Why it matters
Recent evidence cited in the analysis finds no systematic unemployment increase among highly exposed workers since late 2022, while real-world use remains much narrower than benchmark capability. That gap matters because jobs consist of interconnected tasks: if AI cannot complete the whole chain reliably, automating one part can increase demand for the judgment, coordination, and accountability that people still provide.
The absence of an immediate employment shock does not guarantee an easy transition. Models continue to improve and adoption can accelerate, but high error costs, expensive computing infrastructure, rapid depreciation, and data-center electricity demand all constrain the business case. Policy should track task change, job quality, productivity, distributional effects, and resource costs together instead of treating a single headline employment number as the outcome.
Go to the source
Read the evidence behind this analysis. External links open in a new tab.
The Guardian — Why the AI jobs apocalypse may not be imminent
