The ChatGPT Moment for Robots is here 🤖

Remember where you were when ChatGPT launched in November 2022? Within 5 days it had 1 million users. Within two months, it had 100 million users. 

The technology wasn't born overnight, large language models had been brewing in labs for years, but something clicked. The interface was right, the timing was right.

And suddenly, everyone understood.

For the past three years, AI has been building the BRAIN: language, reasoning, vision, decision-making. Now, that brain is getting a BODY, and the body is learning fast, although most people haven't noticed yet.

You can now buy robots directly from Amazon:

The Numbers No One Is Talking About

In 2025, global humanoid robot shipments surged over 500% year on year. Somewhere between 13,000 and 18,000 humanoid robots were shipped worldwide. That number might sound small until you remember that the entire installed base was basically zero three years ago. 

Bank of America projects these shipments will reach 1 million units annually by 2030, with an 88% compound annual growth rate over the next decade. The humanoid robot market, valued at roughly $2–3 billion in 2025, is projected to hit $4–5 billion this year and could exceed $50 billion by 2035.

But the story isn't just humanoids. Amazon quietly crossed one of the most staggering milestones in industrial history last year: it deployed its one millionth robot. One million.

Across more than 300 facilities worldwide, Amazon now operates the largest fleet of industrial mobile robots on the planet. Its newest fulfillment center in Shreveport, Louisiana, 3 million square feet, five floors, is essentially a robot-first building where humans supervise rather than carry. 

The company says 75% of its global deliveries are now assisted by a robot in some way. And here's the kicker: Amazon's robot count is approaching parity with its human workforce of 1.56 million. The packages handled per worker jumped from 175 in 2016 to nearly 3,900 in 2025. Amazon expects automation to save $12.6 billion between 2025 and 2027.

This isn't a pilot program.

Why 2026 Is the Year Everything Changes

Here's what makes 2026 different from every previous year of robot hype:

The foundation models have arrived. Just as GPT-3.5 unlocked a new paradigm for text, Vision-Language-Action (VLA) models are doing the same for robotics. NVIDIA's Isaac GR00T N1, launched publicly with live demonstrations, allows robots to learn tasks without extensive manual coding. The shift from "programmed to perform" to "learned through observation" is the single most important technical breakthrough in physical AI. Robots are no longer limited to the tasks they were explicitly built for, they can be taught new ones.

The economics are crossing over. With labor shortages reaching critical levels across the many countries and region, where job vacancies often exceed the number of unemployed individuals, the economic case for robots has flipped. It's no longer about whether robots are cheaper than humans. It's that there simply aren't enough humans to do the work. An aging global population and declining birth rates make this structural, not cyclical.

The hardware is finally ready. The shift from hydraulic to electric actuators, lighter materials, longer-lasting batteries, and compact processors means robots can now operate full shifts in real environments. Unitree's G1 just set a record for autonomous walking in extreme cold: 130,000 steps at minus 47.4 degrees Celsius. These are machines approaching industrial reliability.

The home is next. 1X Technologies' NEO opened consumer preorders in late 2025 at $20,000 (or $499/month), with deliveries starting this year. Figure 03 was named one of TIME's Best Inventions of 2025. We're not talking about another Roomba, we're talking about a robot that can fold laundry, load a dishwasher, and hold a conversation. Imperfectly, yes, but improving rapidly through real-world learning.

The Great Power Robot Race

If the ChatGPT era sparked an AI arms race between the U.S. and China, the robot era is following the exact same script.

🇨🇳 China is winning the opening round. Nearly 90% of all humanoid robots sold globally in 2025 were Chinese. Six of the top-selling humanoid companies were Chinese. Unitree Robotics shipped over 5,500 units last year; AGIBOTshipped over 5,100. These aren't lab curiosities, they're commercial products deployed in factories, logistics centers, and retail environments. 

At CES 2026 in Las Vegas, Chinese humanoid makers showed up in force, showcasing models that are cheaper, faster to market, and increasingly capable. Just a few weeks ago Unitree's robots performed martial arts on the CCTV Spring Festival Gala, China's equivalent of the Super Bowl, watched by hundreds of millions. The symbolism wasn't subtle: this is national strategy now.

