I keep hearing people talk about the “AI bubble.”
Usually when we hear the word bubble, we picture something that gets bigger and bigger until eventually it pops—and then it is gone.
That is the picture that bothers me.
It is dramatic. It makes a good headline. But it is not really what I have watched happen with technology throughout my lifetime.
I have watched technology go through cycles over and over again.
At first, something new appears. People get excited. Money starts pouring in. New companies appear everywhere. Everybody wants to be the company that figures it out first, builds the best version, or captures the biggest piece of the market.
Prices can be high. Profit margins can be high. Expectations can become unrealistic.
Then the market starts sorting things out.
Companies disappear.
Prices come down.
What once seemed rare becomes common.
And eventually the technology stops feeling revolutionary at all.
It simply becomes part of everyday life.
I've Seen This Movie Before
I remember companies like Compaq Computers.
At one point, Compaq was a major name in personal computers. Today, most people probably do not think about Compaq at all.
That does not mean personal computers failed.
Quite the opposite.
Computers became so common and so deeply woven into business and everyday life that the market changed around them.
I have watched similar transitions with the Internet, smartphones, websites, digital advertising, social media, and many other technologies.
Some of the companies that seemed unstoppable disappeared.
Other companies were acquired.
New leaders emerged.
Prices dropped.
The technology improved.
And adoption continued.
That is why I think the phrase “AI bubble” can give people the wrong picture.
The Bubble May Be in the Companies, Not the Technology
There are thousands of companies racing into AI right now.
Everybody is trying to have the best model, the best AI assistant, the best image generator, the best video tool, the best automation platform, or the best AI solution for some particular industry.
They cannot all win.
Some of them will fail.
Some will merge.
Some will be bought.
Some products that look remarkable today may barely be remembered ten years from now.
And some businesses are undoubtedly being valued based more on expectations than on sustainable profits.
That part of the AI boom can absolutely correct itself.
But that does not mean AI disappears.
The important distinction is this:
The bubble can pop without the technology disappearing.
A technology industry can go through a major financial correction while the technology itself continues spreading.
We saw that with the Internet.
The dot-com crash did not end the Internet.
It helped sort out which businesses had something sustainable and which ones did not.
Meanwhile, the Internet continued moving deeper into commerce, communication, advertising, entertainment, banking, education, and nearly every other part of life.
AI may follow a similar path.
Watch the Price of Intelligence
One of the things I have been watching closely is the falling cost of using AI.
Early in a technology cycle, the scarce thing commands the money.
Computing power was expensive.
Storage was expensive.
Bandwidth was expensive.
Long-distance telephone minutes were expensive.
Website hosting was expensive.
Then engineering improves. Competition increases. Capacity expands. Prices fall.
Eventually the underlying unit becomes so inexpensive that most people stop thinking about its individual cost.
I believe something similar is beginning to happen with AI intelligence.
Models become more efficient.
Hardware improves.
Competition drives prices down.
Companies include more AI use inside subscriptions.
The cost of producing an answer keeps moving downward.
That tells me something important.
AI itself is gradually becoming less scarce.
And when something becomes abundant, its economic value begins moving somewhere else.
Cheap AI Can Hurt AI Companies and Accelerate AI at the Same Time
This is one of the strange parts of the cycle.
Falling prices can make life harder for companies that are trying to sell a capability that competitors can increasingly provide for less.
Margins shrink.
Products become harder to distinguish.
Companies that looked valuable because the technology was scarce may discover that scarcity was their advantage.
Some will not survive.
But cheaper AI also makes it easier for more people and more businesses to use it.
That is why I think this sentence matters:
The collapse in the price of AI may destroy AI companies while accelerating the adoption of AI.
Those two things can happen at the same time.
The company can disappear while the capability spreads.
If Intelligence Gets Cheap, What Becomes Valuable?
This may be the most important business question.
If nearly every business eventually has access to powerful AI, simply “having AI” is not much of a competitive advantage.
Your competitor will have it too.
The value moves to what you do with it.
