Large Language Models (LLMs) have demonstrated impressive text generation capabilities, prompting us to reconsider the future of human-AI co-creation and how humans interact with LLMs. In this paper, we present a spectrum of content generation tasks and their corresponding human-AI interaction patterns. These tasks include: 1) fixed-scope content curation tasks with minimal human-AI interactions, 2) independent creative tasks with precise human-AI interactions, and 3) complex and interdependent creative tasks with iterative human-AI interactions. We encourage the generative AI and HCI research communities to focus on the more complex and interdependent tasks, which require greater levels of human involvement.