Skip to Content
Artificial intelligence

Facebook is now dabbling with robots to help make its AI smarter

One of Facebook's robots.
One of Facebook's robots.Facebook

Facebook’s efforts to build intelligent helpers show how badly AI needs some sort of grounding in the physical world.

Robots rising: Facebook is showing off a new robotics lab at its Silicon Valley headquarters, along with several research projects that involve robots learning how to do different things. So is the company looking to add arms and legs to its Portal devices to let them explore your home? 

Advancing AI: Nope (luckily). The project is actually a very natural extension of the company’s existing AI research. Machine learning can be used to teach robots how to do things that would be impossible to program. So the project might have practical applications somewhere down the line. More important, though, the way robots learn about the world can feed back into the algorithms applied elsewhere.

Curiouser and curiouser: As with Facebook’s other AI research, the robotics team is focused on developing algorithms that learn with minimal human supervision. An interesting aspect of their approach is a kind of machine curiosity algorithm that lets the robots learn without trying to achieve a specific end goal.

Real experience: This is important when you consider Facebook’s ongoing efforts to build a virtual helper of some kind. Simply training an AI algorithm on conversational text isn’t enough to create something capable of having a decent chat. As experts will tell you, that’s partly because those algorithms lack the understanding of the physical world that we take for granted.

Bio-inspired: “The only example of intelligence we have is humans or animals,” says one of Facebook’s AI research scientists, Roberto Calandra. “If we want to re-create the kind of intelligence we have in humans and animals, it seems very plausible that we might need that physicality.”

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.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

Providing the right products at the right time with machine learning

Amid shifting customer needs, CPG enterprises look to machine learning to bolster their data strategy, says global head of MLOps and platforms at Kraft Heinz Company, Jorge Balestra.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.