Generative AI is accelerating the development of humanoid robots. What it means for the workplace

Shenzhen-based LimX Dynamics shows off one of its humanoid robots. Limx Dynamics

ChatGPT Such artificial intelligence is accelerating research and bringing humanoid robots closer to reality in China, where many of the world's factories are located.

AI has been around for decades. What has changed with the advent of OpenAI's ChatGPT chatbot is AI's ability to better understand and generate human-like content. While the American technology is not officially available in China, local companies such as Baidu have already released similar chatbots and AI models.

According to Li Zhang, chief operating officer of Shenzhen-based LimX Dynamics, in robotics, the development of generative AI can help machines understand and perceive their environment.

About three months after joining the two-year-old startup, Lee said he had scaled back his expectations for how long it would take LimX to create a humanoid robot that could not only work in a factory, but also help with household chores.

Lee originally expected the whole process to take eight to 10 years, but now believes some use cases will be ready in five to seven years. "After working for a few months, I've seen how the capabilities of various tools have improved thanks to artificial intelligence," he said in a mandarin translated by CNBC.

"It has accelerated our entire R&D cycle," he said.

Companies are rushing to capitalize on this opportunity. OpenAI itself is supporting startups to create humanoid robots, and Elon Musk's Tesla is developing its own robot called Optimus.

According to PitchBook, electric car giant BYD last year invested in Shanghai-based Agibot just months after it was founded.

And in November, Chinese state media published a photo of Chinese President Xi Jinping examining a humanoid robot at an exhibition center during his first trip to Shanghai since the pandemic. The robot was developed by Fourier Intelligence.

Before humanoid robots show up in homes, as LimX eventually plans, factories could be a profitable closed scenario for their introduction.

According to the latest Stanford's AI Index report, China overtook Japan to become the world's largest producer of industrial robots in 2013, and now accounts for more than 50% of the world's robots.

According to the report, the top three sectors in which industrial robots are installed in China are electronics, automotive, metallurgy and machinery.

Impact on people's workplaces

However, when it comes to completely replacing humans, AI advances alone are not enough.

Even if AI allows a robot to think and make decisions on par with humans, mechanical limitations are the main reason why humanoids can't yet replace humans, LimX's Lee says.

One of LimX's backers, Future Capital, has also invested in Pan Motor, a company that specializes in motors for humanoids.

Generative artificial intelligence won't help robots directly, notes Eric Xia, partner at Future Capital, a LimX investor. But "advances in large language models can help humanoid robots with task planning," he said in Chinese, translated by CNBC.

Other investors in LimX include Lenovo Capital.

The transition to factory robots may accelerate once the cost per robot comes down.

Steve Hoffman, chairman of startup gas pedal Founders Space, said he is working with Chinese startup Fastra, which he predicts could start mass-producing robots in as little as a year. He said he spent time in China this year teaching local companies how to integrate generative AI.

"We have already received six orders from research institutes," he said, noting that the startup intends to reduce the cost per robot to $50,000 to $100,000 by the time sales begin.

"If we can get to a $50,000 price point, we can sell a lot of robots," he said, noting that the robots' batteries can be recharged while in operation, 24 hours a day. "We can pay off a robot in a year."

In pharmaceutical research, generative AI can reduce costs without reducing human labor.

"In our business, you don't save on costs by having fewer people. You actually save money by having fewer failed experiments," says Alex Zhavoronkov, chairman, CEO and executive director of Insilico Medicine, which has offices in Hong Kong, New York and other parts of the world.

He noted that large pharmaceutical companies typically have to spend thousands of dollars to reproduce a molecule for testing - and run several thousand such tests per program. He claims that with Insilico's artificial intelligence, only about 70 molecules per program need to be synthesized.

In March, the company published a paper in the journal Nature claiming to have reached a Phase II clinical trial for an artificial intelligence-derived drug.