Why Analysts are Freaking Out About the AI Investment Numbers
I have been building AI systems and using them daily for years, so I understand the genuine technological potential of this breakthrough technology. AI tech is very real and very capable, but what I am seeing in the financial markets is not rooted in higher intelligence at all. Rather, it resembles a speculative mania fueled by cheap debt and fairy-tale projections.
Consider OpenAI: the company is hemorrhaging cash, projecting $44 billion in losses over five years with no profits expected until 2030 at the earliest. HSBC analysis reveals that OpenAI requires an additional $207 billion by 2030 just to remain solvent, driven by astronomical cloud compute costs expected to hit $792 billion annually by that same year. [1] This is not a business; it is a savage black hole that swallows capital and makes it vanish forever.
Then there is the SoftBank loan. SoftBank recently lent OpenAI $40 billion on an unsecured basis at high interest rates with a 12-month maturity. This is a massive gamble that reeks of the same reckless behavior that destroyed WeWork. OpenAI is now missing revenue and user targets, and its own CFO reportedly fears that $1.5 trillion in commitments cannot be paid. [2] Meanwhile, Nvidia is essentially lending money to OpenAI so that OpenAI can buy Nvidia’s own chips, creating a circular risk that benefits no one but the insiders cashing out before the potential collapse. As Wall Street money manager Ed Dowd recently warned, “The real part of this economy is not doing well. Just wait ’til the AI bubble bursts.” [3] I see a house of cards built on promises, not profits.
The Amazon Mirage: Why OpenAI Has No Moat
The bulls love to compare OpenAI to Amazon, arguing that massive upfront investment will eventually create an unassailable moat. But as Brian Dumaine documented in Bezonomics, Amazon built a unique logistics network with economies of scale that are nearly impossible to replicate. [4] Amazon’s moat came from physical warehouses, delivery infrastructure, and a relentless focus on customer experience. OpenAI, by contrast, trained its models on the same public data available to everyone. There is no proprietary data advantage, no physical network, and no barrier to entry. Every AI lab can download the same internet corpus and produce essentially the same AI model, given similar compute.
The first-mover advantage has evaporated. I and thousands of other developers now routinely use Anthropic, DeepSeek, and Qwen instead of OpenAI. Chinese models like DeepSeek deliver superior math and engineering at roughly 1% of the cost of frontier U.S. models. [5] This is not a niche; it is the market speaking. OpenAI’s pricing is impossible to justify when a Chinese competitor offers similar or better performance for pennies. The narrative that OpenAI is the next Amazon is a mirage. Without a real moat, the billions being poured into this company are not investments – they are subsidies for a collapsing revenue model that doesn’t have a clear path to solvency.
SoftBank’s Reckless Bet and the Coming Liquidity Crisis
In my analysis, SoftBank’s $40 billion unsecured loan to OpenAI is a textbook example of financial insanity. The firm is borrowing some of those funds at 8.5% to fund a company that has no discernible path to profitability, repeating the same mistake it made with WeWork on a vastly larger scale. As Jamie K. McCallum explains in Worked Over, the drive to optimize everything with data and round-the-clock labor has created a culture where reckless bets are disguised as innovation. [6] But when interest rates rise and the economy strains under the weight of war and inflation, refinancing in 2027 might become nearly impossible.
With the Persian Gulf conflict driving diesel to $5.60 per gallon and the Fed facing stagflation, the conditions for a liquidity crisis are already forming. [7] If OpenAI defaults, the ripple effects will cascade through the banking system. Private credit markets are already locking up – BlackRock and other firms are freezing redemptions. [3] This is a 2008-style crisis waiting to happen, except the leverage is hidden in unsecured loans and overvalued AI startups. The financial system is not prepared for the contagion that will surely follow when the AI bubble bursts.
The Physical Impossibility of Scaling Data Centers
Even if the money were available, the physical infrastructure simply cannot be built fast enough to satisfy scaling demands. As I explained in my article on the compute crunch, there are not enough transformers, gas turbines, or GPUs to construct and operate the data centers being promised. [8] The supply chain disruptions from the war in the Persian Gulf have crippled semiconductor fabrication – helium, a critical gas for chip manufacturing, is primarily sourced from that region. [7] Without helium, you cannot make advanced chips. Without chips, you cannot build GPUs. Without GPUs, you cannot run AI data centers.
