Explore Future Healthcare Solution with AI Agent
With the current LLM, we can imagine the quality increase of the outcome in a much quicker speed. However, without proper design process, the work is lack of evidence and hard to understand.
In this semester, working with D-Droid (an AI agent based on LLM), I explored and tested how system thinking, strategic choice making, and research through design are all empowered by technology. At the same time, it inspired my thinking of how might we reimagine our design process with new technologies.
Research through Design
How might we solve the wicked problem of healthcare system in Chicago? The goal is to empower people who lived in Chicago with more affordable, efficient and inclusive healthcare solutions with Low-altitude technology.
Discussed and designed with D-Droid, we analyzed different features, affordance, impacts and goals using system design frameworks like Anatomy of Infrastructure.
Data Visualization as system narrative
Further exploration: With another class I took this semester, Data Visualization, LLM inspired by the homework I did in that class, and imagine a new presentation with an interactive experience of Anatomy of Infrastructure. Usually, methodologies we used for system thinking is complex to visual every relationship in one diagram. In this case, interactive diagrams is a way to solve the problem. Working with LLM helped me generate more solution and data that can support the complex system diagrams.
With chatbot and other tech support, making the design intelligence and thinking process interactive can empower the future design methodology development.
How medicine reaches a patient, snapshot by snapshot
The line runs from the oldest delivery model at the top to the speculative ones at the bottom. Hover a point to enlarge its image. Click it to open the scene with its visible and invisible agents.
How medicine reaches a patient, snapshot by snapshot
The line runs from the oldest delivery model at the top to the speculative ones at the bottom. Hover a point to enlarge its image. Click it to open the scene with its visible and invisible agents.