Published on Jan 30, 2026 • Updated on Nov 7, 2026
This article was written by Adem Tumerkan
Content Strategist, Dunham | 2025 ThinkAdvisor Luminaries Award Winner | 2026 Wealthies Finalist, Thought Leader of the Year | Macroeconomics, markets, geopolitics & global trends
Subscribe to get the latest insights delivered straight to your inbox.
We care about your data in our privacy policy
The current AI boom shows classic bubble traits. Valuations have exploded, corporate AI spending is increasingly financed with debt, and each new dollar buys less innovation than the last. At the same time, today’s shortages in power, chips, and data centers are setting up tomorrow’s overcapacity and margin crunch. The risk is that expectations, spending, and leverage outrun the revenue and productivity gains needed to support them.
Since ChatGPT went public in late 2022, artificial intelligence (AI) has dominated every headline, every conversation, and every investment theme.
We’ve heard it all - “AI will take over the world,” “AI will reshape the economy,” and “Anyone not investing in AI is a fool."
And while there’s significant merit for the upside of AI, I’m not convinced it’s the end-all, be-all the crowd expects.
If history means anything, the success and hype surrounding AI are already planting the seeds of its own demise.
So the real question becomes: Is this another hype cycle - like the roaring 1920s, the dot-com bubble, or the housing boom?
Many may believe that this time is different.
But, as Sir John Templeton warned,
I believe three things are carrying AI right now:
Of course, at first glance, these sound like tailwinds. And they certainly have been.
But remember. In markets, what fuels the boom cycle often ends it.
| Force | How it fuels the boom | What could reverse it | Potential market impact |
|---|---|---|---|
| Hype and high expectations | Promises of productivity gains and future profits attract investors and lift valuations. | Revenue and profits fall short of expectations. | Lower valuations, weaker sentiment, and less funding for speculative ventures. |
| Corporate spending and debt | Spending on chips, servers, and data centers drives expansion. Borrowing helps finance it. | Investment returns disappoint while financing obligations remain. | Pressure on cash flow, spending cuts, and financial stress for leveraged firms. |
| Infrastructure shortages and overcapacity risk | Scarce power, chips, and data center space support pricing power and attract new investment. | New supply outpaces demand, creating overcapacity and greater competition. | Weaker pricing, thinner margins, and potential consolidation. |
*These are potential risks, rather than a prediction that all three will unfold together.
Those are the same ingredients that fueled the dot-com bubble, the housing boom, and every mania before it.
So, let’s take a closer look at what’s really behind the boom - and why it may not last.
We all know that emotions drive markets.
And right now, markets are drunk on optimism about AI.
Analysts, CEOs, investors, entrepreneurs – all convinced that AI will change everything and mint fortunes along the way.
And so far, they seem right.
But what if it doesn’t hold? What if the promised productivity revolution never quite materializes?
Well, if that happens, trillions in paper wealth could vanish – fast – as hype turns to fear.
And that risk is growing.
Back in August, the Massachusetts Institute of Technology (MIT) published a worrying paper showing that most companies aren’t seeing the revenue boom AI was supposed to deliver.
According to The GenAI Divide: State of AI in Business 20252, only 5% of AI pilot programs achieved meaningful revenue growth. The other 95% stalled, delivering little measurable impact on profit. The researchers found the problem isn’t exactly the models – but rather the execution (aka companies are overspending on tools that don’t integrate, don’t learn from workflows, and don’t scale).
The result is an economy where AI adoption is soaring, but returns aren’t.
Even worse is that spending could start to backfire. If companies keep pouring money into underperforming systems, the payoff gap will grow wider - and valuations will have to adjust (more in the next section).
Still, falling hype doesn’t mean AI is useless (far from) - just that it might be overpriced right now.
For example, as Gartner’s Five Phases of the Hype Cycle shows, technology booms often follow the same pattern:
Figure 1: Gartner, Dunham, 2025
I’d estimate we’re around the peak of inflated expectations and sooner than later will plunge into the trough of disillusionment.
Hype seems surreal in this market. But that can quickly unwind.
Big tech is spending fortunes in this AI arms race.
And the spree isn’t even close to slowing. . .
Morgan Stanley estimates nearly $3 trillion in AI infrastructure spending from 2025–2028, enough to add about 0.5% to U.S. GDP growth this year and next.
Figure 2: UnderstandingAI, Substack, October 2025
Such massive investments are helping push companies’ profits (and their share prices) to records.
But this will – inevitably – lead to a significant squeeze in margins.
Why? Because of the law of diminishing returns.
The early money spent in AI created breakthroughs - ChatGPT, copilots, image generation, and autonomous code. Each dollar then delivered exponential returns in capability.
But now? The spending curve has gone parabolic while the innovation curve is flattening – aka more spending for less innovation.
Said another way - the marginal gain from each new model or chip is shrinking - yet the cost of chasing it keeps drastically rising.
Figure 3: Dunham, 2025
Simply put, the weight of this enormous AI spending will eventually collapse under its own mass.
Meanwhile, the top players in the AI space are increasingly engaging in sketchy circular financing deals - which is historically a fragile sign.
The same risk now looms over AI.
