The AI leaders can't be trusted for AI predictions - investingLive

TEXT START: One of the most-radicalizing events of the past year in AI has been the co-ordinated change in messaging from AI leaders.

The Dissection

This is not a neutral AI forecast. It is an incentive autopsy followed by a lag-trade investment thesis. The article correctly identifies leaders softening their rhetoric because public opposition threatens data centers, capital, political permission, and institutional power. It then treats economic inertia as an investable moat.

The core maneuver is simple: replace “AI will not replace labor” with “AI will replace labor slowly enough for incumbents to keep monetizing the delay.”

The Core Fallacy

It mistakes delayed displacement for failed displacement. Travel agents and realtors still existing proves only that occupations can survive after their monopoly is destroyed. It says nothing about headcount, wages, fees, bargaining power, or whether their labor remains economically necessary at scale.

Altman’s inertia argument is an admission of timing lag, not a rebuttal of the Discontinuity Thesis. DT is governed by P1, P2, and P3: cognitive automation becomes superior, institutions cannot preserve stable human-only domains, and mass productive participation collapses. Social adaptation can slow the process. It cannot reverse the mechanism.

The article also confuses unreliable spokespeople with unknowable economics. Leaders may lie about timing and consequences; cost curves, capability, deployment, and ownership still determine the outcome.

Hidden Assumptions

• Occupation survival equals labor-demand survival.
• Revenue persistence equals productive human participation.
• Institutional inertia is a permanent moat rather than a consumable buffer.
• Data-center resistance can permanently constrain deployment instead of delaying or redirecting it.
• Airlines are insulated because they are physical, ignoring their exposed cognitive, administrative, and coordination layers.
• Leaders’ statements are necessary to determine the direction of travel.
• A large total addressable market requires immediate disruption rather than eventual capture after a long adoption lag.

Social Function

Partial truth with transition management and ideological anesthetic. The article usefully exposes promoter self-interest, then offers investors a comforting tactical story: buy businesses protected by inertia and let the old order keep paying.

That is not a survival thesis. It is carcass management while the animal still twitches.

The Verdict

The article identifies the salesmen’s incentives but misses the structural verdict. AI leaders are unreliable narrators because their institutions depend on expansion, political consent, and continued financing. Their reversals make the transition more legible: minimize displacement publicly, capture the upside privately, and use inertia to buy deployment time.

The evidence supports slower, disguised, politically managed automation—not human economic permanence. This is a useful lag indicator and a tactical investment warning, but it is no refutation of the Discontinuity Thesis. The post-WWII labor-to-wage-to-consumption circuit can die slowly. Slow death is still death.