China Is About To Pop The AI Bubble

The Comparison of AI Development Between the U.S. and China 00:10

"The AI buildout relative to the TMT buildout of 1999 to 2000 is multiples, even as a percent of the economy."

• The conversation highlights that the current AI boom may be larger than the internet bubble of the late 90s, suggesting a significant economic transformation in progress.

• The U.S. stock market's current stability relies heavily on the narrative that American technology companies will generate massive profits due to global reliance on their innovations.

• Meanwhile, there is growing skepticism around this narrative, as competition from other countries, particularly China, is increasing and could undermine the U.S.'s technological dominance.

The Impact of Decisions on AI Companies 01:01

"A letter was sent to a company in San Francisco... the most advanced AI in the world was shut down."

• A significant event occurred when the U.S. government ordered a major AI firm to restrict its advanced models from operating outside the U.S., which affected not just China but also several Western nations.

• This incident reflects a concerning trend as countries reconsider their dependencies on America's technological infrastructure amidst fears of information control and competition.

• Various European nations have expressed intentions to halt contracts with American AI companies, indicating a shift towards independence and alternatives that may be more cost-effective.

Growing Distrust in AI Value and Applications 03:00

"Fundamentally, large language models are not the future... they’ve run out of hyper-growth ideas."

• A critical viewpoint is introduced regarding the sustainability and future of AI technology, pointing out that tech companies may be diverting attention to AI due to a lack of innovative ideas.

• This discussion questions the profitability of investing in AI, suggesting that companies are currently overspending in an area with uncertain returns and practicality.

• Leading figures in the industry express skepticism about AI's true potential, cautioning that operational effectiveness cannot be reliably measured, raising concerns about corporate investments in AI.

The Risks Associated with AI Adoption 07:04

"The fear for all these CEOs is that when your company uses these models, your data flows through them..."

• The apprehension among corporate leaders is centered around data security and the potential for AI systems to become competitors by mining proprietary information.

• Many corporations are investing substantial amounts in AI technologies, yet they lack clarity on the return on investment and are risking exposing their critical business models to external AI providers.

• This results in an awkward position where companies feel they might be inadvertently training their competition while paying for uncertain benefits.

The Problems with AI Companies 11:04

"Even if every company in America trusted these AI companies, the money still does not make sense because the business model is broken in a way we've never seen from tech companies."

• The first problem with AI is a pervasive lack of trust among companies, and the second, even larger issue, lies in the fundamental business model of AI itself. Traditional software businesses operate on a model that allows them to profit from a base cost spread over new customers, where new sales often translate into significant profit without additional costs.

• However, the AI model disrupts this norm by incurring costs with each query or operation due to the electricity and resources it consumes. This implies that, contrary to traditional software, the addition of more customers leads to an increase in costs rather than a decrease in operational expenses.

The Financial Struggles of AI Companies 13:05

"OpenAI burned $20.9 billion in 2025."

• OpenAI exemplifies the struggles faced by AI companies, having reported massive losses that raise concerns about their long-term viability. Their financial trajectory indicates that as they scale up their operations, their profit margins continue to deteriorate rather than improve.

• Furthermore, these companies face a crucial realization: as technology advances, the costs associated with running AI models are escalating with each iteration, leading investors and insiders to question the potential for sustainable profit in the AI space.

Market Assumptions and Global Competition 16:34

"The profits will come, it's going to be the American companies that will make the profits because the world has no other option."

• At present, there is a prevailing assumption that American companies will dominate AI profitability due to the lack of alternatives. However, the emergence of competitive models from countries like China may challenge this narrative.

• China's AI investment is significantly lower than that of the US in proportion to their economy, highlighting a strategic advantage in deploying cost-effective AI solutions. This situation poses a complex dilemma for the American tech giants who may not achieve the expected returns on their considerable investments in AI.

AI Competition Between the US and China 18:35

"The US still has the smartest AI in the world, but if you ask the question that every business is asking, do I need the fastest AI model to manage my business? The answer is no."

