Learning

Editors' Recommendations

Christian de Looper
Christian’s interest in technology began as a child in Australia, when he stumbled upon a computer at a garage sale that he…
Google’s AI image-detection tool feels like it could work
An AI image of the Pope in a puffy coat.

Google announced during its I/O developers conference on Wednesday its plans to launch a tool that will distinguish whether images that show up in its search results are AI-generated images.

With the increasing popularity of AI-generated content, there is a need to confirm whether the content is authentic -- as in created by humans -- or if it has been developed by AI.

Read more
Google’s ChatGPT rival is an ethical mess, say Google’s own workers
ChatGPT versus Google on smartphones.

Google launched Bard, its ChatGPT rival, despite internal concerns that it was a “pathological liar” and produced “cringeworthy” results, a new report has claimed. Worker say these worries were apparently ignored in a frantic attempt to catch up with ChatGPT and head off the threat it could pose to Google’s search business.

The revelations come from a Bloomberg report that took a deep dive into Google Bard and the issues raised by employees who have worked on the project. It’s an eye-opening account of the ways the chatbot has apparently gone off the rails and the misgivings these incidents have raised among concerned workers.

Read more
Google Bard vs. ChatGPT: which is the better AI chatbot?
ChatGPT versus Google on smartphones.

Google Bard and ChatGPT are two of the most prominent AI chatbots available in 2023. But which is better? Both offer natural language responses to natural language inputs, using machine learning and millions of data points to craft useful, informative responses. Most of the time. These AI tools aren't perfect yet, but they point to an exciting future of AI assistant search and learning tools that will make information all the more readily available.

As similar as these chatbots are, they also have some distinct differences. Here's how ChatGPT and Google Bard measure up against one another.

Read more