Owning a humanoid robot still comes with a price tag that puts most families off, anywhere from $40,000 to $100,000 for a full-size machine, and $2,000 to $8,000 for a robotic dog. Jerry Wang, Executive Chairman of Faraday Future and Co-CEO of its Nasdaq-listed subsidiary AIxCrypto Inc. (AIXC), thinks he has found a way around that barrier: let people rent a robot the way they’d rent a car through Uber or Lyft. “Very similar thoughts, if the car can be shared, the robot should be able to share as well,” Wang said.
The new platform, called RoboShare, is AIxCrypto’s answer to a problem the entire humanoid robotics industry is grappling with: high unit costs and low everyday utility. Rather than asking people to commit tens of thousands of dollars upfront, RoboShare lets them try a robot for a few hours, a day, or longer, before ever considering a purchase.
Wang’s example is entertainment. Hosting a birthday party, an anniversary, or a corporate event? For around $500, RoboShare customers can rent a humanoid for a couple of hours. “They can sing, it can dance, they can talk to the guests, they can provide some jokes, making people fully entertained,” Wang said. The pitch is simple: lower the cost of entry, let people experience a robot firsthand, and hope that a good experience nudges them toward eventually buying one, whether a full humanoid or a smaller robotic dog.
A Marketplace Built on Existing Owners
Like any sharing-economy platform, RoboShare depends on supply as much as demand. Wang expects that supply to come not only from Faraday Future’s own inventory, but from private owners as well. “If you purchase, no matter a humanoid or a robotic dog, you’re not going to be using it 24/7, right?” Wang said, noting that owners can list an idle unit for around $200 a day while they’re traveling or simply not using it. For Faraday Future, every rental is also a potential sales funnel: hobbyists, parents renting a smaller unit for a child’s education, or developers testing the platform’s open API and SDK could all become future buyers.
Renting out a humanoid isn’t as simple as handing over a set of keys. Today’s machines aren’t fully autonomous, and setting one up for an event still requires two trained staff to transport, calibrate, charge, and monitor the robot throughout its use. Off-the-shelf tasks like dancing work well because the movements are pre-recorded; anything beyond that, new routines, new tasks, still requires retraining, sometimes via motion-capture rigs and VR controllers. For now, Wang suggested, owning a full humanoid mostly makes sense for developers, engineers, or performance companies rather than the average household.
Liability is handled through dedicated insurance coverage for robotic devices, something Wang said isn’t trivial to source, combined with trained technicians present at every humanoid deployment. The priority, he stressed, is straightforward: “Most important thing is to not hurt people.”
Launch Markets and Early Signals
RoboShare is starting in the United States, with an initial focus on dense urban markets like Los Angeles before expanding to cities such as New York. Europe isn’t on the immediate roadmap; Wang cited differences in regulation and market experience as reasons to prove out the U.S. model first before considering international expansion.
Early pilot testing has already produced a notable signal: more than half of the people expressing interest say they’d rather rent a robot first than buy one outright, unsurprising given that a few hundred dollars for a day’s rental is a far smaller commitment than tens of thousands for a purchase.
Use Cases
Zooming out from RoboShare itself, Wang laid out three use cases where he believes physical AI is already mature enough to deliver value, even as broader capabilities continue to develop:
Education. Faraday Future is running robotic summer camps in Los Angeles in partnership with public and private education institutions, multi-week programs introducing kids to hardware, software, cloud connectivity, and AI fundamentals. The company also plans to offer ongoing course subscriptions (around $399 annually) tied to companion robotic dogs with programmable, open APIs.
Safety and surveillance. A camera-equipped robotic dog, priced around $5,000, combines front-facing facial recognition with 360-degree monitoring and lidar-based patrolling. It can recognize family members, issue warnings to unfamiliar visitors, escalate to an alarm and police notification if someone continues to approach, and integrate with existing camera systems. Wang sees applications ranging from residential yards to overnight monitoring of warehouses and factories, including risky scenarios like checking on a fire without sending a person into danger.
Industrial logistics. Rather than legged robots, Faraday Future is deploying wheel-based humanoid platforms for manufacturing and final-assembly environments, where stability and lower cost outweigh the need for legged mobility.
On the topic of legs versus wheels more broadly, Wang confirmed that wheel-based designs, cheaper, simpler, and more mechanically reliable, are increasingly the pragmatic choice for industrial deployment, even as legged humanoids remain the more visible, flashier form factor.
Chinese Hardware, U.S. Software
Faraday Future’s underlying approach, which Wang calls a “bridge strategy,” leans on partnerships with established Chinese robotics manufacturers for hardware. “We do less on the hardware,” Wang explained. “Instead we partner with the best China robotic companies to bring their best-selling power and mature parts into the United States.” Faraday Future itself handles final assembly, U.S. regulatory certification, and the software, AI, connectivity, and data layer. All data is kept on Faraday Future’s own servers under U.S. control, addressing potential privacy and regulatory concerns.
Asked to grade the state of physical AI, Wang put hardware at roughly 50–60% of where it needs to be, but software and real-world task capability at only 10 to 20 percent. He illustrated the gap with a simple example: asking a robot to “bring me a coffee” requires it to understand what coffee is, locate or prepare it, decide on the right tool and sequence of actions, and physically manipulate objects of varying fragility and weight, a far more complex chain of decisions than autonomous driving, which Wang described as operating along essentially “four degrees of freedom: front, back, left, and right.”
The scale of the data problem is stark by comparison. “Tesla have millions of cars as device to generate data,” Wang said, contrasting that with the robotics industry’s much smaller footprint: “How many robots do we have? Like 10,000.” Closing that gap, he said, will take considerable time, computing power, and training data, but for now, platforms like RoboShare are betting that letting more people simply try a robot is itself a meaningful step forward.
