The Discontinuity Thesis v3.3: Interface Collapse

The standard of proof every policy on this site is tested against. Read the full essay →

The Core Argument

Postwar capitalism dies when wage labour no longer provides mass economic agency. Not total labour extinction — mass agency collapse. The majority of working-age adults cannot sell labour at socially sustaining wages without subsidy, protection, artificial scarcity, makework, or political intervention.

The Four-Layer Cascade

Layer 1: Task-Level Unit-Cost Dominance

AI + thin human oversight produces professional task outputs at equal or better quality, faster, cheaper. Already crossed for many cognitive tasks. The question isn't whether AI can do the task — it's whether the human price point survives.

Layer 2: Interface & Workflow Dominance

AI operates through the same software environments where work happens — sees screens, clicks, types, browses, calls tools, manipulates files. The old moat was software fragmentation: humans as the integration layer between systems. That moat is collapsing.

Layer 3: Job-Level Dominance

Whole roles become unnecessary as enough task volume is stripped out. A job is a bundle of tasks. Once decomposable, the worker is decomposable with it.

Layer 4: Labour-Market Dominance

Wage labour stops being the mass route to economic agency. This manifests as non-absorption: fewer entry-level roles, broken training ladders, productivity gains not reaching wages, incumbents held while new entrants fail to launch.

The Three Complementarity Stages

Why "humans will work with AI" is not an answer:

Genuine Complementarity

AI raises the marginal value of human labour. Real but not guaranteed to persist. This is the stage policy makers point to. It's also the most temporary.

Transitional Complementarity

Humans supervise while AI improves. Unstable by design — the human role thins with every model generation. Today's supervisor is tomorrow's rubber stamp.

Theatrical Complementarity

Humans remain for trust, liability, or optics. Not augmentation — managed displacement. The pilot who hasn't touched the controls in two hours. The doctor who rubber-stamps the AI diagnosis.

Why Regulation Fails

Regulation defines categories; economics decides which survive. AI erodes task boundaries continuously: spell-check → drafting → composition → analysis → decision-making → execution. Where does spell-check end and full automation begin? No law can draw that line — the boundary between AI assistance and AI replacement doesn't exist.

Coordination fails on two axes: actors who keep expensive human labour lose to those who automate, AND task boundaries blur faster than regulators can redefine them.

Why Redistribution Doesn't Save the System

Capital redistribution may preserve consumption while conceding production. Recipients eat, vote, live — but no longer hold structural leverage from being economically necessary. Citizenship without productive agency. Dividends are a successor system, not a rescue.

The Four Scoring Lenses

Every policy on this site is scored against these four dimensions:

💰

Unit-Cost Survivability (30%)

Does this policy survive when AI drives marginal cost of cognitive work to zero?

🔌

Interface Collapse (25%)

Does it account for AI becoming the integration layer?

📉

Propagation Blindness (25%)

Does it see the full cascade from task to workflow to job to labour market?

🎯

Coordination Feasibility (20%)

Can it be enforced when defection = competitive advantage?

The Threshold

Not the last worker. The majority. Postwar capitalism dies when the majority of working-age adults cannot sell labour at socially sustaining wages without subsidy, protection, artificial scarcity, makework, or political intervention.

The Cope Report
Weekly. Free. No cope.
The week's most revealing AI coverage,
scored for omission. Every Monday.
Got feedback?

Send Feedback