AI job displacement debate: scenario planning maps four futures for workers

Economists disagree on whether AI drives mass unemployment. The article cites layoffs and hiring slowdowns (e.g., Block’s 40% workforce cut, Salesforce reducing thousands of support roles) alongside studies finding limited earnings impact or mixed outcomes (Vanguard job growth/wages for AI-exposed occupations; NBER Denmark showing zero measurable effect on hours or earnings). PwC’s 2025 AI Jobs Barometer also claims AI-skill workers earn a 56% wage premium and productivity rises sharply in AI-exposed industries. To handle conflicting evidence, Tim O’Reilly argues for scenario planning rather than single-point forecasting. He proposes two uncertainties for the AI job debate: (1) AI capability vs. adoption speed, and (2) whether AI is used mainly for efficiency in existing tasks or for “doing more” (new products and expanded markets). He notes capability is accelerating (SWE-Bench coding performance surging in 2023–2024, plus further model improvements), while adoption may be the near-term constraint (Goldman: AI suppresses hiring more than it destroys jobs). Crossing these vectors yields four futures: augmentation (gradual adoption, broad worker gains), slow squeeze (efficiency use, fewer entry-level jobs), displacement crisis (rapid capability + rapid efficiency, heavy job cuts and “AI redundancy washing”), and great transformation (rapid adoption with new work creation, especially in relational sectors like healthcare, education, and human-centered services). “News from the future” is mixed: 2025 layoffs attributed partly to AI, rising employee anxiety, and planned workforce reductions contrast with wage-premium and job-growth signals for AI-exposed occupations. For traders, the key takeaway is that AI labor-market effects are likely uneven and data-dependent, implying continued volatility around tech adoption narratives rather than a single macro outcome.