Meta and Tesla are competing to lead the humanoid robot market using AI-powered devices

Mark Zuckerberg stepped on stage with AI glasses, but what looked like a gadget reveal is now tied to a much bigger story.

This is about Meta and Tesla heading into a direct fight over humanoid robots, according to the Wall Street Journal.

For years, Zuck and Elon Musk circled each other with social media barbs, competing platforms, and even a teased cage fight that never happened. This time the clash could move from the internet to the world of machines built to act like people.

Up until now, the rivalry was more about pride and money. When Meta launched Threads in 2023 to compete with Elon’s X, that seemed like the loudest move. But humanoid robots are different. They promise a new market, powered by artificial intelligence, where companies compete not just for users but for control of the future of work and life.

Meta uses AI glasses to build a robot future

The newest Meta glasses went on sale Tuesday. They come with cameras inside the lenses and a screen that shows video right in front of the eyes. The glasses can capture exactly what the user sees, and Meta has said that some of this audio and video could be used to improve its products. That means video data that can train machines.

On September 17, Meta tried showing off these glasses by having a chef cook for a party with help from the AI in the device. The demo crashed, but the message was clear. This technology was designed to watch people, learn tasks, and later hand those lessons over to robots.

Adam Jonas, an analyst at Morgan Stanley, told investors last week that Meta could have 20 million glasses in use within two years. That would be almost double the number of Tesla cars expected on the road by then.

He wrote, “Every Meta glasses user may be training a humanoid avatar iterated in simulation across billions of scenarios in a digital omniverse.”

That means even the most boring home videos, like making dinner, could become training clips for robots learning how to function in daily life.

Meta has also built Project Aria, another wearable designed to collect AI and robotics data. Andrew Bosworth, the company’s chief technology officer, said in June:

Tesla pushes Optimus to learn like humans

Tesla is leaning on its fleet of self-driving cars to create a bridge to humanoid robots. With more than eight million cars on the road, the company has already built up one of the largest pools of video data in the world. That footage trains Tesla’s computer vision system, and Elon has said the natural next step is robots.

The company’s robot, called Optimus, is being trained in new ways. In May, Milan Kovac, who was then a Tesla vice president, posted that “One of our goals is to have Optimus learn straight from internet videos of humans doing tasks… We recently had a significant breakthrough… and can now transfer a big chunk of the learning directly from human videos to the bots.” For Tesla, that means taking third-person video footage and teaching a machine how to copy it.

By contrast, Meta’s first-person glasses can record what people actually see as they move. That difference could give Meta an edge. But Elon has set his own scale. He predicted that by 2040, there would be 10 billion humanoid robots in the world.

Meta recruits experts as Elon sets bold numbers

Meta’s push into robots is still early, but it is moving. Earlier this year, Zuck hired Marc Whitten, former CEO of General Motors’ self-driving unit Cruise, to lead a new robotics group. He also brought in Sangbae Kim, an MIT robotics professor known for building a four-legged “cheetah” robot that could run at speed.

At the same time, Meta has been recruiting big AI talent, offering packages worth hundreds of millions of dollars. Zuckerberg calls this race for advanced AI “superintelligence,” and he says it will be the foundation of everything from wearables to robots.

On the “ACCESS” podcast earlier this month, Zuckerberg admitted the costs are huge but said speed matters more than money.

“If we end up misspending a couple of hundred billion dollars, I think that is going to be very unfortunate obviously,” he said, “but… if you build too slowly… then you’re just out of position on what I think is going to be the most important technology that enables the most new products and innovation and value creation in history.”

If you're reading this, you’re already ahead. Stay there with our newsletter.