Emergent Edge

Emerging patterns in AI use

Emergent Edge explores how people use AI at its edges: new practices, unexpected interactions, and emerging risks. It turns posts from blogs and forums into evidence-backed case cards, then compares them with an evolving library of patterns.

  1. Discover. Semantic searches find candidate posts. Early filters select material for closer review.
  2. Describe. A model extracts the behavior, supporting quotes, and reported consequences into a case card.
  3. Compare. Embeddings retrieve related cases and patterns. A separate judgment considers whether the behavior fits an existing pattern or needs a new one.
  4. Review. The map makes the collection explorable; the pipeline walkthrough exposes saved inputs, prompts, and results.

The collection brings together saved research runs, with repeated cases shown once. Some pattern assignments were curated. Summaries record reported experiences and research claims; they are not independent verification. Recorded model output remains available alongside editorial review notes.