The hysteria driving opposition to data centers and now AI in general is not a threat to the global economy. China welcomes it, of course, but that won't cause the next great recession.
The entire planet is swimming in a sea of bad debt, and the crisis we're seeing in the debt and bond markets is a harbinger of a much deeper crisis, with AI as a major catalyst.
The current global debt situation revolves around two primary pressures: soaring sovereign debt in major developed economies (especially the U.S.) and escalating financial distress in developing nations. Overall global debt sits near record highs, driven by a combination of high interest rates, persistent post-pandemic budget deficits, and massive capital expenditures on infrastructure, defense, and emerging technologies.
Because central banks raised and sustained elevated interest rates to tame inflation, the cost of servicing existing debt has skyrocketed. Net interest payments have become one of the largest spending line items in the U.S. federal budget, rivaling defense and entitlement programs.
A massive volume of existing low-interest bonds issued during the 2010s is maturing. Third-world governments, the beneficiaries of these low-interest bonds, must refinance this debt at much higher interest rates, creating a vicious cycle of rising interest expenses as these bonds continue to mature.
Countries in Africa, Asia, and Latin America are spending substantial portions of government revenue purely on debt service. When default or restructuring becomes necessary, negotiations are often drawn out due to split debt holdings between traditional Western lenders (like the IMF and Paris Club), non-Western creditors, and private bondholders.
To secure emergency financial aid or IMF bailout packages, nations are forced to slash subsidies, freeze public aid, or raise taxes, frequently triggering political instability and civil protests. It's the Greek sovereign debt crisis on steroids repeated half a dozen times across Southern Africa.
Refinanced loans will crowd out money for home loans, car loans, and small business loans, making them more expensive and likely pricing many middle-class buyers out of the market.
Throw into this mix unrealistic profit expectations and massive spending on AI infrastructure (like data centers and specialized chips), which could lead to sharp market corrections. Unprofitable AI startups close down, tech stock valuations fall drastically, and investors lose billions.
Crash. Boom. Bang. How bad can it get?
Similar to the dot-com crash of 2000, the market will clear out overhyped, low-value products ("workshop") while durable, genuinely useful technologies survive. Even after a crash, the physical and digital infrastructure built during the boom, such as massive computing power, cheaper hardware, optimized open-source models, and specialized tools, remains in place.
Then what? Once the hype quiets down, AI tools fade into the background. Instead of trying to create all-knowing "human replacements," developers pivot to focused, practical applications, like specialized medical tools or automated workflows.
What becomes of companies like OpenAI, Anthropic, and Google that have invested hundreds of billions of dollars into "Gee-Whiz" technology and proprietary AI models? The future of AI is probably closer to the Chinese "open-weight" models that don't have most of the bells and whistles of the proprietary U.S. models, but are much cheaper and easier to adapt into workflows.
Because of strict export controls on exports to China, they're denied the best, most sought-after chips. Chinese labs cannot simply use brute-force computer clusters that cost billions of dollars to achieve progress.
Chinese AI labs aim to achieve 85–90% of frontier capabilities at a fraction of the training and inference cost. While Silicon Valley often focuses on consumer chatbots, speculative AGI timelines, and massive multi-billion-dollar hype cycles, China's official national policy treats AI primarily as a "quality productive force" to modernize the real economy.
China is hardly home free. Breakthroughs by Anthropic and ChatGPT in artificial general intelligence (AGI) could make China's open-weight approach a loser for most companies. But for that to happen, Bernie Sanders and his anti-AI, anti-data center caucus will have to lose at the ballot box, and the hysteria over AI and data centers will have to die down so that the continued build-out of data centers can proceed and the full potential of AI can become a reality.
If a perfect storm of a debt-market collapse and a stock-market correction over poor AI-company earnings happens, the downturn will likely be sharp but relatively brief. The underlying U.S. economy is strong despite the $40 trillion in debt we're carrying. AI companies will also likely switch gears and do whatever it takes to become profitable.
