The debate around artificial intelligence investing has become the defining market argument of the decade. Bulls see AI as the next electricity or internet revolution — a structural force capable of reshaping productivity, corporate margins and global growth. Bears argue the market has entered speculative territory, with valuations detached from commercial reality and echoes of the dot-com bubble everywhere.
The truth is more nuanced: parts of the AI market almost certainly exhibit bubble characteristics, but the broader AI investment theme still looks like a durable structural growth story.
🟢 Why investors think AI could be a bubble
There are genuine warning signs.
A small cluster of mega-cap technology companies now dominates global equity performance. Capital expenditure from hyperscalers is exploding into the hundreds of billions of dollars annually, with investors assuming future profits will justify unprecedented spending.
BlackRock noted that ‘AI’s buildout is happening at potentially unprecedented speed and scale,’ while also acknowledging that ‘investors have started to fret about equity valuations and whether an AI bubble is forming.’
UK-based BNY Investments fund manager Jon Bell argued that ‘rapid, outsized AI investment has created bubble dynamics in equity markets.’
There are several reasons for caution:
- Semiconductor and AI infrastructure stocks have massively outperformed broader markets.
- Valuations in some AI names imply years of flawless execution.
- The market narrative is increasingly concentrated around a handful of winners.
- AI monetisation still lags the scale of infrastructure spending.
The parallels with the late-1990s internet boom are obvious. Then, as now, investors were convinced a transformational technology would change everything — and they were right. The internet did reshape the global economy. But many early winners disappeared after valuations collapsed.
That historical analogy matters because technological revolutions and investment bubbles can coexist.
🚀 Why the structural growth case still looks compelling
The counterargument is equally powerful: unlike many speculative booms, the AI surge is already generating real revenues, real productivity gains and real infrastructure demand.
Goldman Sachs recently argued Wall Street is increasingly treating AI as a ‘defensive trade’ rather than speculative momentum, supported by hyperscaler capital expenditure growth of roughly 38% year-on-year.
BlackRock remains ‘pro-risk’ on AI-linked equities and sees the theme as ‘the main driver of US equities.’
Mike Seidenberg of Allianz Technology Trust (ATT) described AI as one of those technology transitions that ‘come once every 12–15 years’ and ‘tend to create significant value over the cycle.’
Read Mike Seidenberg’s most recent commentary
Royal London Sustainable World manager Mike Fox also dismissed concerns that AI investment is simply speculative excess, arguing that the scale of corporate spending is rational given the productivity potential.
🌍 The most persuasive evidence supporting the structural growth thesis includes:
1. AI spending is backed by cash flow
Unlike many dot-com companies, today’s AI leaders are enormously profitable businesses with fortress balance sheets.
2. AI is becoming infrastructure
AI increasingly resembles a foundational utility layer — like cloud computing or electricity — rather than a niche software trend.
3. Productivity gains are already visible
Enterprises are deploying AI across coding, customer service, legal work, design and research workflows.
4. Governments are strategically committed
The US, China, UK and EU all see AI leadership as economically and geopolitically critical.
5. Second-order winners are emerging
The opportunity is expanding beyond chips into power infrastructure, cooling systems, networking, industrial automation and software platforms.
🧠 Investment reality: not a bubble, but possibly a bubble inside a megatrend
The most likely outcome is not a complete collapse in AI investing — but a separation between sustainable winners and speculative excess.
That is exactly what happened after the internet boom:
- Amazon (AMZN) survived and thrived.
- Cisco (CSCO) stagnated for years despite being essential infrastructure.
- Thousands of speculative companies disappeared.
The same pattern is likely in AI.
👉 Investors should expect:
- Periodic violent corrections
- Extreme volatility
- Rotation away from overhyped names
- Expanding opportunities outside mega-cap tech
But the broader secular trend remains intact.
📈Six Best AI Stocks to Own
| 1. Nvidia (NVDA) | Price: $222.32 | Market cap: $5.38tn |
The core infrastructure provider of the AI era.
