Few themes have captured investor imagination like artificial intelligence. Over the past two years, AI has transformed from a niche frontier into the centerpiece of virtually every market conversation. With billions of dollars flowing into AI startups, staggering valuations for chipmakers, and the S&P 500 increasingly driven by AI-linked names, it’s natural to ask a simple but uncomfortable question: Are we in an AI bubble?

As someone who spends a lot of time analysing market cycles, we see signs of both exuberance and genuine economic transformation. The difficulty is separating hype from reality. So, in this piece, we want to break down where we genuinely stand today.

Signs of Speculation: The Ingredients of a Potential Bubble

Let’s start with the uncomfortable part: yes, some indicators of speculation are clearly visible.

  1. Sky-High Private Market Valuations

Valuations for private AI startups have surged at a pace that feels eerily familiar. We’ve seen an explosion in unicorns — startups valued above $1 billion — and many of them have reached that status with minimal revenue. Investors are aggressively piling into anything labeled “AI,” reminiscent of earlier speculative episodes:

  • The late-1990s dot-com mania
  • The SPAC boom of 2020–21
  • Even the iconic tulip mania often invoked in these comparisons

While the technologies differ, the pattern is similar: a flood of capital chasing big promises.

  1. AI’s Outsized Role in Public Market Valuations

Roughly 40% of the S&P 500’s market capitalization is now tied, directly or indirectly, to AI expectations. That level of concentration can be concerning. When a single narrative commands such a large share of market value, there’s always a risk that sentiment—not fundamentals—drives prices too far, too fast.

  1. Venture Froth and IPO Volatility

In venture capital circles, each funding round for AI companies often comes with a massive mark-up compared to non-AI peers. We’re also seeing newly listed AI-centric IPOs and SPACs with dramatic first-day pops—usually a hallmark of investors chasing the next big thing rather than sober valuation math.

This doesn’t mean a bubble has already burst—but it does mean the ingredients are forming.

The Other Side of the Story: Real Earnings, Real Growth

Now here’s the part many overlook: AI’s current rally is not built solely on hype.

The biggest players in this ecosystem — the cloud hyperscalers and chipmakers — are generating real, measurable, and accelerating revenue growth.

  1. Nvidia: The Clearest Example

Nvidia’s stock performance often dominates the headlines, but the more important story is this: its profits have risen even faster than its share price. Demand for AI chips continues to outpace supply, and the company’s free cash flow generation has skyrocketed.

This isn’t speculation. It’s earnings.

  1. Hyperscalers Are Turning AI into Cash Flow

Companies like Microsoft, Amazon, Meta, and Google are not just experimenting with AI — they are monetising it right now.

Across cloud services, enterprise software, developer tools, and AI infrastructure, we’re seeing:

  • Tens of billions of incremental revenue from AI-related services being added every quarter
  • Strong free cash flow growth tied to AI adoption
  • Expanding pipelines in AI-enabled productivity, security, and computing solutions

This is fundamentally different from the dot-com era, when many companies had “eyeballs,” but no business model.

  1. The Market’s Leaders Are Earning Their Valuations

If AI stocks in the public market were soaring without profits, the bubble argument would be much stronger. But so far, the rally has been supported by:

  • Revenue growth
  • Margin expansion
  • Demonstrable cost efficiencies
  • Real customer demand

In other words: prices have risen, but so have earnings.

Where We Stand: Not a Bubble Yet — But Risks Are Building

After analysing the data, I don’t believe we are currently at bubble-peak conditions. Instead, I’d frame the situation like this:

We are in the early stages of a technology supercycle, but speculation risk is rising.

A bubble could easily form later if:

  • Valuations detach from earnings
  • Investors chase unproven AI startups
  • Growth expectations become unrealistic

But today, the public-market leaders — the companies actually generating returns — remain grounded in strong fundamentals.

How I’m Positioning: Stay Invested, Stay Selective

Given this backdrop, my approach is straightforward.

  1. Stick with companies that have proven business models

The winners so far have clear revenue engines, access to capital, and unmatched scale. They are best positioned to commercialise AI sustainably.

  1. Be cautious with speculative or lesser-known AI names

Some narratives sound compelling, but without:

  • A working product
  • A path to profitability
  • A durable competitive edge

they’re simply stories — not investments.

  1. Focus on who will ultimately capture lasting value

Every major tech cycle creates huge expectations early on, but over time the market decides who the real winners are. History shows:

  • First movers don’t always finish first
  • The best returns often come from companies that enter later but execute better
  • Consolidation tends to favour firms with deep moats, not shallow hype

So my focus remains on quality, durability, and long-term value creation rather than chasing every new AI trend that flashes across social media.

Conclusion: Exuberance Is Rising, But the Foundation Is Real

So, is AI in a bubble?

Not yet — but we are moving closer to conditions where one could form.

Right now, the market is being driven by:

  • Genuine earnings power
  • Massive productivity gains
  • Real customer demand
  • Structural changes in computing and software

Yes, there is hype. Yes, private market valuations look frothy. But at the core of today’s AI boom is a set of businesses generating extraordinary profits from transformative technology.

As long as we stay disciplined — focusing on fundamentals, ignoring noise, and identifying which companies truly capture value — the current AI cycle offers more opportunity than danger.

And that, in our view, is the real story.