Because subsidized momentum fueled by self-financed spending = more fragility.
What started as a cash-flow story is now entering its credit-driven chapter.
Keep in mind that most companies here remain financially strong, but the trend shows that AI spending is stretching even Big Tech’s balance sheets.
Thus, as credit replaces cash, the story of organic, self-funded growth is giving way to one of rising leverage, opaque financing loops, and uncertain payback - the classic signals of a boom nearing its peak.
Put simply, every boom runs on hype until it starts borrowing against it. Then it’s a recipe for disaster.
A ripple effect from all this AI hype and spending has been the extreme shortage of supplies - from commodities, power, and space.
Because of this, prices for copper, aluminum, and energy have soared, and some regions are already rationing electricity for new projects.
Many Americans already dislike the idea of AI taking their jobs. And they’ll like it even less when it starts raising their power bills further.
Figure 4: WIRED.com, October 2025
Thus, what started as an AI revolution is now colliding headfirst with the limits of the physical world and rippling into something far bigger.
And when prices rise, entrepreneurs rush in to chase profits.
But – like every bubble before - we’re not seeing the true innovators anymore.
We’re seeing what I call “slop-preneurs” - the latecomers who pile in when the real opportunity’s already been picked clean.
Thus, they’re not creating efficiency - just chasing momentum.
And in every cycle, it’s the slop-preneurs who show up last - and leave first.
But - there’s upside when they do show up. . .
Because these slop-preneurs will bring on extra supply – helping ease shortages and push prices down.
Sure, many will likely go bust from excessive competition and falling prices that crush margins (from overcapacity). But the last firms standing will pick over their bones and grow stronger.
The only problem is that this may take a while to play out.
The point is, shortages always create gluts, and gluts always crush margins.
This boom has already planted seeds of its own destruction.
So, am I saying AI itself isn’t important? No - it’s clearly revolutionary.
Am I saying it’s not worth watching? Absolutely not. It will reshape industries, workflows, and maybe even economies.
But is the hype worth the price being paid for it right now? That’s where I have doubts.
That’s where we are with AI. The technology isn’t the problem. The hype and valuations are.
A correction could mean lower valuations, spending cuts, weaker supplier orders, and financial trouble for heavily leveraged projects. And if new data center capacity outpaces demand, competition could push prices down and squeeze margins further - which is dangerous with so much debt being used now.
And none of that requires AI demand to up and disappear. Revenue could keep growing and still fall short of what investors and lenders assumed.
For instance, say an AI lab grows revenue by 25% in a year. That’s still strong growth. But if investors priced it for 50%, that same result could suddenly look disappointing.
That’s the risk.
History offers plenty of examples of transformative technologies attracting more capital than investors could profitably put to work
And I believe AI will too.
And when it does, it won’t mean the end of AI - far from it. It’ll simply mark the end of the illusion that it could defy gravity at any cost.
Thus, AI won’t collapse from failure - it’ll stumble under the weight of its own success.
Because in markets, progress is never the real problem. Hype, valuations, and scarcity always are.
But as always, this is just some food for thought.
Are we in an AI bubble? The AI boom carries bubble risks, but strong demand alone doesn’t tell us whether today’s investments will pay off. The question is whether future revenue, profit margins, and productivity gains can justify current prices and spending. AI can keep improving and find wider use even if some investors earn disappointing returns.
Could AI data centers face overcapacity? AI data centers could face overcapacity if usable supply grows faster than demand at prices that cover costs and support returns. That’s a future risk, rather than an established broad glut. CBRE reported 1.4% vacancy across primary North American markets in the first half of 2026. The concern is whether projects remain profitable as capacity grows and customer needs change.
What is circular financing in AI, and why is it risky? Circular financing happens when an AI supplier invests in, lends to, or helps finance a customer that buys its products or services. Sales may then depend partly on ongoing financial support rather than customer demand that can sustain itself. If funding or revenue falls short, losses can spread among suppliers, customers, and lenders. The arrangement alone doesn’t prove fraud.
What happens if the AI bubble bursts? An AI bubble bursting could bring falling stock prices, cuts in capital spending, and financial stress for companies whose revenue can’t cover their commitments. Suppliers, data center owners, and lenders could face losses as investment slows. AI adoption wouldn’t necessarily stop. The damage would depend on debt levels, customer demand, contract terms, and whether infrastructure generates steady cash flow.
This communication is general in nature and provided for educational and informational purposes only. It should not be considered or relied upon as legal, tax or investment advice or an investment recommendation, or as a substitute for legal or tax counsel. Any investment products or services named herein are for illustrative purposes only, and should not be considered an offer to buy or sell, or an investment recommendation for, any specific security, strategy or investment product or service. Always consult a qualified professional or your own independent financial professional for personalized advice or investment recommendations tailored to your specific goals, individual situation, and risk tolerance.
Information contained in the materials included is believed to be from reliable sources, but no representations or guarantees are made as to the accuracy or completeness of information.
Dunham & Associates Investment Counsel, Inc. is a Registered Investment Adviser and Broker/Dealer. Member FINRA/SIPC. Advisory services and securities offered through Dunham & Associates Investment Counsel, Inc.