• The United States leads in the development of advanced AI technology, yet many businesses only require sufficient AI capabilities for routine tasks like customer service. This dynamic indicates that China may be gaining ground in the AI race through more cost-effective strategies.

• While the US spends heavily on pioneering AI research, China capitalizes on existing models, using a method that allows them to create effective AI with significantly lower investment. Essentially, they replicate and compress advanced AI models, which reduces costs and allows broader usage.

The Economic Impact of AI Spending 19:11

"Every dollar of US AI spending is kind of like a donation to the Chinese AI industry."

• The United States invests billions in AI research and development, which inadvertently benefits China as they utilize these findings without incurring the same expenses.

• Chinese companies, including those not primarily focused on AI, are now able to produce competitive AI technologies mirroring those of major US firms. This raises questions about the valuations of traditional tech companies, as their market dominance may not be sustainable in the face of reducing costs from international competitors.

Potential Market Bubble Triggers 21:47

"According to data, the bubble could pop a lot sooner than when corporations stop their capital expenditures."

• Historical analysis reveals that market corrections do not always align with corporate spending patterns. For example, during the dot-com bubble, companies continued to spend on infrastructure long after stock prices had begun to collapse.

• The tipping point for potential market change may come from minor indications during earnings calls, where CEOs express caution regarding investment spending. The first company to signal reduced capital expenditures may set a precedent, leading others to follow suit, which could catalyze a broader market downturn.

Current Signals and Bond Market Indicators 24:11

"When data center debt stops being issued, that will be when it's bedtime for this industry."

• Analysts warn that the bond market serves as a critical indicator for economic stability, focusing on promised interest payments rather than speculative growth narratives that dominate the stock market.

• A significant widening in credit spreads can forewarn of economic distress, with history showing that tight spreads during periods of hidden risk can delay recognition of potential crises. Understanding the relationship between risk perceptions and market behaviors is vital in identifying true economic conditions.

Valuation Concerns Around Tech Stocks 27:50

"Chip stocks are trading at the top of their 15-year valuation range, basically at the same peak they hit before the 2024 correction."

• Current market conditions suggest that technology stocks, particularly in the semiconductor industry, are being evaluated at inflated prices based on optimistic future growth scenarios. Historical patterns indicate that such peaks are often followed by corrections.

• The market’s perception of semiconductor valuations suggests that significant capital has already been assumed to be realized, raising questions about sustainability and future profitability in these sectors.

Analysis of AI Stock Performance 28:21

"The AI winners are the companies selling the chips and the equipment, while hyperscalers like Microsoft, Google, and Amazon show little stock value growth."

• Recent analysis divides AI stocks into three groups, highlighting that companies providing essential technology, such as chips and equipment, are experiencing stock increases of up to 200%.

• In contrast, hyperscalers, which are major players in the tech industry, are not seeing similar financial rewards despite massive spending on AI development.

• This disparity indicates a market sentiment where investors are skeptical about the returns on investments made by these larger tech companies.

Declining Value of AI Tokens 29:10

"The price of AI is down almost 20% from its high in May, despite the largest AI build-out in history."

• A significant indicator of the AI sector's health, the Silicon Data LLM token expenditure index, reveals a troubling trend with a notable decline in AI-related token prices.

• The drop in prices during a time of rapid technological expansion raises questions about future demand and pricing strategies within the AI marketplace.

• Bloomberg suggests that the decline may be due to a shift in demand towards cheaper models, signaling possible market saturation or changing consumer preferences.

Uncertainty Surrounding the AI Market 30:10

"The real answer to how long it will take for the AI bubble to pop, if ever, is that no one knows."

• The ambiguity in AI market trends suggests that predictions regarding a potential bubble burst are speculative at best.

• Data trends indicate early signs of instability, but past predictions about market crashes have not always materialized, casting doubt on current forecasts.

• The lack of clear signals means that investors should remain cautious while observing further developments within the sector.