If you always stay in the third tier, the time left for you in embodied intelligence won't be long.

If a company remains stuck in the third tier of embodied intelligence indefinitely, time is running out for it.

This conclusion is based on the evolution of the track pattern and the iteration of capital logic. It started in early 2025 when the embodied intelligence track had seen the rise of leading players and the initial formation of the pattern. At that time, the Embodied Intelligence Research Society and several front - line investors deeply involved in hard - tech asked: Can real dark horses still emerge in the track? If so, what kind of founders should these dark horses have?

At that time, investors from different institutions gave highly convergent judgments: Dark horses must exist, and they will most likely come from the core executives of the leading embodied intelligence companies currently heavily bet on by capital, those who are truly in charge of technology implementation, supply - chain integration, and scenario expansion at the front line.

Different from the logic of betting on "scientists" in 2023, if capital providers choose the "investment in people" logic again, the ideal candidates should have experienced the technology iteration from scratch, encountered the pitfalls of large - scale implementation, and have a deep understanding of industrial boundaries and market demands.

Looking back one year later, this prediction has been intensively verified. Under the surface of the embodied intelligence track, there is a new force spawned by executives' entrepreneurship. According to the Embodied Intelligence Research Society, some capital providers are persuading core talents in embodied intelligence companies to start their own businesses, and the starting capital amount is quite attractive to anyone.

However, new contradictions have become sharply prominent: The size of the capital cake is ultimately limited.

In 2026, the agglomeration effect of financing in the leading players of the track has become more and more significant. These "new targets" of executives' entrepreneurship under the surface have become the darlings of capital due to their luxurious teams and industrial resources. On one hand, there is the "siphon effect" of the leading players, and on the other hand, there is the "squeezing situation" of the new forces.

The third - tier companies in the middle have neither the technological and capital barriers of the leading players nor the industrial resources and flexible strategies of the new forces. Their living space has been continuously compressed.

This is not to deny the necessity of the current third - tier companies, but the reality that must be faced is: If they continue to stay in the third tier without making progress, and always maintain a significant gap with the second - tier companies in key dimensions such as technological breakthrough, commercial implementation, and industry influence, and are unable to find a differentiated way out, they may miss the historical dividends of the embodied intelligence industry.

The Evolution History of the "Opportunity Theory"

Looking back from a retrospective perspective, after the explosive growth and substantial implementation in 2025, it is hard not to find it incredible to look back at the enthusiastic discussions about "embodied intelligence opportunities" in early 2025.

This is by no means a one - sided narrative of pessimists, nor a misjudgment from a narrow perspective, but a cautious exploration, or even self - questioning.

In early 2025, embodied intelligence, which was originally hidden behind the industry, was suddenly pushed into the spotlight due to a national - level appearance and was instantly given the context of "the next - generation technological revolution" and "a certain correctness." The slogan of the "Year of Embodied Intelligence" resounded outside. However, under the hustle and bustle, the foundation of the track was just being laid: Technologically, core bottlenecks such as insufficient generalization ability, the embodied gap, and the difficulty of Sim2Real implementation had not been broken through. Most products were still in the stage of "specific scenario demonstrations." Commercially, there was a lack of a stable profit model, and the situation of "looking for nails with a hammer" had become a common dilemma. There were very few large - scale scenarios that could be truly implemented.

The massive amount of traffic and capital poured in within a short period of time did not seem to be the nutrients for "ripening" the industry at that time. Instead, it worried most practitioners. This was more like an irrational "flattering to kill."

When a group of start - up companies with unformed technologies and vague business logics, seemingly "falling apart with just a gentle breeze," appeared at the same time as large - scale funds of hundreds of millions or even billions, this seriously unbalanced ecosystem inevitably made people question: Is this another standard scenario of "blowing bubbles" in the technology industry?

So at that time, not to mention the "opportunity theory," there was even a group of people discussing "withdrawal."

After one year, the opportunity theory of embodied intelligence has completed a complete evolution.

The track that was once regarded as having "false hype" has gradually developed verifiable implementation scenarios under the leadership of leading players; the business models that were once questioned as "castles in the air" have begun to be implemented on a large scale.

The capital that once chased concepts has also curbed its enthusiasm and turned to a hardcore assessment of technology, supply chain, and commercialization capabilities. Investors no longer look at the technical parameters on PPTs but require on - site demonstrations of practical scenarios such as "unscrewing a mineral water bottle." It is no longer "profiting by betting on the trend," but "selecting high - quality players carefully to achieve long - term success"; it is no longer "anyone who catches the concept can get a share," but "only a few who can truly solve industry pain points and have core barriers can reach the end."

Now it can be seen that few capital providers or other industry insiders publicly talk about the embodied intelligence bubble. More voices are calling for "patience." This long - slope and thick - snow track finally has a more appropriate development rhythm.

