Hey everyone,

👋 Welcome to the 21 new readers who joined since last week!

This week was another big one. We're really getting into the business end of 2025.

Still building out our email nurture series and SEO architecture, which is a joy with AI tooling.

I also attended my first-ever startup pitch event last night, which was fun.

As I get back into writing about AI, I thought it might be useful to take a look at the current state of the industry.

It’s a funny time since everyone’s talking about the ‘AI Bubble.’

Last weekend, I spent 3 days analysing 15+ hours of podcasts to cut through the noise and figure out what’s going on.

Excited to dive into what I found and potentially burst some bubbles.

Bubble Talk

Humans love mental shortcuts.

When we see a new piece of information, our brain quickly tries to make sense of it by anchoring us to past experience.

It seems most analysts are comparing AI today to the dot-com bubble of the late 90s.

For someone browsing headlines, this argument is convincing.

Before last weekend, I felt the same way. But I’ve realised today’s market is fundamentally different from any other market cycle we’ve seen.

The real story is far more nuanced.

Broadly, two camps have taken each side of the AI bubble debate:

1. Investment analysts

2. Tech bros

Since I see myself within both worlds, I initially felt conflicted. But I realised that I have a decent shot at walking the tightrope between the two.

Instead of feeding the fervour of one side, I thought it’d be far more valuable to lay out both sides of the argument.

First, I'll present all the evidence for an AI bubble’s existence. Then, I'll present the evidence to the contrary.

Honestly, I find both arguments compelling. The risks feel very real to me, but so does the promise of the technology.

I’ve tried to make this analysis as objective as possible, drawing from resources across both technology and finance.

Of course, it’s impossible to strip out all bias. I hope I’ve captured the fundamentals so you can decide for yourself where you stand.

Evidence for an AI bubble

The investment analyst’s argument has 5 factors.

1) Market Dynamics

This is the primary driver of the Dot-com bubble comparisons.

• The Magnificent 7 (Apple, Microsoft, Amazon, Alphabet, Meta, Nvidia, and Tesla) now account for 25% of global stock market valuations.

• OpenAI was recently valued at $500B, 38x revenue of $13B. The creator of ChatGPT reportedly is not making any material profits, even on their most expensive $200/month subscription tier.

These companies are a single point of failure for stock markets. Any slight knock could have an undue impact on global markets due to their enormous valuations.

2) Circular Deals

The report above illustrates the interconnected web of recent AI deals.

Basically:

• Nvidia invests $100B in OpenAI → OpenAI buys Nvidia chips

• Microsoft invests in OpenAI → OpenAI buys Microsoft cloud services

• Oracle gets $300B from OpenAI → Oracle buys Nvidia chips

This is referred to as financial ‘roundtripping’, which can create artificial demand for AI, which fuels stock prices.

One default might cause the entire Jenga tower to topple.

3) Capex Blowout

OpenAI alone has committed to spending $1.4 trillion on AI infrastructure over the next 5 years. Genuine questions have been asked about how an unprofitable startup plans to fund this.

That may not even be enough. McKinsey projected that the industry needs $5.2T to build their roadmap of AI systems.

And Bain estimated that AI companies need to generate $2T in annual revenue to justify current levels of spending.

That is a lot of money that people are concerned won’t materialise.

4) Resource Constraints

Satya Nadella, the CEO of Microsoft, recently stated that power is the ultimate bottleneck to building AI:

It can befuddle the brain to consider how all this infrastructure is built out so quickly, especially considering it can take up to 10 years to deploy a nuclear power plant.

There are genuine concerns that the entire industry will come to a screeching halt when we reach our energy limits.

5) Taiwan Risk

Over 90% of the world’s computer chips, which power your iPhone and ChatGPT itself, are built by a single company in the most geopolitically precarious country in the world.

That company is TSMC, whose headquarters are conveniently located on China’s doorstep.

There are genuine US intelligence reports that mark a potential Chinese invasion of Taiwan by 2027.

