Human–AI collaboration
Designing AI to finish or amplify a human's work rather than replace the human step — keeping judgment with the person and handing the drudgery to the model.
Human–AI collaboration is a design stance: instead of automating a task end-to-end, the system augments the person doing it. The human supplies judgment, intent, and the bits only they know; the model handles structure, recall, and tedium. The best AI products often pick this split deliberately — they finish the human's step instead of removing it, which keeps the output trustworthy and the user in control.
The automation framing — "the AI does it for you" — tends to produce output nobody trusts or reads. The augmentation framing keeps a human in the loop at the point where judgment matters, and uses the model where it's genuinely better: structure, phrasing, recall, breadth.
It's an old idea with new urgency. As models get more capable, the design choice of where to put the human becomes the difference between a tool people rely on and one they quietly stop using.