You Will Not Get an Official Warning

Hey Everyone - Hope you had a great weekend. This morning a story dropped that I think is the cleanest bubble-era signal we have had all year, and it has almost nothing to do with the stock price of any AI company. It is about who inside the government is quietly writing the opposite of what officials are saying in public. Let me walk you through it.

This week:

Let's dive in.

On June 25, Treasury Secretary Scott Bessent gave a speech in New York praising the largest US technology firms for their $750 billion AI buildout this year. He favorably compared the current environment to the dotcom boom. He asked, out loud, "Could we do at least that? Can we do maybe more?" He told a G7 audience shortly after that the only real risk of AI is China getting ahead of us. That is the public voice of the institution charged with financial stability.

This morning, NOTUS published a report obtained from inside the same institution. Career Treasury analysts have prepared a draft report for Bessent, Fed Chair Kevin Warsh, and federal financial regulators, warning that the AI market shares key traits with the dotcom bubble and that AI firms are now more deeply entrenched in the US economy than the dotcom firms ever were. The analysts wrote that "much of the financial system now rests upon AI meeting expectations for productivity gains and profitability." They mapped the ripple. Stock markets, private credit, data center financiers, cloud providers, chipmakers, utilities. A notable detail: fewer retail investors are exposed than in the dotcom era, so the damage from a sustained dip would fall harder on the institutional investors that anchor economic stability.

The report has been completed for weeks. It is sitting in the queue awaiting approval, which is why you have not read it in a press release. A Treasury spokesperson called the findings "unvetted."

The gap between the two documents is the entire story.

I am not going to tell you the bubble will pop. Neither is Treasury, publicly. What I will tell you is that the institution charged with predicting this exact scenario is privately modeling it while the person who runs it is publicly telling the market to lean in. Whichever way the AI cycle resolves, that gap is the truest read you can get on how much certainty exists behind the official confidence. There is not much.

For readers of this newsletter, the update to the framing is subtle. The bubble argument has always been that preparation works either way. If the buildout is real, you want durable skills, real relationships, and independent income because those keep compounding. If the buildout overshoots, you want durable skills, real relationships, and independent income because those keep you standing when other people's positions unwind. The Treasury story hardens that argument. You will not get an official warning before the turn, because the official warning has been drafted and shelved. The people close enough to see the risk are the people who cannot be seen worrying about it in public.

The practical translation for your week is smaller than the story sounds. Do not sell your house. Do not close your Roth. Do not react. Do run a small audit. Where in your income, your career, and your household is the underlying assumption "AI capex keeps going up and to the right"? If your job depends on a firm whose stock is up because it is in the AI supply chain, notice. If your consulting work depends on clients whose growth story is AI-adjacent, notice. If your personal savings are concentrated in the same handful of names that would take the ripple, notice. You do not have to change anything today. You do have to know what you are betting on.

The Bank of England said it in April. The IMF has said it. The Federal Reserve has said it in fragments. Now the US Treasury has written it down and buried it. Preparation is the only signal you get to control.

I have hinted at this for a few weeks and now I want to name it. I am running a small agency that builds and grows newsletters for content creators and business consultants. It is the direct extension of the last seven months of work I have been doing on my own distribution stack, and it started because friends kept asking for it and I kept saying yes.

The economics work in a way I did not expect. My typical customer earns back their investment inside about three months, because a newsletter is one of the few marketing channels where the compounding is measurable and the operational levers are learnable. Most of them are not going to build the same skill I built over half a year, and paying somebody who already did is the cheaper move.

I am starting with friends and family at a discount, which serves two purposes. It gives me the tightest possible feedback loop with people I trust to be honest with me, and it lets me price test in a low-stakes environment before I open it to the broader network. The plan is to leverage my existing network as the second wave, and only open beyond that once the delivery process is truly tight.

There are two reasons I am doing this now, and both matter for the newsletter's overall thesis. First, cash flow. The agency work funds everything else I want to build, which lets me stay patient on the longer bets. Second, the skill compound. Every client I take on makes me better at running my own newsletters, which is the underlying asset the whole business rests on. The agency is not a distraction from the audience. It is a subsidy to it.

The skill is to consciously operate on two time horizons at the same time. A short-term game that generates cash flow this quarter, and a long-term game that compounds hedges, relationships, and durable skills over the years. Most people play one and neglect the other, and both failure modes are expensive.

The short-term game is the cash flow. Client work, agency work, freelance work, the day job. This is the game with a clear scoreboard and a fast feedback loop. If you play only this game, you eat well but you never build anything that outlasts the current contract cycle. In the AI era specifically, the risk is that the short-term game gets repriced or replaced at a pace that would have been unimaginable five years ago.

The long-term game is everything that compounds. In-person relationships with people you would trust in a real problem. Skills you keep sharp with unassisted reps. Physical assets and independent income streams that are not exposed to one platform or one employer. A local reputation that has been earned in the same room over years. If you play only this game, you starve while you are building it, and you cannot fund the very hedges you are trying to accumulate.

The trick, and the reason it is a skill, is that the long-term game is fuzzy in 2026. Nobody knows exactly what a durable career or a durable business looks like in 2030. So the practice is to take the clearest available long-term bet (usually the one this newsletter keeps pointing at, in-person relationships and hard-to-fake skills) and invest toward it with what the short-term game earns. When the long term goes fuzzy enough, revert to a greedy short-term algorithm until the picture clears, then reinvest.

The mistake most people make is treating the short-term as a distraction from the long-term. It is not. The short-term is the funding mechanism for the long-term. The agency work I described above is a short-term game. The audience, the community, the skill stack it feeds are the long-term one. Both matter. Play both.