A working LLM built inside Age of Empires 2 — out of goats — to show AI isn't sentient

A vivid argument against mistaking fluency for sentience. As recounted in the thread, a Microsoft researcher built a working LLM inside Age of Empires 2 — wiring the game's goats together as the yes/no switches that make up any computer — purely to show that a model replying in a warm, human tone is still nothing but logic gates flipping. Zy passes it along with a flat endorsement: "I would like it if everyone read this paper."

@ZyMazza on X →