Chinese technology platforms could begin reaping significant benefits from artificial intelligence within the next two to three years, but only if current restrictions on chip supplies and infrastructure development are eased. This scenario is directly tied to a redistribution of market power in the AI sector.
Shift in the Center of Profit
At present, the primary margins in the AI sphere are captured by suppliers of hardware and related services. The market is hitting a severe shortage of computing power, making "hardware" the main beneficiary of the boom. However, as my observations of the industry underscore, the situation could change dramatically. If export restrictions on advanced chips are relaxed, pricing power will inevitably shift to internet platforms that boast vast audiences, unique data, and well-developed distribution channels. For giants like Tencent and Alibaba, this means a return to a phase of actively monetizing their AI developments.
Rising Costs and Market Anxiety
The forecasts emerge against a backdrop of sharply increased capital expenditures by key players. Alibaba, for instance, nearly tripled its capital investments in the second quarter, bringing them to 52.8 billion yuan (approximately $7.86 billion). Notably, the company's free cash flow turned negative for the first time, coming in at -13.8 billion yuan ($2.05 billion). Overall, the quarterly outflow of free funds reached 44.7 billion yuan ($6.65 billion), more than double the figures from the previous year, while total quarterly spending soared to 67.7 billion yuan ($10.07 billion).
Such aggressive investments are causing wariness among investors, especially amid a slowing macroeconomic environment in the second half of the year. Based on my industry analysis, Chinese tech companies' annual AI spending is already equivalent to their cash flow over 1.5 years. This is an enormous burden, but it is necessary to survive the competitive race, even if the returns prove weaker than expected.
Efficiency as the Key Advantage
It is worth noting that the combined AI spending of Chinese companies amounts to only about one-seventh of their American competitors' expenditures. This gap stems from limited access to foreign chips and smaller business scale. However, the Chinese sector has an undeniable advantage—cost efficiency. The cost of training local models is estimated at less than 10% of the level of global leaders, and the average API price of major Chinese models does not exceed 20% of international counterparts.
In the long term, the key factor will not be the volume of capital expenditures, but the ability of platforms to monetize their audiences and data. It is precisely these assets that will allow China not just to catch up, but also to seize the initiative in the AI race.
My view: the current phase of "infrastructure hunger" is temporary. Once the chip supply chain normalizes, we will witness a sharp shift in profits from hardware manufacturers to service giants. Investors should prepare for this rotation now.