CITIC Construction Investment: Data Gaps Drive Capital Tilt, Asset Visualization Lays Long-Term Barriers

The Zhitong Finance App learned that CITIC Construction Investment released a research report saying that physical AI is taking over and becoming the core variable for the next phase of AI growth. As a key engine from perception to feedback, physical world data faces a huge gap in demand. Constrained by the “impossible triangle” of fidelity, scale, and cost, data acquisition in the physical world is expected to usher in a triple transformation: the business model side collects data through social crowdsourcing, the technical paradigm side relies on virtual manufacturing of the world model, and the asset attribute side has leapt into the enterprise's high-premium core intangible asset. The imbalance between supply and demand drives the revaluation of physical data assets, and capital is being concentrated in data infrastructure and evaluation at an accelerated pace. The market pattern is yet to be determined, and companies that control scarce data and closed-loop capacity are expected to experience core premiums.

CITIC Construction Investment's main views are as follows:

The value of data in the physical world is high, the gap is large, and the trend of capital inclination is accelerating

Physical AI is becoming a core new variable in the next phase of AI growth. As a key engine that runs through perception and feedback and empowers the center of intelligence, physical data faces serious supply bottlenecks due to high threshold restrictions. The rigid contradiction between supply and demand is driving the revaluation of data assets, and capital is pouring into data and evaluation infrastructure at an accelerated pace, spawning highly valued data unicorns represented by optical wheel intelligence.

The data collection technology path has not yet converged, and “high quality+scale” has become the key to solving the problem

Physical AI data systems face an “impossible triangle” of fidelity, scale, and cost, and there are technical and efficiency barriers in collection, interpretation, circulation, and iteration. In order to break through the bottleneck, the industry is undergoing a three-dimensional transformation of the business model, technical paradigm and asset attributes to accelerate the construction of a closed loop of data supply and monetization.

Trend 1: Business Model Innovation

Through the social “crowdfunding and crowdsourcing” model, the “eGO+UMI+ crowdsourcing” social data collection network is activated to obtain high-frequency, long-tail real-world behavioral data in a distributed manner, showing huge potential for large-scale expansion.

Trend 2: Technological Paradigm Innovation

Based on the “seed data base, world model expansion, and continuous learning closed loop” strategy, we have created a “virtual expansion and continuous closed loop” numerical engine to break the physical time and space limitations, and achieve exponential amplification of seed data collected by the commercial side.

Trend 3: Asset Value Reshaping

The asset attributes of physical world data are undergoing fundamental revaluation. It is leaping from being a “intermediate processing factor” in circulation to a core barrier and high-premium intangible asset for enterprises.

Market development is on the rise, and the competitive landscape is yet to be determined

Physical AI is experiencing explosive growth, driving the physical data and simulation market into a broad incremental space. Sullivan expects China's physical AI simulation and data platform market to exceed 180 billion yuan by 2030. As industrial competition accelerates the paradigm shift to “physical data driven”, technology giants, native physical AI data companies, and traditional data labeling vendors are relying on their core assets and technical advantages to actively seize the strategic window of the physical data era.

risk analysis

(1) Risk of lagging in core technology breakthroughs; (2) risk of imperfect industrial ecosystem construction; (3) risk of scarce application implementation scenarios; (4) risk of policy and regulatory adaptation; (5) risk of imbalance between investment and return.