Do you understand your customers?
Do you understand your business?
Can you connect your information?
Can you build useful workflows?
Can you recognize when AI gets something wrong?
Can you combine technology with experience and judgment?
Can you create a system that continues learning and improving?
Those things do not automatically become commodities just because tokens get cheaper.
In fact, they may become more important.
This is why I have gradually stopped getting excited about individual AI tools.
I have bought and tested a lot of them.
Many are impressive.
But the tool is not the business.
The system is.
When intelligence becomes cheap, intelligence itself stops being the product. What you do with it becomes the product.
Smartphones Did Not Disappear Either
Think about smartphones.
There was a time when having a smartphone seemed remarkable.
Today we do not spend much time marveling that a phone can connect to the Internet, take photographs, provide directions, play video, send email, run a business application, or make a video call.
We simply expect it.
The technology became part of normal life.
AI appears to be moving in that same direction.
Today we still say things like “AI-powered” because AI itself gets attention.
Someday that phrase may sound strange.
We do not advertise most websites today by saying they are “Internet-powered.”
We do not advertise an accounting system as “computer-powered.”
Those technologies became assumed.
AI may eventually become the same way.
What the Word Bubble Can Hide
The media often uses dramatic words because dramatic words get attention.
Bubble is one of those words.
It can leave people with the impression that if the bubble pops, the whole thing goes away.
Business history is usually messier than that.
A market can be overheated.
Investors can overpay.
Companies can fail.
Products can disappear.
Prices can collapse.
And the underlying technology can still become more important than ever.
Those are not contradictions.
They are often part of the same process.
So Is There an AI Bubble?
Maybe.
There may be an enormous financial bubble around some AI companies, investments, valuations, and expectations.
I do not know which companies will survive.
Nobody does.
But I do not think that is the question small-business owners should spend most of their time worrying about.
The better question is:
What is this technology becoming, and how is it changing the way my customers, employees, competitors, and business operate?
That question remains important whether today's leading AI company is still leading ten years from now or not.
Do Not Bet Everything on a Tool
This is also why I do not want small-business owners becoming completely dependent on one AI product.
The product may change.
Its pricing may change.
The company may be acquired.
A better competitor may appear.
A feature you depend on may disappear.
Build around your business instead.
Keep your knowledge.
Keep control of your information.
Understand your processes.
Use AI tools where they make those processes better.
And be prepared to change tools when the technology changes.
That approach survives the cycle.
I've Watched This Happen Before
Maybe age gives me a little different perspective on this.
I have lived through enough technology changes that I have seen companies everyone knew disappear and technologies people dismissed become ordinary parts of life.
Usually, while we are living through the change, everything looks chaotic.
Looking backward, the pattern becomes much clearer.
I think we are inside one of those periods right now.
There is tremendous hype.
There will be failures.
There will probably be some spectacular failures.
Prices will fall.
Business models will change.
Some fortunes will disappear and others will be created.
But underneath all that noise, something much larger is happening:
AI is moving from being a remarkable new technology to becoming part of everyday life.
And that may ultimately matter far more than whether somebody calls today's market a bubble.
The NTA Point of View
Do not chase every AI company.
Do not assume today's hottest tool will still matter five years from now.
And do not dismiss AI because some of today's AI businesses may fail.
Learn what the technology can do.
Watch what it produces.
Build around the real needs of your business.
And create systems that can evolve as the technology does.
Because the companies may come and go.
The technology will change.
But the larger shift has already begun.
Reflection Questions
- When you hear the phrase “AI bubble,” what do you think is actually at risk—the technology, individual companies, investment valuations, or all three?
- What technology have you watched move from expensive or unusual to ordinary?
- Is your business dependent on one AI vendor, or could you change tools without losing your knowledge and process?
- If AI capability becomes inexpensive and widely available, what knowledge, judgment, relationships, or systems would still make your business different?
- Are you buying AI because it is new, or because it improves a real part of your business?
- What would you keep even if the AI tool you use today disappeared tomorrow?