You've probably been hearing that we're in an AI bubble. I think that's both loaded and reductive, and I'd like to take some time to help people understand the nuances of the situation we're currently in, because it's deep. To be clear, I am pro AI as a technology and I have an economic interest in its success (and for reasons I'll discuss, so should you), however there is a lot more going on that I don't agree with that I'd like to raise awareness of.
Bubble Talk Misses the Scale of the Bet
AI capital investments are running far ahead of expected returns, and the pace of investment is accelerating. Analysts estimate AI-linked activity drove roughly 40–90% of H1-2025 U.S. GDP growth and ~75–80% of S&P 500 gains. If it wasn't for AI investments, it's likely the United States would be in a recession right now. According to Harris Kupperman of Praetorian Capital “the industry probably needs a revenue range that is closer to the $320 billion to $480 billion range, just to break even on the capex to be spent this year.” It sure sounds like a bubble, however thinking of it as just another bubble would be doing a disservice to the magnitude of the dynamics at play here. To understand why, we have to explore the psychology of the investors involved and the power circles they're operating in.
Political Alignment Is Warping Market Discipline
The elites of Silicon Valley have cozied up to Donald Trump in a way that's unprecedented in the history of modern democracy. They've lined the pockets of his presidential library foundation, supported his white house renovations, paid for his inauguration and provided a financial lifeline for the Republican party. Between Elon Musk, David Sacks, Sriram Krishnan, Peter Thiel and his acolyte J.D. Vance, Trump has been sold the story that AI dominance is a strategic asset of vital importance to national security (there's probably also a strong ego component, America needs "the best AI, such a beautiful AI"). I'm not speculating, this is clearly written into the BBB and the language of multiple executive orders. These people think AI is the last thing humans will invent, and the first person to have it will reap massive rewards until the other powers can catch up. As such, they're willing to bend the typical rules of capitalism. Think of this as the early stages of a wartime economy.
The Strategic Energy Gap
How much of a "war" this is is debatable; if you think we're going to reach peak AI with GPT6, it's a nothingburger, but if you believe AI is the next industrial revolution, that framing is fairly accurate, as failure to adapt would be an economically existential threat. The Chinese will not pull back on AI investment, and because of their advantages in robotics and energy infrastructure, they don't have to. Unlike us, they're able to do it sustainably. US households don't have sufficient surplus income to support consumer robotics at scale and our industrial base is likely insuffient to drive robotics investments at the scale required, and with the administration's crusade against solar energy and America's unwillingness to build nuclear power, we're at a crippling disadvantage in the energy race. We have a head start in AI, but China can train for less, soon they'll have more training capacity, and they'll be able to use their lead in robotics to capture far more economic value from AI than we will. That means the longer the AI race drags on, the more likely China is to beat us. The people in power aren't willing to risk that outcome, and they've been bewitched by the idea of being the only ones to have superintelligence, so they're willing to go all-in to win big and fast.
Just to put the magnitude of China's structural advantage in perspective:
| China | 3,487 | Apr 2025 | Climate Energy Finance monthly update citing official statistics |
| United States | 1,189 (utility-scale) | End-2023 | U.S. Energy Information Administration, Electricity in the U.S. |
| United States | 1,189 GW (utility-scale, end-2023) | 63 GW (developer plans) | >32.5 GW (part of 63) | 7.7 GW | 18.2 GW | EIA baseline at end-2023; 2025 plans from EIA Preliminary Monthly Electric Generator Inventory; solar plus storage = 81% of expected 2025 adds |
| China | 3,487 GW (Apr 2025, total installed) | >=200 GW renewables (policy floor) | 270-300 GW (industry 2025 forecast; ~212 GW in first half) | ~46-50 GW pace in first-half 2025 | n/a | Climate Energy Finance; NEA & CPIA forecasts for solar and wind build-out; first-half 2025 PV adds ~212 GW |
| Domestic market share in China held by Chinese suppliers (2024) | 57% | - | IFR World Robotics 2025 press release |
| Share of global industrial robot installations (2024) | 54% | ~6% (about 34,200 of 542,000) | IFR World Robotics 2025 country breakout |
| Share of global robotics patents granted (2005-2019) | ~35% | ~13% | CSET global robotics patent analysis |
U.S. figures are often utility-scale only, while Chinese data covers the full grid. China is still building faster and more consistently, and the country's robotics ecosystem is expanding just as quickly. The gap matters because AI's value capture depends on having the energy and hardware to deploy frontier models in the physical world.
Rationality Under Questionable Priors
I'm going to say something that sounds a little crazy, but please bear with me: from a geopolitical perspective, what we're doing is a rational play, and depending on how valuable/powerful you expect AI to be and how hostile you expect a dominant China to be, possibly a near optimal one. If you're a traditional capitalist, it probably looks like a bad move to you regardless of your beliefs about AI; you're going to need to put those aside. This is not a traditional economic situation. We're in an arms race, and we're veering into a wartime economy, or at least that's how the powerful view it.
Emergency Stimulus Is Already Baked In
This perspective is critical to understand how things are going to unfold from here. The current administration has demonstrated that they will nationalize businesses and industries in the interest of security, and AI is the most existential security threat on their radar. That's why the administration has been leaning on allies such as Japan and the Arab states to fund infrastructure buildouts. The take home is that if it looks like we're in danger of being overtaken by the Chinese, the president will bully congress into stimulus or quasi-nationalization for key AI players. Even if the Chinese aren't going to immediately overtake us, remember that these people place incredible value on being the first to superintelligence, so if AI momentum flags, it will be met with rapid stimulus.
