Short answer: both—but on different timelines. In the near term, AI will eliminate or reshape many routine, predictable tasks; over the medium to long term, it tends to create new tasks, roles, and even industries as productivity gains lower costs, expand demand, and make new products feasible. Whether net jobs grow or shrink where you live depends on adoption speed, regulation, and how quickly workers and firms adapt.
How it plays out
•Task, not job, automation: Most roles contain automatable and non‑automatable tasks. AI takes the repeatable parts and leaves (or expands) work needing judgment, context, creativity, and interpersonal skills.
•Price and demand effects: When AI makes something cheaper/faster, firms often sell more of it, hiring for complementary work (sales, integration, QA, support, compliance).
•New work appears: Tooling, integration (MLOps, data engineering), safety/governance, AI‑enhanced product design, domain experts “teaching” models, human‑in‑the‑loop quality control.
•Uneven impact: Routine cognitive and back‑office work face the largest disruption; customer service, basic content, simple analysis, and some coding/testing see big shifts. Jobs involving field work, complex coordination, high‑stakes decisions, or relationships skew more complementary.
What this means for you
•Decompose your role: List your tasks. Anything digital, rules‑based, and repetitive is at risk; plan to automate it yourself and move up the value chain.
•Build durable complements:
•Tool mastery: automate workflows (AI + spreadsheets/scripts/low‑code).
•Domain depth: industry knowledge to set requirements and judge outputs.
•Data literacy: prompt writing, evaluation, and basic analytics.
•Human skills: communication, negotiation, change management, ethics.
•Produce proof of value: Ship small AI‑assisted projects that save time or money; quantify the impact.
•Position your work: Upstream (problem framing, requirements) and downstream (verification, accountability, compliance) tasks are stickier than midstream routine production.
For employers and policymakers
•Redesign jobs around AI, don’t just “add tools.” Measure task mix, retrain, and set guardrails.
•Expect churn: fund rapid reskilling, portable benefits, and transition support to reduce short‑term pain and unlock long‑term gains.
Bottom line: AI will take away some jobs and create others; net outcomes hinge less on the tech and more on how quickly people, firms, and institutions reorganize work. If you adapt your task mix toward judgment, integration, and oversight—while using AI to handle the drudge work—you improve your odds of being on the “more opportunity” side.