Why Australia's quiet labour market isn't proof AI has flopped - Crikey

TEXT START: In 1987, almost two decades after the invention of the internet, Nobel Prize-winning economist Robert Solow quipped that: “You can see the computer age everywhere but in the productivity statistics.”

THE DISSECTION

The excerpt is a framing device, not yet an economic demonstration. It invokes the Solow paradox to establish a lag template: technological capability can become visible long before productivity or employment statistics register it. The headline then uses that template to argue that Australia’s quiet labour market does not disprove AI.

The examples of superhuman hacking, novel mathematics and protein design establish capability spectacle, not economic deployment. Nothing supplied here connects those achievements to lower costs, redesigned workflows, reduced hiring or displaced labour. The paywall interrupts before the article provides that bridge.

THE CORE FALLACY

The likely error is treating the computer-era lag as a template for AI rather than as evidence that implementation and measurement can lag invention. Desktop computers primarily augmented human workers. AI is aimed at automating cognitive work itself. The historical analogy may explain delay; it does not establish that AI will have the same trajectory or social effect.

The narrow claim that a quiet labour market is not proof AI has failed is logically correct. But it is also weak. Under the Discontinuity Thesis, a quiet labour market can be a lag effect produced by attrition, hiring freezes, institutional inertia and gradual task removal. Silence is not exoneration. It may be the period in which the system is still concealing the wound.

HIDDEN ASSUMPTIONS

• Aggregate productivity statistics will detect AI’s effects promptly and accurately.
• Technical benchmark gains will translate into deployed systems and reorganised firms.
• Labour displacement will appear as obvious mass unemployment rather than wage compression, non-replacement, downgraded work or declining entry-level access.
• Firms will adopt AI at a pace that can be observed through conventional macroeconomic indicators.
• The computer age is a sufficiently close precedent for an automation technology that can substitute for cognition.
• Existing institutions can absorb the transition without a collapse in productive participation.

SOCIAL FUNCTION

This is partial truth functioning as transition management and ideological anesthetic. It correctly blocks the lazy conclusion that no visible labour-market shock means no AI impact. But it risks replacing one lazy inference with another: because disruption may be delayed, the eventual outcome is treated as familiar, manageable or benign.

The historical anecdote gives the audience a comfortable script—wait for the statistics—while shifting attention away from ownership, control and whether most people remain economically necessary. It is a lag-defense argument, not a rebuttal of structural displacement.

THE VERDICT

An intelligent opening with a narrow valid point and no decisive evidence. The excerpt establishes neither that AI has failed nor that the Discontinuity Thesis is true. It supplies no proof of durable cost and performance superiority, coordination failure or majority loss of economically necessary work.

Its real value is negative: Australia’s quiet labour market cannot acquit AI. At most, the passage shows that measurement and deployment may lag capability. The corpse has not been certified alive merely because the labour-market monitor is still flat.