And to round out this week, an article by Peter Wildeford that gives a thoughtful defense of the non-bubblic character of AI today.
Wildeford acknowledges the discussion around whether artificial intelligence represents a financial bubble is complex and nuanced. Prominent figures like Goldman Sachs’ CEO and OpenAI’s Sam Altman acknowledge the presence of a bubble, yet they continue to invest heavily in AI infrastructure. This apparent contradiction reflects a recognition of significant timing risks alongside confidence in the technology’s long-term value.
OpenAI exemplifies this dynamic with rapid revenue growth—from $200 million in early 2023 to $13 billion by mid-2025—paired with substantial projected losses reaching $45 billion by 2028. These figures highlight an unprecedented scale of investment and financial risk, far exceeding losses seen in other high-growth companies. The company’s intricate financing arrangements, involving circular investments with chipmakers and cloud providers, add complexity but are not inherently unsound.
The potential bubble is best understood as an infrastructure bubble, akin to historical episodes like Britain’s Railway Mania or the late 1990s telecommunications crash. In these cases, the underlying technology was transformative, but excessive and overlapping investments created financial instability. Unlike those past examples, AI infrastructure offers more flexibility; data centers and GPUs can adapt to various workloads, reducing the risk of wasted capacity. Also, as of today, the Hyperscalers don't have excess capacity....
The critical uncertainty lies in whether AI capabilities will advance swiftly enough to generate the economic returns necessary to justify these investments. While there is a significant chance of a market correction if growth falters, the prevailing expectation is continued progress and eventual profitability. AI’s demonstrated ability to attract users and generate revenue distinguishes it from previous speculative bubbles tied to unproven business models.
TLDR; The future of AI hinges on its capacity to deliver sustained economic value before financial pressures mount, shaping both market outcomes and broader economic impacts.