Federal Reserve Chair Jerome Powell himself has noted that the AI data center boom is contributing to inflation, not reducing it. [9] The energy demands are so massive that even if every dollar materialized, the grid as it exists today cannot support the buildout. The eastern U.S. power grid has zero spare capacity. This is not a financing problem; it is a physics problem. The hype ignores the gritty reality of transformers, copper wire, and construction timelines. You cannot will factories into existence with press releases and FOMO investment fundraisers.
History Rhymes: Dot-Com Collapse 2.0
Every bubble narrative insists that “this time is different.” The dot-com era said the same thing about internet stocks. As Autumn Spredemann noted in The Epoch Times, AI investment today mirrors the early dot-com days: capital floods in, and valuations soar on promises of a future that has not arrived. [10] The same pattern is repeating. Companies with no revenue are valued in the billions. Executive compensation is tied to stock price, not earnings. Insiders are often found selling, while retail investors are left holding the bag.
But there is an even darker twist: if superintelligence actually arrived, it would collapse the economy by eliminating millions of jobs, making it impossible for AI companies to earn revenue from a jobless population. At that point, they wouldn’t care, however, since their superintelligent entities could just hack the world’s governments and banks and magic-wand all the the money they need out of thin air, while ruling the world and cancelling all their previoud debts.
The technocrats pushing this agenda are either delusional or malevolent. My advice is simple: avoid AI stocks like the plague. Invest in real assets like gold, silver and land with water rights. In the mean time, prepare for a massive correction that will shake the financial system to its core. The AI bubble is not a sustainable boom; it is a death rattle for a financial system built on debt and deception.
We Don’t Know When, But We Can See The Damage in Advance
I have seen enough financial manias to recognize the signs. The AI bubble is shaping up to be the largest and most dangerous speculative frenzy in history, dwarfing the dot-com crash and the housing bubble combined.
The numbers do not add up, the moats are imaginary, the debt is unsecured, the infrastructure is impossible to scale, and history is rhyming. The only rational response is to step aside, secure your wealth in gold and silver, and wait for the dust to settle.
When the collapse comes, those who prepared will be the ones rebuilding… and there will be empty data centers on sale for pennies on the dollar to anyone who wants to take on such buildings for a different use case.
Personally, I’m waiting for the fire sale on GPUs so I can build a bigger decentralized mini data center of my own, and use it to build free platforms and knowledge solutions for humanity. The sooner the AI infrastructure bubble crash comes, the sooner we can redirect hardware, investment capital and other resources to a better use.
References
- HSBC: OpenAI Requires Additional $207B by 2030 to Remain Solvent. – NaturalNews.com. Ava Grace. November 30, 2025.
- OpenAI Misses Revenue, User Targets As CFO Fears $1.5 Trillion In Commitments Can’t Be Paid. – ZeroHedge. April 28, 2026.
- The Real Part Of This Economy Is Not Doing Well: Ed Dowd Warns ‘Just Wait ‘Til The AI Bubble Bursts’. – ZeroHedge. May 29, 2026.
- Bezonomics. – Brian Dumaine.
- Mike Adams interview with Seth Holehouse – January 31 2025.
- Worked Over: How Round-The-Clock Work Is Killing the American Dream. – Jamie K. McCallum.
- When The Persian Gulf Supply Shock Meets The Warsh Fed: Stagflation & The Coming AI Bubble Bust. – ZeroHedge. May 9, 2026.
- The Compute Crunch: How AI’s Unstoppable Demand is Creating a Hardware Famine for Years to Come. – NaturalNews.com. Mike Adams. February 18, 2026.
- Fed Chair Blames AI Data Centers for Rising Bills, Inflation Concerns. – The National Pulse. March 19, 2026.
- How The AI Bubble Is Being Masked Within Big Tech. – ZeroHedge. Autumn Spredemann. December 29, 2025.
- AI stock melt up sparks fears of 2025 bubble burst. – NaturalNews.com. Willow Tohi. October 17, 2025.
- 2025 10 15 BBN Interview with Cordon . – Mike Adams.
Explainer Infographic
Editorial Cartoon
Read full article here