The article highlights the growing importance of AI literacy as generative AI tools like ChatGPT and Gemini reshape how people search for and interact with information online. The author, a UX researcher focused on civic tech and digital inclusion, emphasizes that not all users engage with AI tools equally. Many lack the comfort or understanding that tech professionals assume. Two key skills define how effectively users interact with generative AI: prompt fluency and output literacy. Prompt fluency is the ability to craft detailed prompts that convey context and intent, while output literacy involves critically evaluating the AI-generated information for accuracy and relevance. In research with participants aged 23 to 65, the study identifies four user types based on their interaction with AI: AI novices, naive power users, skeptical abstainers, and AI experts. Novices struggle with both prompting and evaluation, while naive power users excel at generating prompts but often accept outputs without question. Skeptical abstainers understand the limitations of AI but engage minimally, and AI experts use the tools strategically, verifying information as needed. Participants demonstrated varying levels of prompt fluency, with more skilled users crafting longer, context-rich prompts, while less experienced users defaulted to simple keyword searches. The study underscores that prompt fluency tends to develop more rapidly compared to output literacy. Some users may produce polished responses without necessarily improving their ability to evaluate the accuracy of those responses. This disconnect can lead to misinformation if users fail to verify the AI's outputs. The article illustrates these concepts with examples, showing how nuanced interactions and follow-up questions enhance the effectiveness of generative AI in real-world scenarios.