When humanoid robots turn science fiction into a $5 trillion business

(Dan Tri) - From science fiction to reality, humanoid robots have entered life, opening up a $5 trillion race between giants and a future of work that is being reshaped.

For decades, robots have been synonymous with giant, millimeter-precise mechanical arms in auto factories. They are tireless, industrious workers, but they are “imprisoned” by a single task. A welding robot will only ever weld. A picker robot will never learn to pack. They are specialized, powerful tools without a soul.

This year, that picture is changing dramatically. We are on the cusp of a new era where robots are no longer tools but versatile partners. The convergence of four key elements – foundational AI, humanoid hardware, agile supply chains and advanced manufacturing platforms – is creating a perfect storm that is blowing old stereotypes away.

The question is no longer “where will robots replace humans?” but “how will robots collaborate, adapt, and work with humans to solve humanity’s greatest challenges?” The next industrial revolution is here, and this time, it has a human face.

Thinking AI brain and humanoid body ready for action

At the core of this revolution lies artificial intelligence (AI). Whereas previously each robot needed a separate program, now underlying AI models are giving them a form of “general intelligence”.

The turning point came earlier this year when Nvidia introduced the Isaac GR00T model, a system that allows robots to learn new skills without having to reprogram them from scratch. Similar to how ChatGPT can handle a multitude of language tasks, GR00T helps robots learn behavioral patterns and apply them to a variety of contexts. Boston Dynamics’ Atlas robot, once known for its acrobatic feats, can now walk, grasp, and manipulate objects using a single AI model.

But a brilliant mind still needs a body to interact with the world. And this is where humanoid robots shine. Once considered expensive and impractical science fiction, the human form is now their biggest advantage. They can walk up stairs, use tools designed for humans, and operate within existing infrastructure without modification.

Companies like Apptronik (which recently raised $350 million for its Apollo robots) and Agility Robotics (which puts robots in e-commerce warehouses) are demonstrating the viability of this model in solving pressing problems like labor shortages and boring, repetitive tasks.

In particular, Google DeepMind has just revealed an amazing step forward with the Gemini Robotics project. They have created a two-part AI system, including Gemini Robotics-ER - a thinking model capable of analyzing a complex request such as sorting clothes and outlining the steps to take in natural language and Gemini Robotics 1.5 - an action model that receives instructions from the thinking model and converts them into precise physical actions.

Kanishka Rao of DeepMind sums up the idea simply: “One of the biggest advances is the ability to think before you act.” This allows robots to have flexibility and improvisation that was previously impossible.

Google DeepMind has just launched the first "thinking" AI robot (Photo: Google).

The $5 trillion gold rush and the global race

These technological breakthroughs are not just technical marvels, they are also the spark for a massive “gold rush.” A report from investment bank Morgan Stanley paints a stunning economic picture: the market for humanoid robots could surpass $5 trillion by 2050, double the size of the current auto industry.

Analysts predict the number of humanoid robots could reach nearly 1 billion units by mid-century, with 90% serving in industrial and commercial sectors.

“This will be a fairly gradual process until the mid-2030s, but then accelerate dramatically in the late 2030s and 2040s,” said Adam Jonas, Head of Global Automotive Research at Morgan Stanley.

The main driver of this boom is the sharp decline in prices. Morgan Stanley predicts that the cost of a humanoid robot will fall from around $200,000 today to just $50,000 by 2050 in developed countries. In countries with low-cost supply chains like China, it could even fall to $15,000.

As prices fall, owning robots is no longer the exclusive domain of large corporations. By 2050, an estimated 10% of American households could own a robot assistant. Venture capital is also pouring into the sector, with Pitchbook predicting it will exceed $20 billion a year by 2026. Investors are no longer betting on a distant future, but are pouring money into solving very real problems of today.

But the race is not all rosy. It is shaping up to be a new geostrategic competition, and for now, China is clearly in the lead.

“The national support for AI in China is greater than anywhere else,” said Sheng Zhong, head of industrial research at Morgan Stanley. “China’s advantage may have to expand further before rivals, including the US, really take notice.”

While American companies may lead the way in design, China dominates the supply chain. Most robotics developers around the world, including those in Silicon Valley, still rely on China and other Asian countries for key components like screws, gears, and motors. As Jonas warns, “The US will need to make major changes in manufacturing capacity, education , and national policy to remain competitive.”

Battlefield of the Giants: Hardware vs. Software

This $5 trillion race is also where tech giants demonstrate completely different business strategies, reminiscent of the battle between Apple and Google in the smartphone era.

Following Apple’s path, Tesla chose a vertically integrated model. They designed and manufactured everything from hardware to software for the Optimus robot themselves. Tesla’s advantage lies in its mass production capacity and experience collecting data from millions of self-driving cars. They believe that with enough real-world data, their robots will learn everything.

Meta, on the other hand, is betting on a strategy similar to Google's Android. "I don't think hardware is the biggest barrier," Meta CTO Andrew Bosworth frankly admitted. "The bottleneck is software." Meta is developing a robotics software platform and plans to license it to other manufacturers.

They don’t aim to be the biggest robot maker, but rather the provider of “brains” for millions of robots around the world. This is considered Meta’s “AR-sized bet,” indicating the billion-dollar investment in this project. Bosworth also did not hesitate to point out the weakness in the competitor’s strategy: “I understand how Tesla has enough data for cars. But I don’t see where they will get the robot data.”

Google is in a unique position to become the “intelligence provider” to the industry. With Gemini Robotics, they’re not just creating a product, but a foundational technology that can be integrated into any type of robot. This approach allows them to partner with multiple hardware manufacturers, making them an indispensable force in the robotics ecosystem.

The clash of these strategies will determine the future of the industry: Will the world have a closed, tightly controlled ecosystem like Apple, or an open, diverse ecosystem like Android? The answer will reshape not only the tech industry but the global economy.

The humanoid robot market is forecast to surpass the $5 trillion mark by 2050, a figure that could double the size of the current automotive industry (Photo: Xpert.digital).

Challenges on the road to the future

Despite the promising outlook, the road to popularizing humanoid robots is still thorny.

Safety and reliability: A robot working in a warehouse or hospital must be absolutely safe. Any serious failure could delay its deployment by decades.

Energy Efficiency: A humanoid robot that consumes as much energy as a refrigerator would be economically and environmentally unsustainable.

Workforce Integration: This is the biggest challenge. The key to success is not replacement, but collaboration. Work processes need to be redesigned so that humans focus on tasks that require creativity, judgment, and emotion, while robots take on the heavy, repetitive, or dangerous tasks.

Legal and ethical framework: As robots become more prevalent in public places, issues of liability in case of accidents, data privacy, and ethical standards need to be clearly defined.

Despite the challenges, one thing is certain: robots will be at the heart of the future of work, industry, and society. In Shenzhen, China, the world’s first Robot Mall has opened, where humanoid robots sell goods, make coffee, and interact with customers. It’s not just a technological showcase, but a slice of the future that’s coming to life.

This revolution is not a distant vision. It is happening right now, in Google labs, Tesla factories, and on the planning boards of venture capitalists.

The next decade will be defined not by replacement, but by collaboration between humans and machines. From the assembly line to the hospital, from the farm to the living room, these versatile partners will quietly but powerfully reshape our world.

The next industrial revolution is here and it has a very human face.

Source: https://dantri.com.vn/kinh-doanh/khi-robot-hinh-nguoi-bien-khoa-hoc-vien-tuong-thanh-thuong-vu-5000-ty-usd-20250930112655712.htm