AI may not take your job but it will take your salary | FMT

TEXT ANALYSIS PROTOCOL

1. THE DISSECTION

This article performs a specific sleight of hand. It correctly identifies the mechanism of extraction—AI captures productivity gains, wages stagnate, workers get the shaft—but then proposes a policy nostrum ("free the labour market") that treats this as a distributional bug fixable by better rules. The author notices the knife wound and prescribes a different grip.

2. THE CORE FALLACY

The article operates entirely within the assumption that the post-WWII employment-for-wages compact remains structurally viable. It treats the wage suppression it describes as a policy failure correctable by labour market reform. The Discontinuity Thesis rejects this premise entirely. The problem is not that the pie is being split unfairly. The problem is that the mechanism for distributing the pie—mass employment generating wages generating consumption—is itself being automated away. No amount of "freeing the labour market" creates bargaining power when the employer no longer needs you in sufficient numbers. The J-curve the article describes is not a temporary lag before wages catch up. Under P1 and P2 conditions, it is the permanent new equilibrium: maximum productivity, minimum labour share.

3. HIDDEN ASSUMPTIONS

• Assumption 1: Employment remains the primary value capture mechanism. The entire argument assumes workers can still extract from the employment relationship. This assumes away the productive participation collapse that is the DT's core prediction.
• Assumption 2: Labour market reform is politically and institutionally feasible. The article acknowledges Malaysia's depressed compensation-to-GDP ratio (~32-33%) versus advanced economy norms (40-50%). It proposes "reforming and freeing the labour market" as the remedy. This ignores that Malaysia's depressed CE ratio reflects structural power asymmetries—employer monopsony, weak unions, educational surplus labour—that "freed markets" would deepen, not correct.
• Assumption 3: The digital dividend is a fixed pie. The article treats AI productivity gains as a known quantity to be distributed. It does not address that the dividend itself requires consumer purchasing power to materialize—creating the very demand collapse the DT predicts as the mechanism of system death.

4. SOCIAL FUNCTION

This is a transitional lullaby with a sharper headline than its content deserves. It performs the socially necessary function of telling workers "your job is safe" (reduce panic, preserve consumption patterns, prevent political disruption) while simultaneously telling power "don't worry, the structural fix is just better market design." It is management consultant-grade analysis dressed as heterodox warning. It identifies the symptom—wage theft via J-curve capture—and misdiagnoses the disease as policy rather than structural inevitability.

5. THE VERDICT

The article gets the mechanism right (wage suppression, not mass unemployment, is the near-term threat) but retreats to comfort-food economics when it matters most. "Free the labour market" is the policy equivalent of "thoughts and prayers." In Malaysia's case, the low CE ratio is not an accident waiting for the right market reforms—it reflects the exact labour surplus, weak institutional, educational inflation dynamics that make developing economies more vulnerable to the DT mechanism, not less. The author correctly fears fully staffed enterprises generating exponential output for stagnant wages. What he fails to confront is what happens when those enterprises need fewer of those "fully staffed" humans, which the J-curve analysis does not actually prevent—it only delays. The article is an autopsy report that still believes in CPR.