The AI Bubble: Learning from History Without Fighting It

Dated: 7-Oct-2025

(Personal notes — not financial advice)

This write-up isn't a stock tip or recommendation. It's part of my personal investing journal — documenting how I think about market cycles, risk, and timing. Writing forces me to test my reasoning and stay honest with myself. If it helps others think differently about investing, that's a bonus — but as always, your money, your decision.

Timing the Market vs Learning from History

Timing the market is a fool's errand. But ignoring what history teaches us is equally foolish.

Across generations, every major technological revolution has inspired a wave of investment, euphoria, and ultimately—correction. From railroads to fiber optics, from electrification to the dot-com boom, capital has always rushed ahead of utility. The pattern is familiar: build the infrastructure, price in perfection, then rediscover gravity.

When Build-Outs Break Balance Sheets

In the 19th century, railways promised to connect nations and open new frontiers. Capital flooded into track-laying schemes faster than demand could justify. Fortunes were made—and lost—as overcapacity led to collapses and recessions.

The same cycle repeated with early electrical grids and telephony: huge upfront fixed costs, speculative financing, and eventual consolidation.

Fast-forward to the late 1990s. The internet boom saw vast sums poured into fiber networks, servers, and telecom switches. Thousands of unknown infrastructure startups—hosting providers, backbone operators, and "B2B internet" ventures—burned through cash chasing limitless growth. Most vanished quietly, leaving their assets to be picked up by survivors for pennies on the dollar.

Why This AI Cycle Feels Different

Yet the current AI wave diverges in crucial ways.

Unlike the dot-com or telecom eras, today's heavy spenders—Microsoft, Google, Amazon, Meta, and Apple—are profitable incumbents with fortress balance sheets. They are not speculative startups chasing growth at all costs; they are cash-generating platforms redirecting profits into AI infrastructure.

Moreover, these same firms are also the first beneficiaries of AI productivity. They can get more done with fewer people. AI is both an investment and a cost-reduction engine.

In other words, the firms pouring billions into AI are also the first to profit from it operationally. That dual position makes this cycle unique—and potentially more durable—than past bubbles.

Why I Wrote This: Capital has started to Loop

Even though this AI cycle feels different, I felt compelled to write this because the underlying pattern of excess investment feels uncomfortable.

The companies building the large language models—OpenAI, Anthropic, Perplexity, and others—are not the ones with the balance sheets to fund the colossal infrastructure they've committed to. They rely almost entirely on external capital to pay for data-center build-outs, GPUs, and power contracts worth tens of billions.

Soon after, AMD signed a multi-year deal with OpenAI, supplying its Instinct MI450 chips and granting OpenAI an option to acquire up to 10% of AMD through warrants. The announcement lifted AMD's stock, but again, it's another example of the same circular flow of capital masquerading as expansion.

It's not that these arrangements are dishonest—but they are Synthentic. The same dollars move around the ecosystem, recorded as growth by each participant. But with OpenAI still deeply unprofitable and without the cash flow to honor hundreds of such supplier commitments, the system depends on an uninterrupted stream of new funding.

My Plan of Action

I don't believe we're at the point of collapse. The fragility lies in private markets, not yet in public equities.

For now, the risk sits with venture funds, sovereign investors, and the large tech backers underwriting the ecosystem. But this won't last forever. Eventually, these private AI builders will need new funding sources—and the only pool left will be public investors.

That moment—when companies like OpenAI, Anthropic, and others begin floating IPOs or raising money directly from public markets at inflated valuations—will be my signal to act.

That's when I plan to start reducing equity exposure and build larger cash buffers—not because I expect an immediate crash, but because that capital rotation will mark the point where risk shifts from private hands to the broader market. When that happens, history suggests the next phase isn't expansion—it's digestion.

The correction, when it comes, will likely drag the broader market down—and that's when I'd rather be sitting on cash, ready to lean in when panic turns into opportunity.