adityac

UI/UX for thought, research and feedback. Not necessarily action first.

For years, software interfaces were built to help users do things. Fill a form, run a report, complete a task. User input is at the forefront and heavily workflow driven. Though some of those concepts still are applicable, I think we will start seeing a new shift in the UI that is purpose built for AI Apps.

With the advent of LLMs, reasoning and reinforcement learning models. There will be a shift in the user interaction with the new AI first systems. User role is not going to be anymore action focused, user will be collaborating with machine in trying to make a decision, uncover something or as simple as provide feedback.

This shift in the role will demand a new UX design that will make us move past the traditional UI which is rigid for the AI world. It's not anymore about buttons, heat spaces and rigid visuals anymore. New applications will demand a UI that need to be more fluid and flexible.

Chat will be there, but we will move beyond chat where the interface is like a whiteboard and provides all the flexible components to the user to either build something from scratch. It could also be something where the focus is on exceptions more considering AI is doing the heavy lifting.

Context, Reasoning, Adaptability and Collaboration (with machine and users) will be some the key concepts that will drive the new UI designs.