AI == :implemented:`Implemented` — in the honest sense: AI *demonstration examples* written in ordinary Karkain arithmetic exist under ``examples/09-ai/`` and run byte-identically on both engines. There is **no ``std.ai`` module and no AI/ML framework surface** in the language. These examples show classic ML *geometries* (distance, weighted votes, linear decision boundaries) implemented by hand with plain arrays, maps and arithmetic — deliberately, so the language itself can be exercised without pretending a framework exists. See :doc:`/status/implemented` for how the status vocabulary is used. .. list-table:: examples/09-ai/ :widths: 30 70 :header-rows: 1 * - File - Demonstrates * - ``01_nearest_neighbor.kark`` - Manhattan-distance nearest-neighbour classification by hand * - ``02_linear_classifier.kark`` - Perceptron-style weight updates and a decision boundary K-nearest neighbour (excerpt) ----------------------------- .. code-block:: kark func distance(a, b) { return abs(a[0] - b[0]) + abs(a[1] - b[1]) } ... Run: .. code-block:: console $ karkain run examples/09-ai/01_nearest_neighbor.kark .. seealso:: :doc:`/examples/machine-learning` — the adjacent hand-rolled ML category. :doc:`/status/planned` — the not-yet-implemented AI framework roadmap.