No one knows what will happen if China does it. Will they still permit the export of chips to the West?

There is no viable alternative if TSMC's production is disrupted, and that could be catastrophic for the AI build-out.

Evidence against an AI bubble

On the other side of the aisle are the tech bros: the AI founders and Silicon Valley VCs who dream of a utopia where work is optional and robots clean our homes.

That might be a bit tongue-in-cheek, but the tech bros have a pretty convincing argument.

1) Different Market

Today's tech industry is fundamentally different to the 90s:

• By the time the dot-com bubble burst, 97% of the internet infrastructure built was never used.

  • Compared to today, 100% of AI infrastructure is fully utilised. There’s just not enough compute to meet demand.

• The Magnificent 7 are worlds apart from the early internet companies, which relied on massive debt to fund themselves.

  • Collectively, the Mag 7 have nearly $1 trillion of cash on hand. These are literally the most profitable companies in history, self-funding the AI buildout.

2) Smart Money

There's that age-old saying to “follow the money” when you want to find true signal.

Well, here's what the smart money in AI is doing today.

The Hyperscalers are ramping up their AI investments with each earnings call. This is not frivolous. They’ve reported a 10-point increase in their returns on invested capital.

Microsoft even kept an AI chip deployment internally rather than renting them out, as they found it to be more profitable.

“Following the money” denies any claims of speculation. Real returns are driving AI investments.

3) Paying Customers

Customers are paying real money for AI services.

• ChatGPT has 60M paying subscribers. And they’re only growing.

  • OpenAI is projected to have a $20B revenue run rate by year-end.

• Cursor, the AI code editor, just surpassed 1T AI tokens, indicating massive developer usage on their platform.

Unlike the dot-com bubble, where most products were vaporware, AI is delivering tangible value today.

4) PE Firms

Private equity (PE) firms are rapidly deploying AI into their portfolio companies.

Jonathan Ross, founder of an AI infrastructure provider, stated recently that they’re “all over us” for cheap AI compute.

PE firms notoriously bring bottom-line discipline to everything they do. If they didn't see real cash flow gains in their portfolio companies, they wouldn’t touch AI.

5) Bubble Paradox

A true bubble requires widespread market capitulation, where every single participant is caught up in euphoria.

Imagine Warren Buffett investing in AI startups. That would be real bubble territory.

The mere fact that we have an AI bubble debate is a healthy sign. It proves that the market hasn't lost its mind. Yet.

Thoughts

What the AI bubble debate actually comes down to is about timeframes.

The investment community is focused on the next 12 months, warning about excessive valuations and potential market corrections.

Technologists are focused on the next decade, discussing the transformational impact of AI on society and the economy.

Two things can be true at the same time.

The only individual I’ve seen capture appropriate nuance is Jeff Bezos. He was a Wall Street quant before founding Amazon, so he knows both worlds intimately.

There is still a whole set of products that have not yet been built, which will unlock extraordinary economic gains for society.

We are literally at the beginning of a decade-long buildout. Therefore, it’s impossible to predict how it’ll play out.

The future is never valued correctly when it comes to new technology. Not even by the experts.

Could the stock market crash in 2026 due to AI hype? For sure.

But will AI be enormously beneficial to society over the next decade? Without a doubt.

Was this newsletter forwarded to you? If you liked it, sign up here.

And, if you loved it, forward it to your friends and family.

Thanks for reading,

— Luca

P.S. Want to work together?

1. Get your free AI audit: If you’re still manually doing work that AI could handle in minutes, I'll show you exactly where you're bleeding time and money. In 25 minutes, I'll audit your current workflow and map out your personalized AI transformation roadmap. This is the same process I used to save 15+ hours per week in my own business. 🤖

2. See my complete tool stack: Beyond what I've shared in my AI operating system, I maintain a living document of every AI tool I test, why I use it, and which ones actually move the needle. Access my complete AI toolkit by reaching out here. 💼

That’s it from me. See you next week, Luca 👋

P.P.S Let’s connect on LinkedIn and X.