The world is just not quite ready for humanoids yet

Famed roboticist and iRobot founder Rodney Brooks has sounded the alarm on a humanoid robot investment bubble. He’s not the only one.

In a recent essay, Brooks calls out the billions of venture dollars being poured into humanoid robot companies like Figure. His take: despite the amount of money being injected into the industry, humanoids won’t be able to learn dexterity — or the fine motor movements with hands– rendering them essentially useless.

His take might surprise some, especially those VCs investing in the sector. But not to the multiple robotics-focused VCs and AI scientists, who have told TechCrunch in recent months that they don’t expect to see wide adoption of humanoid robots for at least a few years — if not more than a decade.

The issues

Fady Saad, a general partner at robotics-focused VC Cybernetix Ventures and former co-founder of MassRobotics, told TechCrunch that beyond sending humanoids into space in place of human astronauts, he doesn’t see a huge market yet.

“People who probably haven’t seen humanoids before, or haven’t kind of been closely following what’s happening, they are impressed with what’s happening now in humanoids, but we continue to be a little bit conservative and skeptical about the actual use case and the actual revenues that will be generated,” Saad said.

Saad is also concerned about safety, especially when humans and humanoid robots share the same space. Safety issues could arise from humanoids and humans working closely on a factory floor, or other industrial sites. Saad says those concerns mount when humanoids enter people’s homes — a goal many humanoid companies are working toward.

“If this thing falls on pets or kids, it will hurt them,” Saad said. “This is just one aspect of a big hurdle that no one is paying attention to, or very few people are paying attention to. The other thing is, how many people are comfortable with having a humanoid in their home sitting there? What if it got hacked? What if it went crazy at night and started breaking things?”

The timeline for this technology also isn’t clear — a crucial factor for VCs who have fund lifecycles and timelines to return capital to investors.

The timeline

Sanja Fidler, the vice president of AI research at Nvidia, told TechCrunch in August that while it’s hard to pin the development of humanoids to an exact timeline, she compared the current swell of interest to the excitement in the early days of self-driving cars.

“I mean, look at self-driving cars, in 2017 and 2016, I mean it felt tangible, right?” Fidler said at the time. “It still took them quite a few years to really scale and even now, no one really scaled to the entire world, full autonomy. It’s hard. It’s really hard to go and fully delivery on that technology.”

Nvidia chief scientist Bill Dally agreed in an interview with TechCrunch. Dally and Fidler’s comments are especially notable as Nvidia is also pouring money into developing the infrastructure for humanoid companies to follow.

Seth Winterroth, a partner at Eclipse, said while it can be easy to get excited as each new technological development happens, or the latest demo drops, humanoids are incredibly complicated. He added that it will be a while before they reach their full capabilities.

“It’s difficult to do software releases to six degrees of freedom systems, what we’re talking about with some of these humanoids is 60 plus degrees of freedom systems,” Winterroth said, regarding a robot’s ability to move on 3-D axis. “Then you need to be able to have good unit economics around that solution, such that you’ve got strong gross margin, such that you can be building an enduring business. I think we’re pretty early.”

In most cases, humanoid robots aren’t ready for the world yet, either.

Tesla is a great example of the struggles companies are running into. The company announced it was building its humanoid, Optimus, back in 2021. The following year, Tesla said the bot would be introduced in 2023.

That didn’t happen. When the bot was introduced in 2024 at Tesla’s “We, Robot” event — it was revealed later that the bots were largely being controlled by humans off scene. The company claims it will start selling the bots in 2026.

Robotics startup Figure, which was valued at $39 billion in a September fundraise, has also drawn skepticism regarding how many of its humanoids the company has actually deployed, a claim the company staunchly defends.

What is working

That doesn’t mean humanoids won’t have a future market or that the technology is not worth working on.

Brooks himself said he doesn’t doubt that we will have humanoids in the future. But instead of what the market pictures when they hear humanoids, a robot with a human form, he predicts they’ll likely have wheels and other inhuman features and won’t be coming out for more than a decade.

There are startups working on the dexterity technology Brooks is skeptical humanoids will be able to reach, including Y Combinator-backed Proception and Loomia, which built a kit that can help robotics companies start to incorporate touch into their machines.

There are also numerous humanoid companies that are starting to take orders and gather interest in their robots. K-Scale Labs received more than 100 pre-orders for its humanoid bot in the first five days, surprising even the founders, CEO Benjam Bolte told TechCrunch.

Hugging Face has also seen strong demand from developers for its two humanoid bots. The company opened up pre-orders for its smaller desktop version, the Reachy Mini, in July. The reaction was palpable. Just five days after opening up orders on its Reachy Mini robots, Hugging Face had logged $1 million worth of sales.