Musk challenges legendary AI researcher Karpathy to an AI coding showdown against Grok 5 — gets a polite 'no' to an IBM Deep Blue-like showdown

Elon Musk proposed a coding contest between xAI's Grok 5 and Andrej Karpathy, likening it to the Kasparov vs. Deep Blue showdown. Karpathy declined, preferring collaboration over competition. The idea of a model-versus-human coding contest is not far-fetched, with DeepMind's Gemini 2.5 and OpenAI's GPT models achieving perfect scores. Musk would need to subject Grok 5 to the same conditions for it to be taken seriously in that class. Karpathy's decision not to participate reflects a shift towards focusing on how models can accelerate human output rather than head-to-head contests.