AI jobs debate intensifies
Economists and business leaders are publicly split over whether AI will cost mass jobs or simply transform work — Raghuram Rajan argued job‑loss fears are overstated while others warn many roles will disappear as automation accelerates argued reported. At the same time, reporting flags a hidden cost: highly skilled professionals are spending time training AI systems, shifting where value and bargaining power land reported.
Raghuram G. Rajan argued in a Project Syndicate column on March 10, 2026 that runaway predictions of an immediate, economy‑wide white‑collar job apocalypse are overstated and that the outcome will depend on adoption speed, market structure and public policy. (project-syndicate.org) At a recent industry event, RDR Business Solutions founder Reymond delos Reyes warned that many workers who “cannot adapt” will be displaced by AI, comments reported around March 15, 2026. (mb.com.ph) A Philippine union and labour groups have flagged scale: the Federation of Free Workers estimated roughly 5 million Filipino jobs were at risk from AI and climate shocks in 2025, illustrating how national labour markets are already being stress‑tested. (msn.com) Evidence of a new hidden cost appears inside firms: Harvard Business Review found AI tools often intensify work because senior staff spend substantial time auditing and verifying AI outputs rather than doing other tasks. (hbr.org) Academic and investigative reporting also documents human labour embedded in model building — Oxford’s Mark Graham calls AI an “extraction machine” that depends on human annotation, and Pulitzer Center investigations have traced low‑paid data‑labour chains that support generative models. (ox.ac.uk) The backend economics are steep: an arXiv cost‑model paper estimated the amortized cost to train the most compute‑intensive models has grown about 2.4× per year since 2016, while industry charts put some state‑of‑the‑art training runs at $100m‑plus. (arxiv.org) Corporate behaviour reflects both hope and hurry — a December 2025 executive survey highlighted by HBR shows companies are already using AI as a rationale to cut roles or slow hiring, and research firms warn that even optimistic scenarios put a few percent of jobs at direct risk if AI is broadly applied. (hbr.org) Rajan’s policy pitch — avoid platform oligopoly and deploy targeted measures like price regulation, taxation and retraining — is the explicit alternative he offers to dystopian forecasts, a set of interventions he said in Project Syndicate could shape whether gains are concentrated or broadly shared. (project-syndicate.org)
