Human communication is a complex process involving verbal and nonverbal cues, language, and context. With the advent of large language models (LLMs) like ChatGPT, humans will need to adapt their communication style when interacting with these computerized systems. Effective human-LLM interaction requires providing context, structuring conversations as interactive chats, and using coarse-to-fine prompts to gradually create context. It also involves considering the purpose and goals of the conversation, whether practical or exploratory in nature. Moreover, humans need to be aware that LLMs may not have accumulated enough information or context to achieve a goal and should ask directive questions to lead to the expected outcome. Ultimately, human-LLM communication will require a deeper understanding of each other's capabilities and limitations.