From Boom to Bot-tom: The $40 Trillion AI Fantasy That’s About to Burst!
Had an interesting conversation with my son, Josh, last night. As a first-year Business Management student, his recent lectures on the “AI Bubble” really got me thinking. I’ve been following the surge in AI with great curiosity, which even inspired me to enrol in an AI course at the University of Helsinki – truly fascinating material! I’ve also been tracking this in the media and recently the narrative is shifting!* Picture this: it’s the late 1990s. Everyone is chasing “.com” URLs. Startups bloom overnight. Valuations skyrocket. And then… POP. The Dot‑com bubble bursts. Fast-forward to today: we’re hearing echoes of mania again — but this time for artificial intelligence. Is the Artificial Intelligence era simply tech’s sequel to “Yes we can invest in everything .com”? Or is this something genuinely different and sustainable? Here's my 10c worth – with data, a smidgen of sarcasm, and a dollop of cheeky-ness…
1. What happened last time (the “boom then boom-gone”)
Between 1995 and 2000 the Nasdaq Composite index soared about 400% (roughly five-fold) as internet‐related stocks captured the imagination of money.
The peak came around 10 March 2000: Nasdaq hit about 5,048 points.
From that high, the Nasdaq then plunged about 78 % by October 2002.
The S&P 500, more diversified, still lost around 45–50 % in the same period.
What triggered the pop? A mix of sky-high valuations (many companies “.com” with no profit), speculation, loose capital, and then the real world caught up (i.e., profits matter).
Lessons: When everyone says, “this time it’s different,” it often isn’t. Until… it isn’t.
2. What’s happening now (soaring AI hype)
The AI “boom” is very real. According to the Federal Reserve Bank of Richmond, the share of businesses expecting to use AI in the next six months rose from ~6.3 % to ~14.0 % from Sep 2023 to Sep 2025.
Venture capital sees the trend: in one year, 58 % of all VC investment went into AI companies (per Silicon Valley Bank) — yes, more than half.
The market’s begun to lose their sh*t about valuations: a survey by Bank of America Global Research found 54 % of fund managers believing AI stocks are in a bubble (vs 38 % who say “nah”).
Some analysts warn of a potential wipe-out: one article claimed that a burst AI bubble could wipe out up to US $40 trillion from the Nasdaq‑100 and similar indices.
3. Why many believe the AI bubble is about to burst
A recent study from the Massachusetts Institute of Technology (MIT) found that 95 % of generative AI projects currently yield zero return. Ouch.
The Bank of England warns that equity-market valuations appear “stretched… on some measures comparable to the peak of the dot‐com bubble”.
If the underlying technical or infrastructure bottlenecks occur (e.g., data, chips, power), the promised AI gains may not come fast enough — leaving valuations exposed.
Fundamental mismatches: big spending, big expectations — but until profits follow, the market is relying on templates like “future cash flows ~ magical”.
4. But—it might be different this time
Yes, I said it. There’s genuine reason for optimism too:
Unlike many dot-coms which had zero business model, many AI/tech companies today do have revenue, customers, and scale. For example, some massive firms are investing because they already have product/market-fit.
AI isn’t a niche fad like “underwaterpopuptoaster.com” or “.com pizza delivery”. It’s embedded in cloud computing, edge infrastructure, chips, data. So even if some firms collapse, the infrastructure could remain.
Some analysts argue this is an “industrial boom” (big factories, big data centres, real bills) rather than pure speculation. That suggests the downside may be smaller than the pure internet bubble.
5. My provocative take (with humour): What if the bubble bursts?
Imagine this scenario:
A darling AI chip company that everybody had to have, sees its growth slow. Investors tweet memes of “AI hype so 2024”.
AI service providers who promised “100 % automation by 2025” admit—awkwardly—that humans still review 80 % of the work. Remember this piece I stitched together a while back! https://shorturl.fm/n9GdQ
The market smells trouble. The fund manager survey (54 % see a bubble) turns into “sell the unprofitable AI stuff”.
A domino effect hits broader tech stocks. Suddenly you’re back in 2002-vibe land: “What do you mean the valuations were premised on infinite growth?”
The good news: By the time this is over, some of the best companies will survive (the “Amazon”s of AI). The bad news: A lot of money will evaporate; valuations will reset; dreamers who bought “AI will replace all humans” will own a lot of broken promises.
In short: the hype train could meet a tunnel. If you’re holding tickets, maybe check where the brakes are.
6. What to watch if you’re worried (or curious)
Valuations vs earnings: Are AI-focused firms delivering profit growth, or just revenue promises?
Breadth of market: Is the rally narrow (a few large firms) or wide (many firms)? Narrow rallies often precede trouble.
Use case reality: Projects that invest billions in AI yet deliver minimal improvement = red flags (see the MIT 95 % zero-return stat).
Infrastructure cost: If power/data/chip bottlenecks intensify, the cost side may blow up profit templates.
Investor sentiment: When fund manager surveys show >50 % believe we’re in a bubble (as they do now) that’s a caution.
Macro shocks: Interest-rates rising, or some policy/regulation surprise could trigger the correction.
7. Final word: Are we doomed, or just cautious?
I’ll channel my inner cheeky narrator: No, we’re not doomed… yet. But yes, we should be cautious. The parallels with the dot-com era are striking, but the fundamentals are slightly stronger now. That said: when optimism goes wild (“AI will solve everything, buy everything!”), the downside risk rises.
If you were around in 2000 you remember the glimmering dot-com IPOs, the “we’ll be profitable next quarter” that never came, and the sobering hangover. With AI, we may see a similar hangover — albeit perhaps with a stronger core (since AI does seem to do something). But until the profits catch up, the bubble might be ready to deflate.
So: buckle up. Keep your seat belts on. If you’re in the AI investment train, maybe make sure the emergency brake works.
Comments thoughts welcomed, as always!
MC
