Most organizations "do AI" by handing it to the engineers, and six months later nothing about the real work has changed. In this episode, Josh makes the case that AI adoption fails at the human layer, not the machine layer. He explains why most failures are baked in before an engineer ever touches a tool, and why the people best suited to lead adoption are often the ones every org wrote off as a support function. Along the way he draws a hard line: the market took two different jobs, AI implementation and AI adoption, and stapled them into one job description. Whether your rollout stalled or you are realizing the door is open to you, this one is for you.