/
Published on 25.02.2026
TLDR: Citrini Research published a speculative "memo from the future" exploring a 2028 scenario where AI succeeds so rapidly it creates an Intelligence Displacement Spiral, crashing labor markets and generating "Ghost GDP." The report moved actual markets, contributing to IBM's worst stock drop in 20 years.
Look, I've been in tech long enough to know that every few years someone publishes a "the sky is falling" report, and we all share it around and then move on. But the Citrini Research report, "The 2028 Global Intelligence Crisis," is different. Not because its predictions are necessarily right, but because the market actually listened. IBM dropped over thirteen percent. That's not speculation anymore, that's capital reacting to a thesis.
The core argument is straightforward but uncomfortable: what if AI doesn't empower workers but displaces them so fast that the economic engine stalls? Citrini introduces the concept of an "Intelligence Displacement Spiral" where white-collar knowledge workers get hit first and hardest. Software developers, analysts, consultants -- the exact people building and deploying AI are the first ones it makes redundant. There's a bitter irony there that the report doesn't shy away from.
What's genuinely interesting is the "Ghost GDP" concept. In Citrini's scenario, national productivity numbers look great on paper because AI is cranking out work. But those gains don't circulate in the real economy because AI agents don't buy groceries, rent apartments, or take vacations. You end up with a macro picture that looks healthy while the real economy hollows out underneath. Their numbers are stark: unemployment at 10.2%, labor's GDP share dropping from 56% to 46%, and cascading defaults in private credit markets as prime borrowers in tech hubs lose income.
For architects and engineering leaders, the uncomfortable question isn't whether to adopt AI -- that ship has sailed. It's whether the systems you're building are contributing to a displacement pattern that eventually destroys the demand side of your own market. If your customers are the knowledge workers your AI is replacing, your growth model has a shelf life. The report also flags the private credit angle: $18 billion in software debt downgraded, mortgage market impairment in tech hubs. If you're at a startup funded by venture capital, the financial plumbing matters more than you think.
What the author of the newsletter, Michael Spencer of AI Supremacy, dances around but doesn't quite say is this: the tech optimism narratives are largely coming from people whose wealth depends on AI adoption continuing at full speed. Venture capitalists boosting AI narratives aren't doing it out of philosophical commitment to human progress. The report's timing alongside Anthropic's Enterprise Agents briefing isn't a coincidence either -- it's the market having two conversations at once that fundamentally contradict each other, and nobody is reconciling them.
Link: The Case for Dystopian AI
TLDR: Data scientist Jeremy Ney's analysis shows the AI wealth gap is accelerating at staggering rates -- the top 10% gained $5 trillion in a single quarter while the bottom 50% gained $150 billion. Recent college graduates face the worst job market in recorded history, with AI-exposed young workers seeing a 13% employment decline.
Jeremy Ney's work on American Inequality brings something that's desperately needed in the AI discourse: actual data instead of vibes. His analysis cuts through the optimism-versus-doom binary by showing exactly who benefits and who gets crushed, with numbers that are hard to argue with.
The headline stat is brutal: during the AI surge of the past two years, the top 10% of households saw wealth increase by $5 trillion in a single quarter (Q2 2025), while the bottom 50% saw a gain of just $150 billion. That's a 33x gap. The bottom 50% of Americans own just 1% of all U.S. stocks, so when NVIDIA or Microsoft post record numbers, that wealth flows to an increasingly narrow slice of the population. Ney frames this as the emergence of an "AI Aristocracy" and the numbers support it: the richest 0.00001% of Americans now own wealth equivalent to 12% of national income, up from 4% in 1910. We've passed peak robber baron levels of concentration.
The three categories of at-risk workers Ney identifies tell a story that should concern anyone in tech leadership. Recent college graduates are facing the worst job market in decades. For the first time in recorded history, recent graduates have higher unemployment rates than all other workers. Stanford economists found that young workers aged 22-25 in highly AI-exposed jobs -- software developers, data analytics -- experienced a 13% decline in employment since ChatGPT launched. These workers haven't built the skill depth to pivot, and advanced AI tools can accomplish the same type of work companies once paid $50,000 to $150,000 for.
For architects and team leads, this data should reshape how you think about hiring pipelines and team composition. If junior developers are being squeezed out of the market, your future senior developer pipeline is drying up. The industry assumption that you can always hire fresh talent and train them up is breaking down. Meanwhile, corporate profits are at their highest levels in 80 years while American workers just took home their smallest share of national wealth since 1947. The money exists -- it's just not flowing to the people doing the work.
What Ney is too academic to say bluntly, I'll say: the AI industry is building its growth story on a foundation of labor displacement that it refuses to own publicly. Every "10x developer productivity" pitch is implicitly a "we need fewer developers" pitch. The fact that AI datacenter investment is still only one-sixth of 1880s railroad investment but generates far higher returns for far fewer people tells you everything about the structural dynamics at play. Higher electricity bills for consumers, lower wages, higher profits. That's not a technology story; that's a redistribution story.