# notes --- - [[{6.3a} ai prompts are just context]] - [[{6.3} chatbots are not the right interface for AI]] # summary --- Chatbots are not the right interface for large language models. It doesn’t tell users *how* it should be used; you have to continuously offer additional context; responses are not adequately connected to prompts; users are forced to switch between implementation (asking a question) and evaluation (reading the response) without any sense of flow; and it’s pushing users further toward shorter, lower-quality content rather than being a tool for thinking — which is where the real magic will be. # highlights --- >**Good tools make it clear how they should be used.**.. The only clue we receive [in a chat interface] is that we should type characters into the textbox… Of course, users can learn over time what prompts work well and which don't, but the burden to learn what works still lies with every single user. When it could instead be baked into the interface. > static text is dead >**Good tools let the user choose when to switch between implementation and evaluation.** When I work with a chatbot, I'm forced to frequently switch between the two modes. I ask a question (implement) and then I read a response (evaluate). There is no "flow" state if I'm stopping every few seconds to read a response. The wait for a response is also a negative factor here. As a developer, when I have a lengthy compile loop, I have to wait long enough to lose the thread of what I was doing. The same is true for chatbots. >There's an ongoing trend pushing towards continuous consumption of shorter, mind-melting content… Most of the products I've seen built with LLMs push us further down this road: why write words when an AI can write that article for you? Why think when AI can write your code? >we should be embracing our humanity instead of blindly improving efficiency > Let's build tools that offer suggestions to help us gain clarity in our thinking, let us sculpt prose like clay by manipulating geometry in the latent space, and chain models under the hood to let us move objects (instead of pixels) in a video.