Part 3 of a 3 Part Research Series
The Difference Between A Technology Failure And A Capital Cycle Failure
In the first two articles of this series, I argued that the most important risk facing the artificial intelligence sector may have very little to do with the technology itself.
The first article explored the reliability problem. Artificial intelligence is becoming increasingly capable, yet hallucinations and verification requirements may prevent it from replacing knowledge workers to the extent currently assumed by financial markets.
The second article examined the economic consequences of that possibility. If AI remains an assistant rather than a replacement, the return on investment available to businesses may be lower than expected. Revenue growth may disappoint. Infrastructure spending may prove excessive. The issue is not whether AI succeeds, but whether the economic value ultimately created justifies the extraordinary amount of capital currently being committed to the sector.
The obvious question is what happens next. History provides a useful guide.
One of the most common mistakes investors make during technological revolutions is assuming that a successful technology automatically leads to successful investments. In reality, the opposite is often true. The greatest investment losses frequently occur in sectors built around technologies that genuinely change the world.
- The railway boom transformed transportation.
- The telecommunications boom transformed communication.
- The internet transformed commerce.
Yet each experienced periods of extraordinary financial destruction because investors overestimated the economic returns available from the technology.
This distinction matters because the current AI cycle increasingly resembles a capital cycle rather than a technology cycle. The technology is real. The question is whether the capital structure built around it is sustainable.
Consider what has occurred over the past several years. Technology giants have committed hundreds of billions of dollars to AI infrastructure. Private equity and infrastructure funds have financed data centres. Private credit has funded projects throughout the supply chain. Venture capital has poured money into AI startups. Public markets have rewarded virtually any company able to position itself within the AI narrative.
The entire ecosystem has become increasingly dependent upon assumptions of future growth.
What happens if those assumptions prove too optimistic? The answer is unlikely to be an immediate collapse in AI adoption. Businesses will continue using artificial intelligence. Productivity gains will continue. New applications will emerge. The more probable outcome is that spending grows more slowly than expected.
That may sound like a minor distinction. It is not.Financial markets are built on expectations.
A company valued on the assumption of thirty percent annual growth can lose a substantial portion of its market value if growth falls to fifteen percent, even though the business itself continues to expand. The same principle applies to entire sectors.
The danger facing the AI industry is not that demand disappears. The danger is that demand fails to justify the infrastructure built to support it. At that point, the conversation shifts from technology to finance.
- Data-centre utilisation becomes important.
- Refinancing becomes important.
- Private credit becomes important.
- Exit opportunities become important.
History suggests that these are often the points at which investment booms encounter difficulties. A useful comparison can be found in the dot-com collapse.
When technology stocks peaked in early 2000, the dominant narrative was remarkably similar to today’s AI story. Investors believed the internet would transform the economy. They were correct. What they failed to appreciate was the difference between technological adoption and shareholder returns.
As the bubble deflated, the Nasdaq eventually fell almost eighty percent. Many companies disappeared entirely. Yet internet usage continued growing throughout the collapse.
The technology succeeded. The capital structure failed.
The distinction is critical because it influences how investors should think about Bitcoin. Many Bitcoin investors instinctively assume that a major market correction would be bullish for Bitcoin. History suggests otherwise.
The first phase of a financial crisis is almost always characterised by a scramble for liquidity. Investors sell whatever they can. Leverage is reduced. Risk assets fall together. The objective is survival rather than optimisation. Bitcoin has never been immune to this process.
During periods of acute financial stress, Bitcoin has historically behaved more like a risk asset than a safe haven. Any AI-driven market correction would likely produce a similar outcome. If technology valuations collapse, private credit experiences stress and financial conditions tighten, Bitcoin would almost certainly be caught in the initial liquidation.
This observation is important because it prevents investors from drawing simplistic conclusions. The question is not what happens during the first phase of the crisis. The question is what happens afterwards. To answer that question, it is worth examining the response to previous market collapses.
Following the dot-com crash, central banks reduced interest rates aggressively. Following the Global Financial Crisis, they introduced quantitative easing. During the pandemic, they expanded monetary support to unprecedented levels.
Each crisis produced a larger intervention than the one before it. The details differed, however the direction remained remarkably consistent. Financial systems built on credit struggle when liquidity disappears. Eventually policymakers are forced to respond.
Some observers argue that this time may be different. The Federal Reserve’s current leadership has expressed a preference for balance sheet discipline and has been notably cautious regarding future monetary accommodation. Yet central bankers do not operate in a vacuum. Their actions are constrained by economic reality. A central bank can ignore falling asset prices for some time. It becomes much more difficult to ignore rising unemployment, widening credit spreads and deteriorating financial conditions.
This is where the AI story intersects with the Bitcoin story.
If artificial intelligence ultimately disappoints relative to expectations, the consequences extend beyond technology stocks. The infrastructure supporting the AI build-out has been financed through a complex web of public equity, private equity, private credit and corporate borrowing. A significant repricing would affect far more than a handful of technology companies.
The resulting pressure on financial markets would inevitably influence monetary policy. Historically, periods of monetary intervention have altered the relative attractiveness of scarce assets. Gold benefited from this process after the dot-com collapse. Bitcoin did not exist. Today the landscape is different.
Bitcoin has matured from a niche experiment into an asset class held by institutions, corporations and, increasingly, governments. Exchange-traded funds have expanded access. Regulatory clarity continues to improve. Discussions surrounding strategic reserves and sovereign adoption have moved from speculation to policy debate.
None of these developments guarantee future appreciation. They do, however, change the context. The investment case for Bitcoin has traditionally been framed around monetary debasement, sovereign debt and the long-term sustainability of fiat systems. An AI-driven capital cycle introduces a different pathway to the same destination.
Bitcoin does not require artificial intelligence to fail. It merely requires that expectations exceed reality. If AI ultimately creates less economic value than investors currently anticipate, a substantial repricing of financial assets becomes possible. Such a repricing would likely expose the fragility of the credit structures supporting the current expansion. Policymakers would once again face pressure to stabilise the system. Liquidity would eventually return.
That sequence is not guaranteed, neither is it unprecedented. In many respects it follows a pattern that has repeated throughout modern financial history. The irony is that Bitcoin’s strongest long-term opportunity may emerge not from a failure of technology, but from a failure of expectations.
Artificial intelligence may continue improving for decades. Businesses may continue adopting it. Productivity may continue rising. None of those outcomes are incompatible with an investment bubble. The internet changed the world while destroying vast amounts of capital. Artificial intelligence may yet do the same.
If it does, investors will once again be forced to distinguish between assets whose value depends upon future growth and assets whose value depends upon scarcity.
That distinction may prove to be one of the defining investment themes of the decade ahead.
