Andre Ye

I am an MIT EECS PhD student, PD Soros Fellow, and NSF GRFP Fellow working in human-AI interaction, advised by Mitchell Gordon. I want to build AI concept tools that help us define, develop, and collaborate on humanistic concepts (personal, social, ethical, political, scientific, …) in the form of neologisms, metaphors, connections, graphs, and more beyond the chat interface. I have philosophical interests in work that centers the structuring role of concepts in life and thinking, from Deleuze to Foucault to Brandom.

I am currently working on a framework enabling humans to better understand and intervene on the intermediate decisions that LLMs rely on in “solving” complex open-ended tasks (specifically, in philosophy, AI alignment, and poetry). Stay tuned! In the past, I have explored what philosophers think about improving LLMs, how the “woke Gemini” controversy could instruct us in more critical creative practices, and other ideas in computer science and philosophy. Find what I’m thinking about for now and the future here.

My email is andreye [at] mit [dot] edu.

This is a randomly chosen image from Better Images of AI, an excellent project which seeks to supplement and challenge dominant visualization of AI as a Terminator-robot-agent by emphasizing how people, data, labor, resources, and ideas go into making AI.