Why the collapse of the middle class is actually bullish for my AI portfolio

There is a palpable sense of dread washing over the global markets right now, and honestly, I find it a little embarrassing. People are looking at the massive sovereign bond selloff in Asia, the persistently elevated oil prices, and the very real threat of runaway inflation, and they are doing what amateurs always do. They are panicking about the cost of living.

But as I watched Goldman Sachs Chief Regional Equity Strategist Timothy Moe on Bloomberg this morning, confidently explaining that the artificial intelligence boom will simply shrug off these macroeconomic headwinds, I felt a profound sense of validation. Moe and I see the world clearly. We understand that a devastating collapse of the middle class’s purchasing power is not a tragedy. It is a filter.

The thesis is quite simple, though apparently too complex for the retail investors currently crying about their grocery bills. High interest rates are a peasant problem. Generative artificial intelligence operates on an entirely different plane of capital destruction, one that is delightfully immune to the cost of government debt.

I had lunch with a senior partner at Goldman last Tuesday at the Mandarin Oriental to discuss this exact phenomenon. We ordered the white truffle risotto and laughed about the ten-year yield curve. He showed me a deck outlining how their high-net-worth clients are actively shorting the Japanese yen to fund Series A rounds for chatbots that can mimic the voices of deceased relatives. That is the kind of capital efficiency you simply cannot find in municipal infrastructure.

The consensus among the family offices in my network is unanimous. We have entirely pivoted our portfolios away from tangible assets, physical supply chains, and human-dependent services. We are moving exclusively into companies that promise to replace customer service representatives with software that occasionally hallucinate new legal precedents.

People ask me how the tech sector can survive when borrowing costs are at multi-year highs. I tell them this is actually bullish. When capital is cheap, executives get lazy. They hire human beings. They provide health insurance. They invest in sprawling middle-management layers to do the work of sending emails back and forth.

When the cost of capital spikes, the true visionaries realize it is time to stop subsidizing the working class and start buying compute. A recession is simply the universe’s way of giving you permission to fire your marketing department.

Let us look at the human element, because people always want to bring emotion into finance. Last month, a portfolio company of ours had to let go of four hundred customer success managers due to the rising cost of debt servicing. The media called it a tragedy. I called the chief executive and congratulated him. By replacing those four hundred salaries with a single enterprise license from OpenAI, he did not just survive a high-rate environment. He achieved operational nirvana.

Look at how inflation affects the average household. The price of milk is up. The price of eggs is up. My driver, a lovely man named Thomas, recently asked for a ten percent raise just to cover his heating bill. I had to let him go, of course. It was a matter of principle. I cannot have my personal payroll expanding while the Fed is explicitly asking for labor market slack.

Instead, I invested his severance into a startup developing autonomous sedans. The sedan cannot currently navigate a four-way stop without requiring remote human intervention, but it also does not complain about the price of natural gas.

Let us address the supposed crisis of elevated oil prices. Yes, a barrel of Brent crude is expensive, and yes, that makes commuting to a physical job incredibly taxing for the median worker. But does a large language model need gasoline? No. It simply requires enough municipal electricity to boil a medium-sized lake.

That electricity is a fixed utility cost, easily amortized across millions of automated interactions. While the physical economy grinds to a halt over the price of diesel, our digital infrastructure is quietly consuming the power grid of a developing nation to generate shareholder value. It is the ultimate hedge against supply chain logistics.

The Federal Reserve can keep rates higher for longer. Jerome Powell can try to cool demand in the labor market. But Powell is trapped in an archaic framework. He thinks raising the federal funds rate will stop inflation because he believes the economy is still made of physical goods moving on physical trucks driven by physical people.

He does not realize that the modern economy is just three tech monopolies trading the same digital advertising dollars back and forth while an AI summarizes the transaction. You cannot squeeze liquidity out of a system that is entirely hypothetical. You cannot cool the unbridled enthusiasm of a board of directors that just learned the phrase predictive analytics.

I was listening to an earnings call for a mid-cap logistics firm just yesterday. The chief executive did not report any actual revenue growth, nor did he mention their crumbling warehouse infrastructure. He simply said the word algorithm fourteen times in three minutes. The street loved it. The stock popped twelve percent in aftermarket trading.

The selloff in global bonds is being driven by the realization that governments actually have to pay back their debt. This is a fundamentally flawed business model. A government issues a bond, makes a promise, and is expected to deliver a yield based on actual tax revenue. It is pathetic.

A Silicon Valley AI startup, on the other hand, issues equity based on the promise that in ten years, human labor will be obsolete. They never have to prove it. They just have to raise another round at a higher valuation before the previous investors realize the software is just a series of nested if-then statements.

The Asian sovereign debt market is currently seeing a historic rout, and frankly, they deserve it. Have you ever tried to have a dynamic, paradigm-shifting conversation with a ten-year treasury note? It is completely static. It yields a predictable five percent. In today’s economy, a predictable five percent is essentially a moral failure.

Artificial intelligence, by contrast, promises either infinite returns or the complete restructuring of human society. That is the kind of binary outcome that gets my blood pumping in the morning.

So I have a small request for the working class and the institutional bond managers alike. Please stop letting your localized financial ruin distract from the very real progress we are making in the cloud. The macroeconomic environment is doing exactly what it is supposed to do: making life impossible for the unoptimized.

My only real concern right now is whether my automated wealth manager is hallucinating my quarterly returns, but I suppose I will let the algorithm worry about that.