A sobering piece from Citrini Research jars investors… the dystopian scenario based on AI working as hoped… how and why the dominoes fall… where to invest to keep your money safe
We probably could have figured this out sooner if we just asked how much money machines spend on discretionary goods. (Hint: it’s zero.)
That line is the emotional gut-punch at the heart of a viral essay that hit the markets like a flashbang over the weekend.
It comes from Citrini Research’s THE 2028 GLOBAL INTELLIGENCE CRISIS – a piece written as a fictional “macro memo” from June 2028. It’s not a forecast, not a prediction, and not “bear porn,” as the authors explicitly put it.
But it’s a scenario that traders, investors, and a growing number of strategists are taking seriously because it explores a risk that almost everyone prefers not to model:
What if our AI bullishness continues to be right… so right that it’s actually bearish?
Today, I want to walk you through the essay’s argument in a grounded way – not to scare you, but to help you think clearly about where AI might pressure business models, where it might pressure consumer demand, and where it might concentrate profits.
One more thing worth noting that’s related to this…
Nvidia Corp. (NVDA) delivered strong earnings last night, beating revenue and earnings estimates and delivering better-than-expected guidance – and yet the stock is down 5% as I write Thursday midmorning.
This isn’t a referendum on AI demand. It’s a reminder of how concentrated this cycle has become. When leadership narrows to a handful of names, even excellent results can trigger profit-taking rather than expansion.
Markets eventually start asking whether leadership can broaden. That’s something to keep in mind as we think about where profits concentrate – and where they don’t.
Citrini’s setup… and why it hit a nerve
Citrini frames the essay as a thought exercise and even uses the phrase “left tail risks” – basically, low-probability outcomes with high consequences.
With that caveat aside, the fictional “macro memo” opens with a world that feels plausible enough to be unsettling:
The unemployment rate printed 10.2% this morning, a 0.3% upside surprise.
The market sold off 2% on the number, bringing the cumulative drawdown in the S&P to 38% from its October 2026 highs.
The essay’s point is not “markets will crash exactly like this.”
It’s that AI can succeed and still destabilize the economy – because the economy is not just production and productivity. It’s also wages, demand, consumption, and the velocity of money.
Citrini describes a period where productivity appears to boom, nominal GDP still prints fine, and yet the consumer economy starts to wither.
One of the essay’s most important concepts is what it calls “Ghost GDP”:
When cracks began appearing in the consumer economy, economic pundits popularized the phrase “Ghost GDP”: output that shows up in the national accounts but never circulates through the real economy.
If that sounds familiar, it should – I’ve highlighted a similar paradox in the Digest for months.
From our 11/19/25 Digest:
AI may be fantastic at replacing workers… but it’s terrible at replacing consumers…
From a corporate cost structure standpoint, AI is an efficiency dream.
But there’s a problematic flip side…
AI doesn’t buy anything.
In other words, AI doesn’t drive consumption – and consumption is still nearly 70% of U.S. GDP.
It’s almost as though Citrini read that Digest:
The human-centric consumer economy, 70% of GDP at the time, withered.
We probably could have figured this out sooner if we just asked how much money machines spend on discretionary goods. (Hint: it’s zero.)
The disruptive feedback loop behind this scenario
According to Citrini, the crisis doesn’t begin with a financial accident. It begins with layoffs. “Human obsolescence” layoffs.
At first, those layoffs are bullish. Margins expand. Earnings beat. Stocks rally. And – critically – profits get funneled back into more AI investment.
But then the loop tightens:
AI capabilities improved, companies needed fewer workers, white collar layoffs increased, displaced workers spent less, margin pressure pushed firms to invest more in AI, AI capabilities improved…
It was a negative feedback loop with no natural brake. The human intelligence displacement spiral.
That phrase – “no natural brake” – is why this essay resonated…
In a normal recession, the system self-corrects. Excess cools. The Fed responds. Hiring eventually resumes.
But what if cost-cutting technology becomes the self-reinforcing mechanism?
In a typical downturn, the cure for weakness eventually becomes “stimulus.”
But in an AI-driven downturn, the “cure” may be more automation… which only worsens the problem.
Now, let’s be clear.
This isn’t just about software. Yes, Citrini references pressure in SaaS. But it makes a bigger point…
Software was only the opening act.
Watch out for “friction”
If software is just the opening act, what comes next?
Well, just look for “friction.”
Think of friction as the middleman layer in the economy – the intermediary that exists because human beings are slow, inconsistent, overwhelmed, distracted, or limited.
That intermediary performs a service and charges for it.
AI destroys friction.