China's playbook mirrors its EV strategy. Government-backed funding, deep manufacturing supply chains, aggressive pricing, and the willingness to deploy at scale before the technology is perfect. More than 150 humanoid robot companies now operate in China, backed by a Ministry of Industry and Information Technology plan to build a complete humanoid innovation ecosystem by the end of this year. Chinese firms have a cost advantage that's difficult to overstate: Unitree's G1 humanoid sells for around $16,000. Western competitors often start at six figures.

🇺🇸 The U.S. is playing a different game. American firms dominate in AI, software, and autonomy, the "brain" of the robot. NVIDIA's GR00T foundation model, OpenAI's renewed interest in embodied AI, Tesla's FSD-derived robotics stack, and Figure AI's $700 million in funding from Microsoft, NVIDIA, and Jeff Bezos all represent bets that the intelligence layer will ultimately matter more than the hardware. Tesla is converting its Fremont factory, previously home to the Model S and Model X, to produce Optimus humanoid robots, with stated ambitions of reaching millions of units annually by the end of the decade. 

🇪🇺 Europe is the quiet dark horse. While it lacks the sheer volume of China or the venture capital firepower of the U.S., Europe brings something neither competitor has: industrial depth. Germany's NEURA Robotics raised €120 million in early 2025, one of the largest European robotics rounds ever. The UK's Humanoid (formerly SKL Robotics) signed a partnership with Schaeffler, the German motion technology giant, to deploy hundreds of humanoid robots in its factories starting this year. Mercedes-Benz is running Apptronik's Apollo robots in its European plants. BMW is piloting Figure AI's robots on its lines. And Hyundai Motor Group, through its ownership of Boston Dynamics announced a new U.S. robotics facility with 30,000-unit annual capacity alongside a strategic partnership with Google DeepMind.

Europe's approach is less flashy but potentially more durable: modular collaboration, tight integration with real manufacturing ecosystems, and a regulatory environment that, while demanding, might provide the trust needed for widespread adoption.

The Hyundai Playbook: How You Scale Robots

If you want to understand what scaling robots actually looks like in practice, look at what Hyundai Motor Group just laid out at CES 2026. It's arguably the most complete blueprint any company has presented for taking humanoid robots from lab curiosity to factory reality and it involves three of the most formidable names in technology.

The foundation is Boston Dynamics, which Hyundai acquired nearly five years ago. At CES, the company unveiled the production version of Atlas, not the research prototype that's been doing backflips on YouTube, but a redesigned, production-friendly humanoid built for industrial work. Atlas can lift 110 pounds, has 56 degrees of freedom, 360-degree cameras, tactile-sensing hands, and every component has been designed for compatibility with automotive supply chains. All 2026 deployments are already fully committed.

The brains come from Google DeepMind. Boston Dynamics and DeepMind announced a strategic partnership to integrate DeepMind's Gemini Robotics foundation models into Atlas, aiming to create what they described as the most advanced robot foundation model in the world. The idea: a general-purpose brain paired with a highly capable generalist body, enabling robots to perceive, reason, and adapt rather than simply executing pre-programmed tasks.

This is the template. Hardware excellence (Boston Dynamics) plus AI foundation models (Google DeepMind) plus mass manufacturing and real-world deployment at scale (Hyundai). And Hyundai plans to push robotics beyond automotive into logistics, energy, construction, and facility management.

What This Means for the Rest of Us

Let's be honest about what's coming. Amazon's CEO Andy Jassy has said he expects the company's workforce to shrink in the coming years due to technological advancement. 

Amazon plans to replicate its robot-heavy Shreveport model across 40 facilities by the end of 2027, expecting to employ about half as many workers at those sites as it would without automation. 

It's a productivity and capability story. Countries and companies that embrace robotics will be able to sustain economic output despite shrinking workforces. Companies that deploy robots effectively will see margins expand and delivery times shrink. And entirely new job categories; robot supervision, fleet management, AI training, remote operation are already emerging.

The Bottom Line

The ChatGPT moment wasn't really about one product launching. It was about a technology crossing the threshold from "interesting research" to "obviously useful to everyone." 

Humanoid robots crossed that threshold in 2025. In 2026, we're watching the deployment curve begin to steepen: from thousands of units to tens of thousands, from pilot programs to production lines, from factory floors to (eventually) living rooms.

The race for the BODY is on. 

If you want to book me for a keynote or check out my new book ‘Scale vs Soul’, go to www.dadovanpeteghem.com