We’re Not in an AI Bubble. We’re Watching Big Tech Finally Spend Its Cash.
For the last year, the dominant narrative has been simple: AI is a bubble. Too much hype, too much capital, too many promises — and eventually, it all pops.
The warnings are everywhere. Nobel laureate Daron Acemoglu calls AI models “hyped up” and says “we’re investing more than we should.” Even Sam Altman admits investors are “overexcited” about AI. One analyst claims we’re in “the biggest and most dangerous bubble the world has ever seen.”
I don’t buy that framing — or at least, not the version that predicts a 2008-style collapse.
What we’re seeing right now isn’t a speculative bubble propping up weak fundamentals. It’s something very different: big tech companies finally deploying tens of billions of dollars in capital that’s been sitting idle for years — because until now, there was nowhere sensible to put it.
AI didn’t create the money.
AI gave it a destination.
The Cash Problem Nobody Talks About
If you’ve never worked inside big tech, it’s hard to grasp the scale of the cash problem.
Not revenue.
Not valuation.
Cash.
Between roughly 2015 and 2022, companies like Meta, Google, Apple, and Microsoft accumulated enormous cash reserves. By the end of 2021, Big Tech held a combined $476.7 billion in cash, cash equivalents, and marketable securities:
- Alphabet consistently sat north of $100B (currently ~$100B as of Q2 2024)
- Apple peaked near $167B including marketable securities before aggressive buybacks
- Microsoft hovered in the $80-130B range (~$76B as of Q2 2024)
- Meta held $50-60B in cash and short-term investments
- Amazon reached $89B by Q2 2024
This wasn’t accidental. These companies printed cash quarter after quarter, but there were very few places to deploy it at scale without destroying returns or triggering regulatory scrutiny.
You can’t:
- Buy meaningful companies without antitrust issues.
- Launch moonshots at $50B scale without burning shareholder trust.
- Dump tens of billions into financial markets without distorting them.
So the cash accumulated. Slowly. Relentlessly.
By the late 2010s, big tech wasn’t constrained by capital — it was constrained by deployability.
Why This Is Nothing Like 2008
The 2008 bubble was built on leverage, fragility, and financial engineering.
This moment is built on cash-on-hand and infrastructure investment.
That difference matters.
In 2008:
- Capital was borrowed.
- Risk was hidden.
- The system collapsed when liquidity disappeared.
In 2024-2025:
- Capital is owned outright.
- Spending is discretionary.
- No one is being forced to liquidate to survive.
When Meta spends $65-72B on capex in 2025, it’s not borrowing against the future — it’s spending money that already exists on its balance sheet. The numbers are staggering:
- Amazon: $100B+ planned for 2025 (up from $83B in 2024)
- Google: $91-93B in 2025 (up from $52.5B in 2024)
- Microsoft: $80B in fiscal 2025
- Meta: $65-72B in 2025
Combined, that’s over $320 billion in a single year — deployed into data centers, power infrastructure, and compute capacity.
This isn’t a house of cards. It’s a dam opening.
Even skeptics acknowledge the structural difference. IMF Chief Economist Pierre-Olivier Gourinchas stated: “An AI investment boom may lead to bust, but not likely systemic crisis… Unlike the 2000 dot-com bust or 2008 financial crisis, AI investments are generally not debt-financed.”
That’s the key distinction. The money is already there.
AI as a Capital Sink, Not a Magic Wand
AI didn’t arrive as a miracle product.
It arrived as an infrastructure excuse.
Training clusters. Inference fleets. Power contracts. Data centers. Networking. Cooling. Specialized hardware. Land.
These are boring, real, industrial investments — the kind that absorb enormous amounts of capital without promising overnight returns.
And that’s exactly what big tech needed.
AI is not the thing propping up the economy.
Big tech spending is.
The AI narrative just made it politically, economically, and culturally acceptable to finally deploy that spending at scale.
Why the Spending Is Mostly Infrastructure
Look closely at where the money is going:
- Multi-year data center builds
- Power generation and grid commitments
- Long-term hardware supply chains
- Fiber, networking, and interconnects
- Custom silicon programs
These aren’t moonshots.
They’re not flashy consumer bets.
They’re industrial-scale projects with decade-long horizons.
That’s why calling this an “AI bubble” misses the point. A bubble implies fragile demand and speculative excess. What we’re seeing is closer to a re-industrialization of the digital economy.
What Happens When You Finally Can Spend the Money
For years, people asked: Why doesn’t big tech invest more aggressively?
The answer wasn’t fear — it was scale.
There simply weren’t enough projects large enough, defensible enough, and slow enough to absorb tens of billions without blowing up margins or attracting regulators.
AI infrastructure solved that problem.
Not because it’s magical — but because it’s expensive, slow, and necessary.
Where This Argument Could Be Wrong
I should be honest about the weaknesses in this framing:
“What if AI doesn’t deliver ROI?” If the productivity gains from AI infrastructure don’t materialize, big tech is left with expensive data centers that don’t generate sufficient returns. The spending is real, but the value capture isn’t guaranteed. That said, even skeptical scenarios suggest compute demand will continue growing — the question is rate of return, not total write-off.
“Isn’t this what people said about every bubble?” Yes. Every speculative excess looks like rational investment to someone. The difference I’d point to: this capital is owned, not borrowed; the spending is on physical infrastructure with residual value; and the companies deploying it are profitable on their core businesses regardless of AI outcomes.
“What about the smaller players?” My argument applies to big tech with fortress balance sheets. The AI startup ecosystem is a different story — many are burning cash on the assumption that scaling laws continue indefinitely. Some of those bets will fail. The distinction is between infrastructure builders and application gamblers.
None of this means AI is a sure thing. It means the structure of this moment is different from 2008 — and that matters for how we think about systemic risk.
What This Means For…
Founders: The infrastructure buildout creates opportunity. The companies spending $320B need software, services, and tooling. That’s a procurement cycle, not a bubble pop. Build for the builders.
Investors: Watch capex efficiency, not hype cycles. The question isn’t “will AI work?” — it’s “what’s the return on this infrastructure at scale?” The winners aren’t necessarily the model makers; they might be the picks-and-shovels players.
Policymakers: The systemic risk isn’t financial collapse. It’s concentration — a handful of companies controlling the compute layer of the economy. That’s an antitrust question, not a bubble question.
This Is the Setup, Not the Punchline
If you’re looking for a clean pop — a moment where AI collapses under its own hype — you’re probably waiting for the wrong thing.
The real story isn’t about model capability or chatbots. It’s about capital finally moving.
But if this isn’t a bubble, what should we be worried about? In Part II, I’ll argue that the real economic threat isn’t AI taking jobs — it’s physical automation at industrial scale. And we’re having the wrong conversation about it.
TL;DR
- Big tech accumulated $500B+ in cash with nowhere to deploy it at scale
- AI infrastructure gave that capital a destination — not because it’s magical, but because it’s expensive and slow
- This spending is owned, not borrowed — structurally different from 2008
- The question isn’t “will the bubble pop?” — it’s “who captures the value from this buildout?”
- Watch for concentration risk, not systemic collapse