Robots lack clear large-scale commercial applications and may take 20 years or longer to reach widespread everyday use, Fei-Fei Li, co-founder of World Labs and a member of the U.S. National Academy of Engineering, said on Sunday.

She shared her insights at a discussion themed "On the Eve of the Explosion of Spatial Intelligence" during the 2025 T-EDGE Conference with Jany Hejuan Zhao, founder and chief executive of NextFin.AI and chair and chief executive of TMTPost Group.

Li said the robotics industry remains at an early stage of development despite growing interest from investors and technology companies."Robotics is still very early," Li said. "There are very few real commercial use cases, especially in daily life."

Li said a key constraint is the lack of data needed to train and deploy robots.

"Cars generate massive amounts of data because people drive them every day," she said. "Robots do not have that. There are not enough robots operating in the real world to generate data at scale."

She said collecting robotic data is difficult and costly because robots often need other robots or specially designed environments to gather training data.

Li said progress in generative artificial intelligence, including video generation and simulation, could help accelerate development by providing synthetic data and training environments.

"Simulation and generative models can help," she said. "They offer a way to reduce reliance on real-world data, which is hard to obtain."

Li pointed out industrial robots have been adopted more quickly because they operate in controlled and repetitive environments. "Industrial robots are already in use," she said. "The scenarios are constrained, and data is easier to collect."

She compared robotics development to autonomous driving, which took nearly two decades to reach large-scale deployment.

"Google started self-driving research in 2006, and large-scale deployment came around 2024," she said. "That was almost 20 years."

She said robotics could follow a similar or longer timeline because the sector lacks mature supply chains and standardized commercial use cases.

"Robotics does not yet have the equivalent of a mature automotive industry behind it," she said.

Zhao said the industry faces two linked problems: lack of data and lack of application scenarios. "Without data, applications cannot grow, and without applications, data cannot accumulate," Zhao said.

Li said commercialization, not scientific potential, remains the main barrier. "The question is not whether robots are interesting," she said. "It is whether they can solve real problems at a reasonable cost."

She noted she remains optimistic but urged patience. "Robotics is a long-term effort," Li said. "It will not happen overnight."