The dynamic in the valley is that the people at the top know the game already, and they intend to exploit it to its fullest. They know that the president views AI victory as non-negotiable, and he can be relied on to channel an almost bottomless pool of capital into their efforts, through gangster tactics at first, then through a massive bailout architected by throwing the average taxpayer under the bus. This, combined with their view of AI as a step change moment is causing them to invest with reckless abandon. They're too big to fail, and they know it. I expect the US government will repeat its Intel play with at least one AI company, and there will probably be massive cash transfers in the form of inflated government AI automation and surveillance contracts. I wouldn't be surprised if Larry Ellison already has a contract signed in blood for this stashed away somewhere to whip out once he knows he can get away with implementing it. He's talked openly about building a surveillance state to intimidate the populace like a giant digital panopticon, so it's not just idle thoughts.
The Taxpayer Will Foot the Bill
If you want to understand just how hard the average taxpayer is going to get screwed over here, consider this: these tech oligarchs have given Trump millions. Their ethos is embedded in his administration. They've demonstrated they're willing to kiss the ring (among other things). Trump is objectively corrupt, when he pressures congress to pass a funding bill for national security, who do you think the terms he suggests are going to favor? Nationalization is going to make the owners of these companies a ton of money – Intel stock jumped almost 7% in two days after the government took an interest in it. The public is going to eat the risk while oligarchs capture the majority of the upside.
There's a darker side to this – there's no money to cut in the budget for a bailout and our debt is already unsustainable, so it'll require new revenue, and there's only one source that's realistically on the table: income taxes. You might think this will put the Republicans at an impasse, but I think they already have a play lined up. They've been gutting the IRS and talking about reforming the tax code for a long time, but the plan I see them positioning for is sinister. By raising the nominal tax rate at the same time that they reform the tax code, they can engineer in quasi-legal loopholes that the wealthy can take advantage of by design, probably involving digital coins. They get good talking points ("time to tighten our collective belts for the good of the nation," etc) while letting their friends dodge most real responsibility.
You might be wondering, why do this, it seems kind of crazy right? Sadly, it's not; it'll provide the administration leverage over the wealthy, as they can selectively enforce tax laws as a cudgel against dissent. That means the middle class is going to bear the brunt of the new tax burden while the wealthy grow wealthier and more corrupt. I hope this is just my paranoid take, but given the precedent set by the Eric Adams quid pro quo and the Comey, Bolton and James indictments, I have my doubts.
What Happens If AI Under-Delivers?
Returning to the traditional capitalists, I'd like to note that they aren't wrong; this AI push is unsustainable (for us). I'm not sure how long we can run our economy hot and directed before the wheels come off, but my napkin estimate is between 5-10 years, though it's likely we'll lose the political will to keep pushing before that point if the AI transformation is underwhelming and we still have a democracy. To further support the traditional capitalists' position, if AI unwinds at that point having under-delivered, the economic damage will probably be an order of magnitude greater than if we had just let the bubble deflate naturally. This will be exacerbated by the favorable treatment the administration will make sure the Oligarchs receive; we will suffer, they will coast.
Where does all this leave us? For one, you better hope and pray that AI delivers a magical transformation, because if it doesn't, the whole economy will collapse into brutal serfdom. When I say magic here, I mean it; because of the ~38T national debt bomb, a big boost is not enough. If AI doesn't completely transform our economy, the massive capital misallocation combined with the national debt is going to cause our economy to implode.
If you think I'm being hyperbolic calling out a future of brutal serfdom. Keep in mind we basically have widespread serfdom now; a big chunk of Americans are in debt and living paycheck to paycheck. The only thing keeping it from being official is the lack of debtor's prison. Think about how much worse things will be with 10% inflation, 25% unemployment and the highest income inequality in history. This is fertile ground for a revolution, and historically the elites would have taken a step back to try and make the game seem less rigged as a self-preservation tactic, but this time is different. As far as I can tell, the tech oligarchs don't care because they're banking on their private island fortresses and an army of terminators to keep the populace in line.
The Only Way Out Is Through
This future might have been avoidable if the world had slow rolled AI. Unfortunately, I think we're past the point of no return now. The only way out is through. For you AI skeptics, this is going to be rough to hear, but given the suffering that will occur if AI fails, I believe we have a moral imperative to make sure it succeeds. Think of it like WWII, only only instead of planting victory gardens to beat the Nazis, we're building AI apps and finding ways to create the economic value needed to cover the reckless bets being made by the elites. If helping to bail out folly born of hubris makes you feel bad, use that as motivation to push harder. Disrupt the disruptors. Boycott companies that don't demonstrate integrity. The future isn't lost yet, we can still create the world we deserve.
If you want to try and make some money on the dystopian future we're hurtling towards, any tech stock that is a target for stimulus or nationalization is going to be a safe bet. The government is practically manufacturing a ponzi scheme and the losers will be anyone who doesn't own stock. Google is the most likely target for nationalization given all the antitrust leverage against them and their strong tech portfolio. Oracle and OpenAI are also likely targets; I expect no resistance from Oracle, OpenAI is a little harder to read, but they're boxed in so I expect they'll probably end up cutting a deal eventually.
The sad thing is that even if we as a country stand up and build ourselves out of the hole we've been pushed into, we're only enabling the same actors that created this situation in the first place. We can fight back though, we already have the weapon of our liberation: the power of the purse. You're not powerless. Boycott campaigns forced Disney to walk back Jimmy Kimmel's suspension, that was our power in action. If you care about a just world, don't do business with unethical companies. Demand that the titans of tech change, and if they don't, stop feeding them your dollars. Use an adblocker, don't buy anything from their app stores, don't sell anything through their app stores, use Linux everywhere you can and financially support people who build open source software with love and treat their customers with respect.
.png)

