Generative AI is reshaping design practices, pushing designers to rethink how they interact with users and technology. The traditional model of human-computer interaction has evolved into a three-way dynamic involving humans, interfaces, and AI models. Designers now need to focus on Human-Model-Interface experiences (HMIx), emphasizing how generative models influence design outcomes. This shift requires a deep understanding of the capabilities of AI systems, user intent, and the expected quality of outputs. Successful design now hinges on collaboration and iterative prototyping, where ideal interfaces must be developed alongside AI models. To adapt, design teams must embrace new collaboration methods that integrate the realities of generative technology. They need to consider technical capabilities, input requirements, and user expectations to create effective interfaces. This process can feel chaotic, as designers must balance functional prototypes with product requirements and model expectations. For example, in the development of GenStudio, the team discovered that traditional brand guidelines were often too subjective. They created two key features: one that helps users make brand applications more objective through generative feedback, and another that allows users to request annotations for brand guidelines. Designers must also focus on user journeys and workflows to ensure the generated outputs meet their needs. Questions about user intent, reliability of AI responses, and trust-building communications are now critical. The framework introduced in the article, prompt-plan-show-next, outlines a structured approach to enhance user interactions with generative models. This framework guides users through the interaction process, from articulating intent to understanding how outputs are generated. It emphasizes transparency and actionable next steps, fostering trust and collaboration throughout the user experience.