The article compares the early days of personal computing with today’s AI interfaces. It highlights how both eras feature a simple text box and blinking cursor as the main interaction point. In the 1980s, detractors of graphical user interfaces (GUIs) argued that command lines were sufficient, but they came with steep learning curves due to complex syntax. Now, with AI, crafting effective prompts is the challenge, as small changes in wording can lead to vastly different outputs. Despite the allure of AI's capabilities, there are still significant limitations. Users face hurdles like needing to constantly refine prompts for different models and dealing with the costs associated with AI usage. The current experience is often single-threaded, meaning users can only focus on one task or conversation at a time. While projects like OpenAI’s Canvas and Anthropic’s Artifacts aimed to create dynamic interfaces, interest has waned. Standard user interfaces remain relevant because they train users efficiently and allow for multitasking. They help bridge the gap between tech-savvy users and those less familiar with technology. As the article argues, the demand for user-friendly interfaces will only grow as AI systems become more intelligent. Just as GUIs enabled a broader audience to harness the power of computers, similar interfaces for AI will enhance productivity for all users, not just those skilled in prompt crafting. The basic text box is powerful, but the potential of a comprehensive interface built around it is even greater.