Nvidia still dominates high-performance AI chips, software ecosystems and data-centre acceleration. Its biggest risk is valuation compression, but it remains the clearest ‘picks and shovels’ AI leader.
Why own it
- Dominant GPU market share
- Massive pricing power
- AI software ecosystem lock-in
- Strong free cash flow generation
| 2. Microsoft (MSFT) | Price: $423.54 | Market cap: $3.15tn |
The safest broad AI compounder.
Microsoft has embedded AI across enterprise software, Azure cloud infrastructure and workplace productivity tools. It benefits whether AI becomes consumer-facing or enterprise-driven.
Why own it
- AI monetisation already underway
- Diversified revenue base
- Azure cloud leadership
- Strong balance sheet
| 3. Broadcom (AVGO) | Price: $420.71 | Market cap: $1.99tn |
An underappreciated AI infrastructure winner.
Broadcom supplies custom AI networking and connectivity chips critical for hyperscale data centres.
Why own it
- Beneficiary of AI data-centre buildout
- Strong recurring software revenues
- Attractive cash generation
- Less hype than pure-play AI stocks
Read Sharesify’s Broadcom stock analysis
| 4. ASML (ASML) | Price: €1,258 | Market cap: €484.82bn |
The monopoly powering advanced chip manufacturing.
ASML’s lithography machines are essential for producing leading-edge semiconductors.
Why own it
- Near-monopoly technology
- Structural semiconductor demand
- Long-term AI manufacturing bottleneck exposure
| 5. Alphabet (GOOG) | Price: $393.11 | Market cap: $4.79tn |
The cheapest mega-cap AI platform.
Alphabet combines AI leadership with dominant search, cloud and advertising businesses. The market still underestimates how deeply AI can enhance its ecosystem.
Why own it
- Massive AI research capability
- Attractive valuation relative to peers
- Strong cash generation
- Embedded global user base
| TSMC (TSM) – US ADRs | Price: $395.95 | Market cap: $1.82tn |
Reliable broad infrastructure AI compounder.
Combines crucial manufacturing expertise in advanced microchips, with strong visibility into AI demand, supported by rising hyperscaler capital expenditure.
- Industry leading manufacturing capacity
- Dominant market share
- Rising growth and margin metrics
🚫 Key Risks Investors Should Watch
| Risk | Why It Matters |
| Valuation compression | Even great companies can deliver poor returns if bought too expensively |
| Regulation | Governments may restrict AI deployment or data usage |
| Energy constraints | AI infrastructure requires enormous electricity supply |
| Monetisation delays | Revenue growth may lag infrastructure spending |
| Market concentration | Overdependence on a few mega-cap names increases systemic risk |
🟢 The UK Angle
The UK remains one of Europe’s strongest AI ecosystems, especially in fintech, healthtech and university research clusters. A recent academic study found London accounts for more than 41% of UK AI entities, highlighting both strength and concentration risk.
For UK investors, the challenge is that many of the dominant AI beneficiaries remain US-listed. However, London-listed infrastructure, datacentre, engineering and energy firms may increasingly benefit from second-order AI investment flows.
💡Conclusion
AI investing today resembles the railway boom, electrification era and internet revolution more than a temporary fad.
That does not mean current valuations are universally rational. Some companies will fail. Some stocks may fall 50% or more during future corrections.
But the underlying economic transformation appears genuine.
👉 The critical distinction for investors is this:
- The theme is probably not a bubble.
- Parts of the market probably are.
The winners over the next decade are likely to be companies with:
- Durable competitive advantages
- Real cash flows
- Infrastructure relevance
- Scalable AI monetisation
- Sensible valuations relative to long-term growth
👉 In other words, the AI era may ultimately reward disciplined stock selection far more than indiscriminate enthusiasm.
Disclaimer: The author Steven Frazer has a personal interest in Nvidia and Broadcom.
You might also like:
