Not an AI bubble, but a shortage of AI infrastructure - Larry Fink warns that the real problem is just beginning...
Those who waited to hear the BlackRock CEO talk about an "AI bubble" got exactly the opposite.
In an interview with CNBC, Larry Fink said he is much more concerned about a state of underinvestment in AI infrastructure than a bubble.
According to him, demand for computing power is growing faster than the industry's ability to build new infrastructure - and the main bottleneck is simply a shortage of electricity, data centers, and hardware.
This is precisely the point the market is still struggling to digest.
We already know that Microsoft, Amazon, Google, and Meta plan to continue increasing their capital expenditures at an aggressive pace.
According to analysts' forecasts, these four giants together could approach nearly a trillion dollars in capital expenditures by 2027.
The implication is simple: if demand is still higher than supply, the wave of investments is far from over.
The big winners are not just the model providers, but the entire value chain that enables infrastructure construction:
Chips. Memory. Network equipment. Cooling systems. Power generation. Data center equipment. Companies that build the facilities themselves.
News outlets love to tell stories about bubbles, that's true.
At the same time, the whales are taking advantage of this,
They are absorbing the volatility while retail investors get cold feet.
But meanwhile, Fink tells a completely different story - the world is not building fast enough to keep up with the pace of the AI revolution.