Faraday Future (NASDAQ:FFAI) Global Executive Chairman Jerry Wang says Tesla Inc. (NASDAQ:TSLA) could realistically turn its Optimus humanoid into a commercial product next year, arguing the biggest barrier is no longer building the robot, but teaching it what to do.
“I think it’s very realistic,” Wang told Benzinga when asked whether Optimus could become a commercial product in 2027.
“The hardware is not a barrier right now. The barrier is software and training,” Wang said, pointing to Tesla’s computing power and access to data as advantages. “This will take time, but I think that they will get there.”
Faraday, best known for its electric vehicles, has shifted much of its focus toward robotics. Wang said the company has already delivered more than 500 robots across humanoid, quadruped and wheeled models this year.
Tesla, by comparison, has yet to begin selling Optimus to outside customers. A JPMorgan note last month said commercial sales of Optimus Gen 3 could begin in the second half of 2027.
Prediction traders remain skeptical of a faster rollout. Polymarket prices in about a 10% chance that Tesla makes Optimus available for public purchase by the end of 2026.
The Bodies Are Ready, the Brains Aren’t
Wang said robot hardware is much further along than the software needed to make robots broadly useful, rating hardware about 60 out of 100 and software just 10.
“The hardware as of now is pretty good,” he said, while developing software and training robots to handle different real-world tasks remains much harder.
The problem, Wang said, is partly a data bottleneck. Robots need experience from physical machines operating in the real world to learn what works and what does not.
“Without enough robots, you won’t have enough data,” Wang said. “Even failure is a great data; [a] robot will know what not to do.” He said getting more robots into real-world use should accelerate that learning process.
Tesla CEO Elon Musk has acknowledged that gap, saying during Tesla’s second-quarter earnings call that no humanoid robot can currently perform generalized tasks.
Useful Humanoids Under $20,000
Wang said humanoid robots could become normal purchases for businesses, and potentially households, within three to five years.
“Three years, probably, optimistically,” he said, adding that a useful humanoid could cost less than $20,000 by then.
Faraday already sells a range of robots, including full-size and compact humanoids as well as quadrupeds aimed at education, security and inspection.
Wang expects humanoids eventually to account for roughly half of the robotics market because homes and workplaces are designed around people.
But he said humanoids will not always be the best tool. For simple warehouse jobs, a wheeled robot with an arm can be cheaper and more efficient than a machine that also needs to walk and balance.
Musk Is ‘Right’ on Lidar, Wang Says
Wang also sided with Musk’s long-held skepticism of lidar for cars.
“I think on the automotive side, he’s right,” Wang said.
He said robotics is different because some robots need to map and navigate an area autonomously rather than simply rely on cameras. Faraday’s quadruped security robots use lidar to scan their surroundings, build a map and plan patrol routes, capabilities Wang said cameras alone cannot currently match.
He estimated lidar sensors now cost roughly $1,000 to $2,000, making them relatively inexpensive to add to a robot.
Image: Shutterstock
Login to comment