How AI could change your career: Ex-Meta, Salesforce lead categorises workers into three types

TEXT START: Artificial intelligence disruption continues to effect industries across the globe.

The Dissection

This article translates structural labor displacement into a digestible caste system and a mental-health story. “Build AI, use AI, or be managed by AI” is not three equal career paths. It is a hierarchy:

• Building AI belongs to a narrow ownership and technical-control class.
• Using AI is a temporary servitor position unless tied to ownership, proprietary data, infrastructure, distribution, or scarce accountability.
• Being managed by AI describes the likely mass condition: economically subordinate labor under algorithmic control.

The article correctly identifies that the shock is broad, rapid, and psychologically corrosive. It then routes the conclusion toward training, mentorship, community, and “lasting careers.” That is transition management: treating a change in the economic operating system as a workforce-development problem.

The Core Fallacy

The text assumes that the three categories will form a stable labor market and that displaced people can move upward through adaptation. Under the Discontinuity Thesis, that fails.

P1 makes AI superior across cognitive work. P2 prevents institutions from preserving large human-only economic domains. P3 removes the majority from economically necessary labor. “Using AI” therefore becomes a baseline capability, not a durable moat. Workers using the same tools compete against one another while firms capture the productivity gain and reduce headcount.

The article’s own admission that AI was more capable than many entry-level hires—and that hiring changed because of it—is not merely a transition challenge. It is evidence that the wage-employment circuit is being severed.

The decisive distinction is not build versus use versus manage. It is control of productive AI capital versus dependence on whoever controls it. The article avoids that distinction because it would turn a career story into an ownership indictment.

Hidden Assumptions

• The AI economy will generate enough stable jobs to absorb displaced workers.
• Training can convert most people into economically valuable builders or users.
• Higher productivity will produce broader employment rather than fewer workers producing more output.
• Governments, businesses, and nonprofits can coordinate against the competitive incentives driving automation.
• Being managed by AI remains meaningful employment rather than a lower-status labor tier, rationing system, or transfer-dependent existence.
• Psychological damage can be treated with mentorship and community while the underlying loss of productive necessity remains unresolved.
• AI’s advantage over new graduates is temporary and can be closed through reskilling.
• Employment can continue supplying purpose, identity, and belonging after employment ceases to be broadly necessary.

None of these assumptions is established by the text. Several are directly strained by its own evidence.

Social Function

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

The article accurately records the symptoms: unequal gains, geographically broad disruption, accelerated hiring contraction, and the psychological collapse that follows loss of usefulness. But it rebrands structural dispossession as an employability problem. The proposed foundation may help individuals navigate the wreckage; it cannot create the mass demand for human labor that automation removes.

The “lasting careers” language also performs elite self-exoneration. It implies that the central problem is insufficient preparation rather than ownership of the machines replacing people. It offers a humane landing zone without confronting who owns the aircraft.

The Verdict

This is a warning label attached to a demolition charge. The article sees that AI is reaching knowledge workers faster and more broadly than prior industrial shocks, but misclassifies regime change as career transition.

Build, use, and manage are not three comparable options. They are the apex, intermediary layer, and mass base of an emerging dependency pyramid. Training may help selected people become useful to Sovereigns; it cannot restore majority productive participation. The article’s deepest truth is the fact it fails to follow to its conclusion: once firms prefer AI to new graduates, the post-WWII wage-to-consumption circuit has begun to die.