The label 'AI design' now covers four distinct jobs that require different skills, tools, and mental models. Nielsen Norman Group identifies them as: using AI to accelerate existing design work, designing interfaces for AI-powered products, structuring data and content so AI agents can parse it correctly, and defining the behavioral logic of large language models.
The split matters because hiring managers, job listings, and designers themselves are collapsing four separate disciplines into one phrase. A person optimizing prompts for an LLM's personality is not doing the same work as someone laying out a chat UI, and treating them as interchangeable produces bad teams and bad products.
The full article is worth reading for how it draws the boundary between each role, particularly the distinction between content structuring for AI parsing and LLM behavior design, two jobs that barely existed three years ago and have no settled career path yet.
[READ ORIGINAL →]