Rafomac News, Tech & Trends That MatterThe US stock market’s faith in endless AI profits is facing a serious test. China’s rapidly advancing and affordable AI may trigger the end of America’s dominance—and possibly the AI bubble itself.
Is the AI Boom Just Another Bubble?
The US stock market has leaned heavily on the belief that American tech giants will rake in trillions by owning the future of AI. But this story might be unraveling fast. Some experts argue this AI surge is even bigger than the infamous dot-com bubble of the early 2000s, fueled by hype rather than solid returns.
Ed Zitron, a tech analyst, bluntly states that big tech has run out of breakthrough ideas. Instead, they’re pouring over a trillion dollars into AI, hoping it’s not just a dead end.
Meanwhile, CEOs like Alex Karp of Palantir, a company deeply entwined with government agencies, voice skepticism. AI today charges per token—a ‘token’ being roughly a word processed—which means businesses pay whether the AI adds value or not. They’re frustrated paying for services that sometimes hallucinate facts or steal their trade secrets.
Many companies fear AI providers could turn into competitors by learning from their data. Karp recommends businesses take control by running AI models in-house, using their own data and hardware, avoiding dependencies that can be switched off or compromised.
Why the Traditional Tech Profit Model Fails for AI
Unlike conventional software, where selling to more users costs almost nothing, AI costs rise directly with usage. Every AI query consumes electricity and wears down hardware. In 2025 alone, OpenAI reportedly burned through $20.9 billion, showing that scaling AI isn’t translating into profits yet.
This breaks the golden tech formula of fixed costs plus growing revenue equals huge margins. Investors are starting to catch on, as AI-focused companies delay IPOs and face worsening margins despite increased revenues.
Enter China: The Low-Cost AI Challenger
The real wildcard is China’s rapidly expanding AI sector. While the US spends about $764 billion on AI this year—roughly 3% of its entire economy—China is investing just a fraction, $102 billion, about 0.6% of its GDP. Yet Chinese AI models operate at a fraction of the cost.
For example, a Chinese AI model called GLM completed the same coding task as the US-based Anthropic’s Claude AI at roughly 7 to 12 times lower cost. Chinese companies often open-source these models, making them essentially free for businesses to use.
This means despite the US boasting the smartest AI on paper, most businesses don’t need the top-tier. They need reliable, affordable AI for tasks like customer service—an area where China’s offering is rapidly gaining ground. The US’s trillion-dollar AI spend is, in effect, subsidising China’s cheaper tech ecosystem.
What Could Trigger the Bubble to Pop?
Historical patterns suggest the AI bubble might burst before companies even stop investing heavily in AI infrastructure. Looking back, the dot-com crash occurred a year before major tech investments slowed, showing that investor sentiment shifts faster than spending behavior.
Wall Street watchers believe that when the first big tech company announces slowing AI infrastructure spending, others will follow. According to Goldman Sachs analysts, the market will reward the first hyperscaler to pull back, possibly triggering a cascade.
Credit markets also offer clues. Currently, credit spreads—the extra interest investors demand for riskier corporate debt—are at historically low levels, signaling few fears. But credit markets have been famously wrong before, notably in 2007 ahead of the financial crisis.
Signs Point to Growing Skepticism
Chipmaker stocks, vital to AI hardware, are trading near 15-year highs, echoing peaks before recent market corrections. Meanwhile, companies investing billions in AI infrastructure—Microsoft, Google, Amazon, Meta—aren’t seeing corresponding stock gains. Instead, markets reward hardware suppliers over the giants spending the most.
Additionally, the price of AI compute tokens has dropped nearly 20% since May. Analysts suggest this reflects a shift toward cheaper AI models, which aligns with the rise of Chinese AI alternatives.
But market veteran Michael Bur, who predicted the 2008 crisis, has cautioned that while these trends are telling, their timing is uncertain. Past misfires show AI investors must stay alert yet cautious.
So, What Now?
No one can say for certain when—or even if—this AI bubble will burst. Signs of trouble include capital spending pullbacks, rising credit spreads, and shifts in market sentiment. Most crucially, China’s low-cost AI offerings question the whole premise behind the US’s exorbitant AI investments.
For corporations and investors alike, the choice may soon become clear: adapt to a more competitive AI landscape where cheaper rivals erode margins or face a reckoning. China’s AI evolution isn’t just a tech story—it’s shaping the future of global economic power.
