What should we call the use of LLMs as part of products?

Usually when we have a tool, we have a name for its use. For example, using ML in a product is usually called predictive modeling, classification, or clustering.

I was thinking “Generative interaction” but worry the phrasing anchors and possibly pigeonholes. So I’m curious what folks in HN, many of whom have been regularly involved in the discussion and building of with LLMs, would call it?

Source link