In the previous essay, I argued that several apparently unrelated developments form a single pattern: the quiet erosion of professional value, the absence of visible productivity gains, rising political volatility and a growing dependence on housing wealth to maintain living standards. Together, these point to a slow but profound shift in where value now sits in the economy.
There is, however, a further implication that deserves separate treatment. The pattern described there is not merely uncomfortable. It is structurally unstable.
To see why, it helps to step outside contemporary economic debates and look instead at long-run historical dynamics. In particular, the work of Peter Turchin, who has studied the rise and fall of complex societies, offers an uncomfortably good fit to what we are now observing.
Elite reproduction, not unemployment, is the fault line
Turchin’s central claim is that societies become unstable not primarily when people become poor, but when elite reproduction breaks down. Elites, in his usage, are not simply the wealthy. They are the people who expect status, security and influence to pass reliably from one generation to the next.
For several decades, advanced economies managed this quietly. A growing professional class was absorbed into cognitively demanding, high-status work. University expansion, credential inflation and the growth of symbolic white-collar jobs allowed large numbers of people to be “elite enough”. Not everyone rose, but enough did to keep the system coherent.
The developments described in the previous essay undermine this mechanism.
AI does not primarily eliminate jobs. It eliminates scarcity in cognitive tasks. When scarcity goes, so does bargaining power. Professional work does not disappear; it becomes cheaper, more modular and more easily substituted. This is already visible in translation, journalism, design and analysis. Benchmarks such as GDPval suggest it will not remain confined there.
This matters less for people at the bottom than for those in the middle and upper-middle. The problem is not starvation. It is the loss of trajectory.
Why reassurance does not work
Much current commentary focuses on adaptation. People are told to retrain, to become more flexible, to learn to work alongside AI. Parents are encouraged to reassure their children that they will be fine even if their careers look different from those of previous generations.
This advice may be psychologically sound. It is not socially stabilising.
Elites do not optimise for resilience. They optimise for continuity. They want to know that if they invest in education, networks and effort, their children will remain within the elite. The question quietly circulating among professionals today is not “will my child survive?” but “will my child fall out of our class?”
That question increasingly has no reliable answer.
What is now rewarded in the economy is not a particular subject choice, but a set of meta-traits: adaptability, tolerance for ambiguity, the ability to operate under underspecified conditions. These are not evenly distributed, not easily teachable and not reliably transmissible through money or schooling. They are a form of intelligence, not a plan.
From a Turchin perspective, this is explosive. Societies do not remain stable when success depends on rare personal traits rather than institutional pathways.
Why common policy responses miss the point
This also explains why many proposed solutions feel misaligned.
Universal basic income addresses subsistence risk. The emerging problem is not subsistence. It is status collapse in the presence of continued labour demand. If cognitively prestigious work loses value while embodied, lower-status work remains, a flat income floor does not resolve the insult. It formalises it.
Retraining assumes there is a new ladder. In many cases there is not. There are sideways moves, downward moves and contingent moves, but few paths that preserve the status once attached to professional identity.
Even attempts to restrict AI or carve out “human-only” domains reflect this dynamic. They are not primarily about safety or ethics. They are attempts to re-freeze scarcity in elite professions to preserve predictability. Historically, such freezes buy time but increase pressure.
Housing as a temporary stabiliser
In the previous essay, I described how housing wealth has absorbed much of the shock so far. From a Turchin perspective, this looks less like resilience and more like delay.
A single generation has used rising property values to compensate for eroding professional income. This has maintained consumption and social calm, but at the cost of future flexibility. Crucially, this mechanism is not available to the next generation.
When elite reproduction already looks uncertain, and the asset that smoothed the last transition becomes unavailable or unstable, expectations reset abruptly rather than gradually. That is the point at which societies tend to move from quiet adaptation to open conflict.
Why this points to political turbulence
This framework also clarifies the current political landscape. Movements often described as populist or irrational draw heavily on people who did what they were told and did not get what they were promised. They are not reacting to poverty. They are reacting to downgrading.
When mainstream politics responds by insisting that the economy is healthy, employment is high and the future is bright, it inadvertently confirms the sense that loss is being denied rather than addressed. At that point, the appeal of restoration narratives becomes intelligible, even if the narratives themselves are false.
A final note on AI safety
Some figures within the AI industry, including Dario Amodei, have begun to warn that technological capability is outpacing institutional readiness. This concern is not misplaced. But it often focuses on misuse, alignment and extreme scenarios.
From a historical perspective, the more immediate danger is more mundane. It is not that AI systems will do something catastrophic. It is that they will quietly invalidate the social mechanisms that once tied work, status and stability together, faster than new ones can form.
The uncomfortable conclusion
The pattern described in the previous essay does not converge on a new equilibrium by itself. It creates growing numbers of downwardly mobile elites, blocked intergenerational pathways and political movements driven less by ideology than by wounded expectation.
Historically, societies in this position do not drift gently into a new model. They either find ways to reallocate status and continuity at scale, or they experience rupture.
Understanding this does not tell us what will happen. But it does tell us why so many current responses feel inadequate. They are trying to stabilise a system whose core reproductive logic is already failing.
Which raises the question I cannot avoid on this blog: if AI is not merely another technology but a force that breaks the normal relationship between effort, competence and status, are we heading towards a single “last revolution”, or a sequence of smaller revolutions and counter-revolutions as societies repeatedly attempt to re-freeze a world that will not hold still?