A lot of people think the AI boom is one giant house of cards right now.
And honestly… I understand why.
Nvidia invests in AI companies. Those AI companies then use the money to buy Nvidia chips. Everyone celebrates the revenue growth. Stocks go higher. Rinse and repeat.
Here’s an excellent image I found to explain what’s happening exactly:
To critics, this looks less like innovation… and more like companies passing money around in a circle.
The term they use is “circular spending”.
Imagine if Sheng Siong invested in a chicken farm… then the chicken farm used that money to buy equipment from Sheng Siong’s subsidiaries… and then Sheng Siong proudly reported higher revenues because of it.
You’d probably raise an eyebrow too.
Which is why more people are starting to compare today’s AI boom to the dot-com bubble.
- “It’s a bubble.”
- “It’s all hype.”
- “The spending is fake.”
- “Eventually the money runs out.”
Fair concerns.
But the more I dug into this whole “circular spending” narrative, the more I realised something important:
Sometimes… this is exactly what real industrial revolutions look like.
Big tech companies and chipmakers like Nvidia are pouring billions into AI startups and infrastructure companies.
Why?
Because these companies need enormous computing power to train AI models.
And guess who sells the best AI chips in the world right now?
Nvidia.
So yes – Nvidia invests into the ecosystem… and the ecosystem comes back to buy Nvidia GPUs.
That’s the “circular” part people are talking about. Critics argue this artificially inflates demand.
And to be fair, there’s some truth there.
If companies are only spending because investors are funding them endlessly, eventually the music stops.
That’s exactly what happened during parts of the dot-com era.
Back then, companies were laying fibre optic cables everywhere, building websites everywhere, throwing “.com” behind random businesses and calling it innovation.
The problem? Most of the demand didn’t exist yet.
It was build first, demand later. And many times… nobody came.
That’s why the crash was so brutal.
But here’s where I think today’s AI boom is different.
People are actually using AI.
Not theoretically. Not “someday maybe”. Right now.
- Students use ChatGPT to study.
- Employees use AI to write reports.
- Businesses use AI for customer support.
- Developers use it to code.
- Marketers use it to create ads.
Even you use it to write your emails. I know, cause that’s me too sometimes.
The demand is REAL.
That doesn’t mean valuations are reasonable and that every AI company survives.
But it does mean this may not be the same kind of bubble people think it is.
You see, every major industrial revolution in history looked financially irrational in the beginning.
Railroads, cars, telecom, and even internet infrastructure looked excessive in the beningging,
In fact, during the early automobile era, car manufacturers often had to finance their own suppliers because the ecosystem wasn’t mature enough yet.
They invested in factories, parts suppliers, and infrastructure.
Why?
Because demand existed… but capacity didn’t.
So companies essentially funded the ecosystem into existence.
That’s very similar to what’s happening in AI today.
The big players aren’t just selling products. They’re building an entire industrial ecosystem:
- chips
- data centres
- storage
- cloud infrastructure
- AI software
- energy infrastructure
- enterprise applications
And that kind of buildout always looks messy and expensive early on.
The internet bubble is actually a good example of this.
People often say that the dot-com bubble proved tech was overhyped. But that’s not fully true.
The internet did change the world. The bubble wasn’t wrong about the technology.
It was wrong about:
- valuations
- timelines
- and which companies would survive
That’s a huge distinction.
Because I suspect the same thing may happen with AI.
Some companies today probably won’t survive.
Some valuations are probably insane.
Some investors will almost definitely get burned.
But that doesn’t mean AI disappears.
If anything, the bigger risk may be underestimating how deeply AI embeds itself into the economy over the next decade.
And honestly… I think this is where the investing conversation becomes much more nuanced.
Because people keep assuming that just because AI will change the world, it automatically means that AI stocks will make you rich.
History says it’s not that simple.
The internet changed the world, but many dot-com investors still got wiped out.
Railroads transformed economies, but railway bubbles still collapsed.
Sometimes the technology is real… but investors simply overpay for the future too early.
And I think that’s the biggest risk with AI investing right now.
The risk is whether current valuations already assume too much perfection.
Because once markets believe a technology will dominate the future, prices stop reflecting reality today… and start pricing in dreams 10 years ahead.
That’s where things become dangerous.
Because the winners during technological revolutions are rarely as obvious as people think.
During the internet boom, many assumed Yahoo would dominate forever. Few expected Amazon to become what it is today.
- Sometimes the infrastructure winners outperform.
- Sometimes the application layer wins.
- Sometimes everyone gets disrupted by a company that doesn’t even exist yet.
That uncertainty is exactly why I personally prefer broad exposure over concentrated bets.
Instead of trying to perfectly predict which AI company becomes dominant, I think most investors are better off owning diversified global exposure that can participate in long-term technological growth while reducing the risk of being catastrophically wrong.
That’s why I think the goal shouldn’t be trying to become rich overnight from AI.
It should be:
- staying diversified
- avoiding emotional investing
- and positioning yourself to benefit from long-term innovation without betting your financial future on one narrative.
Because if history teaches us anything, it’s this:
Revolutions often survive. Speculation often doesn’t.
So if you’re on the side of picking which stocks will do best, diversify.
If you’re on the side of avoiding AI because of the bubble, diversify.
Because either way, you’re participating in the upside while protecting your backside downside.
Stay informed, stay adaptable… and stay invested.