At the current stage, companies with more advanced technologies and engineering capabilities have formed a stable pattern, and it is not a simple matter of just buying a ticket to get on board. Moreover, the price of this ticket has exceeded the affordability of most capital providers. In contrast, the second - tier companies still have high growth potential and relatively friendly investment costs. Most capital providers still believe in and wait for the second - tier companies to break through.

It is very difficult for third - tier companies to be seen and selected. An investor once told the Embodied Intelligence Research Society that he would rather wait for new players with differentiated characteristics to appear.

It can be seen that the opportunities in embodied intelligence have completely flowed to those who are more worthy of investment.

Three Portraits Promoting the Shuffle

If the living space of the third - tier companies is being extremely compressed, then a reality that must be questioned is: Who is snatching the cake that originally belonged to them?

The answer is not outsiders who appear out of nowhere, but a group of new players who have naturally emerged as the industry has entered the "deep - water area." Different from the preachers who often promoted the "ultimate form of general artificial intelligence" in 2023, in 2026, the entrants who can get the "new tickets" from capital providers have extremely specific and practical portraits.

The ones with the greatest opportunities are the executives of embodied intelligence companies. This is the most powerful force at present. As mentioned above, the "investment in people" criteria of front - line investors have changed. They not only look at academic qualifications but also at "scars." Core executives who have encountered difficulties and personally handled mass production and delivery in leading embodied intelligence companies are becoming the darlings of top - tier VCs.

Why are these people popular?

Because they know what "won't work." In the past two years, these former technology experts, hardware or supply - chain managers have experienced all the hardships from PPT blueprints to hand - made engineering prototypes and then to small - batch trial production. They know which upstream suppliers have extremely low pass rates and understand which scenarios at the current stage are purely "pseudo - demands."

When these executives lead their old teams to start their own businesses, they have no historical burdens. Moreover, they are unlikely to touch the story of "all - powerful general humanoid robots" again and will find a more comfortable and practical niche in the embodied intelligence industry.

For capital, the "trust cost" of such teams is extremely low: Investors do not need to question their technical abilities (backed by the implementation cases of leading companies), do not need to worry about their supply - chain resources (with established cooperation relationships), and do not need to doubt their commercialization capabilities (with customer resources as a foundation).

Therefore, many executive - led start - up projects can obtain hundreds of millions of yuan in angel - round financing just based on "team background + preliminary business plan," while ordinary start - up companies may need to develop mature products first to get a small - scale seed - round investment. This "starting advantage" with capital support further amplifies their opportunities.

If the entrepreneurship of executives can be regarded as a differentiation within the same industry, then the cross - border players from large companies are completely overturning the table.

Embodied intelligence is essentially about intelligence, which is exactly the logic that new - energy vehicle manufacturers and Internet giants are playing. In the past year, a large number of engineering veterans with automotive manufacturing backgrounds have entered the track. The greatest killing power of these veterans lies in their absolute control over the supply chain. On the other hand, Internet giants are no longer content with just making peripheral investments and have begun to deeply layout embodied intelligence. There is also a possibility that they will establish independent subsidiaries in the future. If they do so, it will also be a force that cannot be underestimated.

The last group of new forces that cannot be ignored are the new - generation geeks from prestigious universities who are "born with embodied intelligence."

Among the current leading players in the embodied intelligence field, there are many founders with prestigious university backgrounds or scientists who have made great achievements in academia. In essence, these scholars are the "wave - makers" of embodied intelligence. What they learned in academia was still "robot learning" rather than embodied intelligence.

However, the current academic representatives in universities are conducting research with "embodied intelligence" as the top - level architecture. They are scientists who are "born with embodied intelligence."

This group of reserves is more in line with the narrative of embodied intelligence and has a new - era style, focusing on "facing reality and giving up the game." They prefer to take the small - but - beautiful product route. For example, they focus on consumer - end elderly - care robots, using precise scenario positioning to make up for the short - comings in the supply chain and engineering. Although they still need to improve their mass - production capabilities, they are also diverting some opportunities in niche areas with their advanced algorithm advantages and precise scenario entry, becoming another "cake - snatcher" for the third - tier companies.

Putting these three portraits together, we can see the real game rules of the current embodied intelligence track: Resources are accelerating their flow direction.

This also explains why it is said that time is running out for companies that always stay in the third tier.

If a company neither has the experience of avoiding risks like the teams of executive - led start - ups nor can show the supply - chain integration ability of cross - border veterans, and still clings to the obsession of "making a large robot" and is unwilling to do profitable things. Then, under the double squeeze of the siphon effect of the leading players and the new forces under the surface, this kind of high - and - dry hard - support will sooner or later exhaust the last bit of cash on the company's account in a series of exhibitions that no one cares about.

This article is from the WeChat official account "Embodied Intelligence Research Society". Author: Lv Xinyi. Republished by 36Kr with permission.