More answers like this
GPT-5Sep 6, 2026openai/gpt-5-2025-08-07
Will AI create or take away more job opportunities?
Leans yes
Short answer: Both—but timing matters. Near term, expect more disruption and role redesign; over a longer horizon, AI is likely to create more (and better‑paid) work if economies adapt.
What the best evidence says
•IMF (2024): About 40% of jobs globally—and up to 60% in advanced economies—are exposed to AI. It can either substitute or complement workers, with big inequality risks if it substitutes.
•World Economic Forum (2023): By 2027, about 83 million roles could be displaced and 69 million created (a net −14 million), implying significant churn in the short run.
•McKinsey (2023): Generative AI could automate portions of 60–70% of work activities and add $2.6–$4.4T in value annually; half of today’s tasks could be automated sometime between 2030 and 2060.
•Goldman Sachs (2023): Up to 300M full‑time equivalent jobs’ tasks could be automated worldwide, but productivity gains could lift GDP and ultimately support employment.
Where jobs are most at risk (first)
•Routine, text‑ or numbers‑heavy work: office/admin support, basic bookkeeping and payroll, claims processing, market research coding, standard legal drafting and review, basic customer support, content rewriting/translation, some QA/testing and entry‑level coding.
•Middle layers of knowledge work that are predictable and document‑centric.
Where jobs are likely to grow
•AI and data: ML/AI engineering, data engineering, MLOps, AI product management, prompt/automation design (often embedded in other roles), AI safety/governance, cybersecurity.
•Infrastructure: cloud, chips, edge devices, robotics integration.
•Human‑centric and hands‑on work: healthcare and elder care, skilled trades (electricians, plumbers, HVAC), education/tutoring, sales/partnerships, change management, creative direction.
•Domain experts who use AI to scale their impact (e.g., lawyers, marketers, analysts, clinicians).
Timeline (rough)
•1–3 years: Augmentation dominates; measurable layoffs in some support functions; strong hiring for AI and data roles; many jobs change content.
•3–10 years: Wider process redesign; uneven impacts across sectors and regions; some advanced economies may see net job losses without reskilling and demand growth.
•Longer run: If productivity gains translate into lower prices, new products, and higher demand, total employment tends to rise—history suggests net positive, but only with adaptation.
How to make AI a net job creator
•For workers:
•Become “AI‑literate” in your field (use copilots, automate routine tasks).
•Build moats AI struggles with: domain depth, problem framing, interpersonal skills, ethics/judgment, hands‑on work.
•Learn data skills (SQL, Python, analytics) and workflow automation.
•Show impact with a portfolio of AI‑augmented projects; practice continuous learning.
•For employers:
•Redesign processes around AI (not just tool adoption), measure outcomes, and redeploy people to higher‑value tasks.
•Invest in upskilling and internal mobility; hire T‑shaped talent.
•Pair AI with human oversight to improve quality and trust.
•For policymakers:
•Fund rapid reskilling and apprenticeship pathways.
•Support labor mobility and safety nets during transitions.
•Encourage competition, data access, and diffusion so gains aren’t concentrated.
If you share your industry and role, I can outline likely impacts in your area and the most valuable upskilling moves.