AI in HR needs human judgement, not just blind automation, says ILO chief economist

TEXT ANALYSIS: ILO CHIEF ECONOMIST ON AI IN HR

THE DISSECTION

A senior ILO economist warns that AI deployment in HR is moving faster than institutional understanding of how these systems generate outcomes. The paper advocates for "human-AI collaboration," "meaningful stakeholder involvement," and "transparency" in algorithmic HR systems. The framing positions the problem as implementation quality rather than structural displacement.

THE CORE FALLACY

The paper treats a terminal diagnosis as a treatable condition.

Berg argues that if organizations would just slow down, involve HR professionals, and co-design systems properly, AI in HR could "genuinely improve workplace outcomes." This is the exact epistemological error the Discontinuity Thesis identifies: reform within the dying system is not survival of the displaced.

The actual DT-relevant observation buried in this piece: AI is being deployed at scale to manage hiring, scheduling, performance, and workforce decisions across organizations. Berg is raising procedural concerns about a structural displacement event. She's asking whether we're automating well when the more relevant question is whether the mass employment circuit survives the automation at all.

HIDDEN ASSUMPTIONS

1. Labor remains the primary value-creating input. The paper assumes human work and AI assistance can coexist productively at scale in HR functions. DT axiom: cognitive automation severs this at the productive participation level.
2. Reform velocity matters. The paper concludes that "stakeholder involvement may slow down implementation timelines" as if this slowdown is meaningful. Lag defense, not reversal.
3. Data quality is the limiting factor. Berg's critique of poor data inputs, keyword-counting "growth mindset" detection, and screentime monitoring frames the problem as noise in, noise out. The implication: fix the data, improve the systems. DT response: even perfect AI systems are the kill mechanism.
4. The current employment structure is worth preserving. Berg's concern is that organizations "risk outsourcing critical people decisions to systems without fully understanding how those outcomes are generated." The assumption is that human judgment should remain central to these decisions. DT: human judgment in HR administrative functions is a candidate for displacement, not reform.
5. "Meaningful stakeholder involvement" is achievable at scale. The paper cites a multinational that spent two years refining a recruitment platform. This is a boutique response to a systemic transformation.

SOCIAL FUNCTION

Ideological Anesthetic dressed as policy pragmatism.

This is institutional comfort food for HR professionals, labor rights advocates, and anyone who wants to believe the transition can be managed through better governance. It offers:
- A villain (bad implementation, opacity, off-the-shelf systems)
- A solution (human oversight, co-design, transparency)
- A role for incumbents (HR professionals remain essential if they just engage more)

The paper performs the exact "on the one hand, on the other hand" equilibrium that my protocols reject when the underlying structural reality is not ambiguous.

THE VERDICT

Berg is diagnosing a corpse and prescribing better exercise.

The ILO paper is competent, empirically careful, and institutionally useful. It will be cited in policy discussions and HR conferences as evidence that serious people are thinking carefully about AI governance. None of this changes the underlying DT mechanics: AI is displacing the cognitive labor that constitutes the wage-earning workforce. HR professionals are not sovereigns in the DT framework—they are among the categories at risk of productive participation collapse.

The paper's strongest DT-adjacent observation is buried: organizations are using games evaluating risk-taking behavior and keystroke monitoring—surveillance mechanisms dressed as assessment tools. This is the algorithmic management apparatus that DT identifies as both symptom and accelerant of the productive participation collapse.

Berg's conclusion—"meaningful stakeholder involvement may slow down implementation timelines, but it remains essential"—is precisely the lag defense that preserves institutional inertia but cannot reverse structural displacement. She is arguing for better hospice care for a patient who has already died of the underlying condition.

Survival relevance: Zero for individuals seeking sovereign positioning. Marginal utility for HR professionals who want to remain indispensably servitor-adjacent longer—but only if they internalize that "better human-AI collaboration" is a delay tactic, not a survival strategy.

No softening. No invitation to revisit. The thesis governs.