Matthew Berman examines competing claims about AI and employment, including company announcements of job cuts and broader labor-market data. His interpretation that AI has been used as a pretext for layoffs is an opinion, not a verified general conclusion.
He discusses the cost of frontier models and argues that enterprises should match model capability to the work, while assessing whether spending produces useful outcomesCost per completed task measures the total AI, tool, infrastructure, retry, and repair expense for each verified useful outcome..
Berman describes organizational change, deployment and human review as continuing constraints on adoption. His expectations about future jobs and automation are forecasts rather than guarantees.
Watch on YouTube
