Citrini Research says AI-enabled robotics is emerging as a major investment trend that could extend the artificial intelligence boom beyond software and data centers, with China’s rapid progress in humanoid robots raising the prospect of a new technology race with the United States.
The research firm, which gained widespread attention for publishing a hypothetical AI-driven economic doomsday scenario, has developed a more bullish thesis around robotics, arguing that physical machines capable of using sophisticated AI systems could become the next growth phase of the AI trade.
Citrini pointed to the World Humanoid Robot Games, held in Beijing in August, as evidence of how quickly the technology is progressing and how China is using public competition to accelerate development.
Register for the next Tekedia Mini-MBA.
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
“Humanoid Olympics may seem like an exercise in the absurd, but really are a top-down strategic effort to push the limits of hardware in a competitive, visible and verifiable setting,” Citrini wrote. “Sending a humanoid at full sprint into a crash wall is literally ‘moving fast and breaking things.’ Scoff at your own risk.”
The event is exposing another shift in the development of artificial intelligence. The first phase of the AI investment cycle was dominated by chips, cloud computing and large language models. The emerging phase is focused on putting those systems into machines that can operate in factories, warehouses, homes, and other physical environments.
The concept, described as “physical AI,” is already attracting technology companies and investors. Atom, the physical automation startup founded by Uber co-founder Travis Kalanick, is expanding rapidly, while UBS strategist Ulrike Hoffmann-Burchardi has highlighted robotics stocks as an area that could benefit from the continued expansion of the AI trade.
Citrini’s argument, however, goes beyond the size of the potential robotics market. The firm says US investors may be underestimating China’s position in the emerging industry.
The report acknowledged that the US could retain an advantage in software and AI decision-making systems, but argued that China has important strengths on the hardware side. Those strengths include a large manufacturing base and established supply chains that could become important if humanoid robots move from experimental prototypes to mass-market products.
That possibility has led Citrini to compare the emerging robotics industry with the electric vehicle market.
“It’s running the EV playbook, with the literal same players,” the firm wrote. “Tesla popularized the modern EV and Chinese companies like XPeng replicated.”
The comparison strengthened the message because the EV industry demonstrated how technological leadership and manufacturing dominance can evolve differently. Tesla helped popularize modern electric vehicles, but Chinese manufacturers subsequently expanded aggressively, competing on cost and production scale.
BYD has become one of the world’s largest electric vehicle manufacturers, while Chinese automakers have been challenging established global brands on pricing and production. The companies are now carrying that industrial competition into robotics.
In June 2026, BYD’s CEO disclosed plans to begin incorporating robots into every showroom to assist with car sales. XPeng has also expanded beyond electric vehicles with XPeng Iron, its own humanoid robot designed to compete with Tesla’s Optimus.
For Citrini, XPeng provides an example of how an established Chinese EV company could potentially use its existing manufacturing capabilities to participate in a future robotics market.
“XPeng is another EV company with a nascent arm but ~100x cheaper,” the author wrote. “I would bet Optimus is in a better position than Iron today, but just look at history. Xpeng’s EV business is rapidly expanding in part by taking share from Tesla.”
The firm is therefore not arguing that XPeng currently has a technological advantage over Tesla in humanoid robotics. Instead, its thesis is that today’s lead may not determine tomorrow’s market structure if robotics follows the same development pattern as electric vehicles.
That stance is of the essence because humanoid robotics is fundamentally a hardware business as well as an AI business. A company needs capable AI models, but it also needs motors, sensors, batteries, actuators, semiconductors, manufacturing capacity, and reliable supply chains. The ability to produce robots at scale and at a commercially viable cost could become as important as the intelligence of the systems controlling them.
Therefore, China’s existing industrial infrastructure is expected to become a significant factor if demand for humanoid robots accelerates.
But there are still substantial hurdles before the technology reaches that stage. Demonstrations and competitions can show that robots are becoming more capable, but commercial adoption will depend on whether machines can perform useful tasks reliably, safely, and at a cost that makes economic sense for businesses and consumers.
That has created a market with both significant potential and considerable uncertainty for investors. The current AI boom has been driven largely by spending on computing infrastructure and software. A successful transition into physical AI could create another layer of demand across robotics, industrial automation, semiconductors, batteries, and related manufacturing industries.
Citrini’s broader warning is that the beneficiaries may not be the same companies that dominate the current AI software race. If robotics becomes the next major application of artificial intelligence, the competitive advantage could shift toward companies that can combine advanced AI with low-cost, high-volume hardware production.
The EV industry provides a precedent for that shift. Tesla demonstrated the commercial potential of a new vehicle technology, while Chinese manufacturers later built substantial scale and competed aggressively on cost.
Citrini’s thesis is that humanoid robotics could follow a similar trajectory, making the emerging contest between US AI capabilities and China’s hardware and manufacturing base one of the important questions for the next stage of the AI economy.