Blackstone President Jon Gray: Current AI Infrastructure is Essentially Different from Telecom Bubble
Faced with growing concerns about an AI bubble on Wall Street, Blackstone Group President Jon Gray offered a starkly different view, stating that the current AI infrastructure development is fundamentally different from the telecom bubble of the 1990s.
At that time, companies often expanded capacity on a large scale and then waited for demand to appear, whereas today, companies are striving to build infrastructure to keep up with the rapidly growing real demand.
Gray described the current AI demand as parabolic growth and revealed that Ode, the AI consulting firm co-founded by Blackstone and Anthropic, is currently unable to meet client demand.
This stands in stark contrast to the pessimistic view expressed by Citadel founder Ken Griffin just days ago, who was almost certain that we are in an AI bubble.
Blackstone is positioning AI as a core investment direction, with investments in the data center sector reaching approximately $200 billion, and is broadly allocating capital to power, transmission facilities, and other supporting infrastructure.
Gray acknowledged that even if real demand exists, it does not mean that every investment will yield ideal returns; some cloud service providers specifically targeting AI, such as CoreWeave, are facing financing difficulties, despite decent performance in their main business.
Building data centers requires massive upfront investment, and investors must weigh whether future cash flows can cover financing costs and the risk of asset depreciation due to technological iterations. The current bubble debate is essentially testing whether the explosive demand for AI can ultimately translate into sustainable commercial returns that cover high costs.
In this context of opposing views, capital continues to flow into data centers, power, and supporting infrastructure, benefiting large-scale players like Blackstone, while pure AI cloud service providers like CoreWeave face pressure.
Source: Public Information
ABAB AI Insight
Jon Gray leads Blackstone in expanding capital from traditional real estate and private equity into data centers, promoting acquisitions in logistics and digital infrastructure after the 2010s, forming a path where heavy assets support technological demand, and has repeatedly emphasized the support of physical assets for growth cycles in public forums.
The decision to allocate approximately $200 billion to data centers and extend into power and transmission, while co-founding the Ode consulting firm with Anthropic, is motivated by capturing the already emerging parabolic demand, positioning infrastructure as a capital anchor for AI workloads.
Similar cases include the capacity expansion of telecom operators in the 1990s, which built first and waited for demand, and Blackstone's own early layout in logistics real estate after 2008; currently, we are in the expansion phase of AI from software applications to heavy asset infrastructure control.
Essentially, this represents a concentration of capital, with funds shifting from pure software and cloud services to data centers and energy support, driven by the mechanism that real demand has outpaced capacity, making upfront heavy asset investments potentially cash flow covering, and shifting pricing power from light asset narratives to infrastructure holders that can match demand growth rates.
ABAB News · Cognitive Law
- Demand leads, capacity gains pricing power
- Heavy assets consume light asset narratives
- Parabolic demand tests high-cost returns