If a company’s moat depends primarily on charging for coordination, paperwork, compliance, or workflow management, then AI is coming directly for that layer.
But when friction disappears, it doesn’t just mean “a tool gets cheaper.” It means companies can’t rely on customers automatically renewing contracts… it becomes harder to quietly charge different customers wildly different prices… and the loads of fees embedded in all sorts of industries start getting squeezed.
And over time, some of those businesses simply stop working.
The essay doesn’t keep this abstract. It name-checks insurance, financial advice, tax prep, routine legal work – even real estate agents once AI bots can tap into nationwide property listing data and decades of transaction data.
So, this isn’t just a software problem. It’s a structural intermediation problem.
And then Citrini widens the lens further.
This isn’t only about industries losing pricing power. It’s about what happens when paychecks weaken:
White-collar workers saw their earnings power (and, rationally, their spending) structurally impaired.
Their incomes were the bedrock of the $13 trillion mortgage market – forcing underwriters to reassess whether prime mortgages are still money good.
That’s the macro dynamic.
The question for us, as investors, is what to do about it.
So, what’s the action step?
To help us navigate this, let’s bring in our technology expert, Luke Lango.
Beginning with “defense,” his prescription is simple:
The most vulnerable zones are friction-based digital businesses… especially commoditized SaaS platforms that charge subscription fees for workflows AI can increasingly automate.
We are not saying “sell all software.” We are saying the SaaSmageddon risk is real.
Selective software will survive. Some will thrive.
But broad exposure to generic workflow automation platforms? That risk profile has changed dramatically.
It’s well worth your time to analyze each one of your holdings through this “friction” lens.
Is this company selling true differentiation? Or is it selling process automation that AI might commoditize?
Even if you want to keep a vulnerable position in your portfolio, does it deserve the same weighting?
Shifting to “offense,” here’s Luke with the gameplan:
The key investment insight is that, in the AI Feedback Loop, capital keeps flowing into the physical layer of AI.
Every iteration requires more compute. More chips. More memory. More networking. More power. More cooling.
That’s the layer that gets paid… every single cycle.
So… if we’re right about this risk… and if we’re right that AI will work brilliantly at scale, the money concentrates in the physical supply chain of AI. Chipmakers. Memory suppliers. Networking firms. Power providers. Cooling companies. Foundational model builders.
Those stocks are not disrupted by the AI Feedback Loop – they fuel it.
That’s a critical distinction.
Broad “tech exposure” is not the same thing as owning the backbone of AI.
Which bucket is represented in your portfolio?
Luke also highlights HALO stocks as an important step
To make sure we’re all on the same page, HALO stands for “Heavy Assets, Low Obsolescence.” It was coined and popularized by Josh Brown, CEO of Ritholtz Wealth Management.
The idea is simple – own businesses with real-world assets that AI can enhance but not eliminate.
Back to Luke:
These are companies with physical operations AI cannot replicate, real assets, pricing power, and AI augmentation potential.
Think large retailers with logistics moats, industrial equipment manufacturers, defense contractors, mining companies, energy producers, etc.
AI can enhance their efficiency… but not eliminate their core value proposition. These businesses sit outside the direct blast radius of AI-driven digital disintermediation.
This aligns with what we’ve already seen in the market. Anxiety hasn’t just hit software – it’s spread across office-heavy and intermediary-heavy industries.
Meanwhile, hard-asset themes have attracted attention as perceived shelters.
Even before Citrini’s essay, Luke had been urging investors to look for “choke-point” providers in the physical AI buildout – the materials and components that must exist for the AI economy to function (metals, chips, power, specialized materials, infrastructure).
As I pointed out in yesterday’s Digest, Luke has recently highlighted MP Materials (MP), Lithium Americas (LAC), and Trilogy Metals (TMQ) as examples of companies tied to critical inputs.
But this theme extends well beyond metals – into semiconductors, power infrastructure, networking, cooling systems – the entire AI backbone.
If you want to go deeper, Luke has already begun repositioning around this thesis. If you want to see how he’s translating it into specific stock ideas, you can access his research report here.
One final suggestion…
This Digest is not a call to flip your whole portfolio because a viral essay spooked social media.
It’s a call to stress-test your assumptions.
If AI keeps getting better – and all evidence suggests it will – where do profits concentrate? Where do moats thin? Where does pricing power migrate?
And does your portfolio reflect those shifts?
Make any portfolio changes deliberately and thoughtfully – not based on fear or kneejerk reaction.
At the same time, let’s not ignore the structural shift unfolding in front of us.
On that note, I’ll give Luke the final word:
The real risk isn’t that AI fails. It’s that it works.
Have a good evening,
Jeff Remsburg
