Res Politica Posthumana - Singularity & Scenarios (4/5)

January 2025 

Conjecture 1 – Value-Based Erosion (Orthodox Marx Outcome)

In this first scenario, the economy actually gets better at producing things. Output keeps climbing. The problem is that we lose a reliable way to turn all that productive power into broad purchasing power and steady profit growth. You end up with what you might call “productivity without value”: more gets made, but the connections between wages, demand, and social stability start to fray. And it’s not because anyone is trying to break the system. It’s because the incentives keep pushing people and firms toward choices that make sense for them individually, but add up to something destabilising overall.

Through a game theory lens, it looks like a Red Queen race that people talk about as innovation. Every company faces the same basic choice: automate aggressively or fall behind. Slowing down sounds nice in theory, but in practice it’s not really available. If your competitors automate and you don’t, your costs look worse, your prices become less competitive, and you lose share. So firms automate because they have to, not because they’re ideologically committed to it. Over time, that pushes the whole system toward widespread automation, a smaller slice of income going to workers, and tougher price competition. Each company is trying to protect itself, but the combined effect is weaker demand, less pricing power for most businesses, and a more fragile economy.

In an ideal world, you’d manage that transition with coordination: maybe you pace adoption, strengthen income support, invest heavily in retraining, or redesign redistribution in a way that keeps demand healthy. The catch is that individual companies can’t coordinate that on their own. Governments can try, but they run into a structural bind: as labour income shrinks, payroll and income tax receipts soften, while capital and business models move more easily across borders. So the state is being asked to stabilise demand just as its most reliable revenue base narrows. That’s why this scenario tends to swing between policy approaches rather than settling into something durable.

If you zoom out and look at the system dynamics, the engine becomes clearer. There’s a powerful reinforcing loop: invest in AI, productivity rises, competition intensifies, and that pressure forces even more AI investment. It kicks in quickly and speeds up over time. Historically, capitalism also had a stabiliser: jobs and wages supported spending; spending supported demand; demand supported profits; profits funded investment. AI weakens that stabiliser because a smaller share of output flows back into wages. So the economy’s ability to produce can outrun its ability to distribute income. When that gap gets big, you get an “unstable growth machine”: output keeps rising, but value realisation gets harder, volatility increases, and markets become more sensitive to policy signals and liquidity.

Timing is the killer. The automation and competition loop plays out fast. The counter-response is slow. Policy, institutions, and social bargains usually move after the fact, once trust has already been damaged and the set of feasible options has narrowed. So you get a pattern of oscillation: stimulus followed by backlash, deregulation followed by crackdowns, optimism followed by crisis. Not because anyone prefers chaos, but because the feedback structure makes smooth adjustment difficult.

Once you hold those pieces together, the consequences stop looking like a random list and start looking like a connected chain.

On jobs, the first hit is pretty direct: routine office and clerical work erodes quickly, and wages compress across a lot of middle-skilled roles. Over time, the deeper damage shows up: more long-term underemployment, less willingness to retrain because it feels like a risky bet, and a growing sense of “growth without progress.” The labour market doesn’t just shift. It splits, with a small number of highly rewarded roles and a much larger group facing weaker bargaining power, more instability, and fewer paths upward.

For governments, the first effect is fiscal pressure. Wage-linked tax bases weaken, while demand for welfare, unemployment support, and social insurance rises. The second effect is political volatility. Some governments tighten spending to protect credibility. Others spend more to protect legitimacy. Both approaches become unstable when the revenue base is eroding and the underlying coordination problem hasn’t been solved. Over time, debt stress builds—sometimes it breaks into crisis, but more often it shows up as a slow loss of room to manoeuvre. The system gets more brittle because the state is expected to provide stability without the revenue structure it used to rely on.

For investors, the first effect can feel counterintuitive at first: outside of monopoly or bottleneck positions, average profit rates compress. Competition gets harsher, prices come under pressure, and demand is less reliable. Investors respond by crowding into perceived winners and leaning more heavily on financial engineering, speculative assets, and liquidity-driven narratives. The second effect is fragility. When productivity growth no longer translates cleanly into earnings growth, valuation becomes more sensitive to discount rates, policy, and confidence. You get more boom–bust behaviour as capital searches for safe places in an economy that struggles to convert its productive capacity into broad-based demand.

Companies, meanwhile, become more defensive. For most firms, productivity gains don’t turn into durable profit expansion because AI tools spread quickly and competition intensifies. Cost cutting becomes the default strategy, not because it’s inspiring, but because it’s how you survive in an arms race. The second-order result is more consolidation and stagnation. If demand is fragile and differentiation erodes, firms merge to regain pricing power, or they lean on cheap capital just to keep margins intact. Innovation still happens at the frontier, but it doesn’t diffuse as broadly, and the middle of the economy becomes sluggish.

When you map this onto asset classes, the same logic shows up.

Public equities become more split. A small number of dominant firms can sustain high margins because they control distribution, data, platforms, or regulatory moats. Most of the market faces valuation pressure because earnings growth doesn’t keep up with productivity. Indexes might look fine for a while, but the risk shifts toward long periods of weak real returns, because broad earnings power is capped by weak income distribution.

In private equity and growth equity, the usual playbooks get tougher. EBITDA growth becomes less dependable because the same automation forces that lift productivity also intensify competition and weaken demand. Exit multiples compress, leverage gets riskier, and underwriting shifts toward resilience and cash extraction rather than multiple expansion. Roll-ups can still work in fragmented sectors that are operationally messy, locally rooted, or hard to automate, but service models that depend mainly on scaling labour structurally weaken.

Venture capital doesn’t vanish, but it becomes more barbelled. The outliers can be enormous because controlling bottlenecks matters more, while the middle thins out as features commoditise and distribution dominates. The best opportunities cluster around infrastructure and enabling layers—compute, data, security, tooling, integration—rather than startups whose main edge is simply replacing labour for a short window.

Credit becomes more fragile too, because the assumptions lenders usually underwrite—stable pricing, predictable demand, reliable EBITDA—start to wobble. Cycles shorten, default risk rises, and lenders move toward collateral, tighter covenants, and stricter structures. Senior secured and asset-backed lending holds up better than loans that rely on optimistic growth and loose terms.

Real assets also split. Office and urban commercial property tied to dense employment and rising wages face structural repricing as work patterns change and labour income weakens. Regulated infrastructure, by contrast, looks more bond-like: expensive to build, hard to replicate, supported by policy, and less sensitive to demand volatility. And as property and wage-linked revenues soften, city finances and local politics get more strained in lasting ways.

The big takeaway is that none of this depends on technology failing. It just requires the incentives to keep pushing in the same direction while institutions fail to respond fast and strongly enough to preserve a stable conversion from capability into demand and legitimacy. The economy becomes incredibly capable. What lags is the set of mechanisms that make that capability broadly stabilising. You get a world with very high output—and persistently low stability—an economy that can do almost anything, except share the gains in a way that holds over time.