For years, artificial intelligence has been framed as a contest over intelligence itself. Public discussion has focused on increasingly capable chatbots, image generators and large language models.

Developers are increasingly relying on large language models (LLMs) for everyday computing tasks such as fixing bugs, ...