AI is being taught how to do your job. Will artificial intelligence replace us, or just change the ...

TEXT START: Music producer Robbie Hiser is confident there will always be a place for creatives like him even though he taught artificial intelligence some of the tricks of his trade.

THE DISSECTION

This article stages a false choice: AI will either replace workers or merely improve their jobs. Under the Discontinuity Thesis, replacement arrives through job transformation. Tasks are automated, headcount falls, and the remaining human role becomes narrower, more prestigious, or more precarious.

The article accidentally documents that process. Mercor’s 100,000 freelancers and temporary $20–$200-per-hour training work are not evidence of durable employment. They are calibration labor: humans are paid to convert expertise into systems that will require fewer humans later. The law-firm personas and AI agents make the mechanism explicit—senior judgment is packaged, distributed, and used to reduce the number of people needed to produce the same output.

THE CORE FALLACY

The article confuses residual human activity with economically necessary human labor.

A song may retain a human “touch.” A client may prefer to look a lawyer in the eye. A courtroom may still require a human advocate. None of this proves that the surrounding labor market remains intact. Human presence can survive as branding, liability cover, luxury service, or a thin interface over automated production.

AI does not need to eliminate every human act. It only needs durable cost and performance superiority across enough cognitive tasks to force firms to reduce labor under competition. The piece also treats AI training as an expanding career, ignoring the obvious contradiction: the more effectively workers train the system, the less labor the system requires.

HIDDEN ASSUMPTIONS

• Firms will choose augmentation over labor reduction even when competitors gain an advantage by eliminating staff.
• New AI-training jobs will scale proportionally with the jobs they help destroy.
• Human authenticity can be monetized broadly rather than reserved for premium niches.
• Retraining can absorb millions of displaced workers at the speed AI is advancing.
• Present AI limitations—lack of life experience, courtroom presence, or emotional credibility—are permanent moats rather than temporary constraints.
• Tax incentives and managerial goodwill can override the competitive pressure toward automation.
• Stable unemployment today means the transition is manageable. It does not; labor-market lag is not structural protection.

SOCIAL FUNCTION

Primary classification: transition management and ideological anesthetic, with a substantial partial-truth component.

The article admits that displacement is real, then wraps the admission in optimism about creativity, human connection, retraining, and policy choice. It makes systemic elimination sound like a skills-upgrade problem. It also avoids the central question: who owns the models, compute, platforms, and productivity gains?

The result is useful to institutions. Workers are told to adapt. Companies are framed as designers of humane AI. Governments are invited to adjust incentives. Ownership and bargaining power remain offstage while the machine is installed.

THE VERDICT

This is an early obituary disguised as balanced reporting. It records all three DT stages: AI absorbs cognitive tasks; institutions fail to preserve human-only domains at competitive scale; and entry-level pathways begin collapsing before mass unemployment becomes visible.

The “human touch” will survive, but mainly as a premium feature, a liability wrapper, or a status signal. It will not preserve universal wage-based participation. AI training, creative work, law, and product development do not refute the thesis; they show expertise being extracted, encoded, and redeployed to compress headcount.

The article’s central question is malformed. AI changes work precisely by replacing the workers who previously performed it. Human labor survives where people own the systems, remain indispensable to them, or occupy protected physical, legal, logistical, and maintenance niches. The mass employment circuit does not.