MoralMachine

For 1952A course (Human-AI Interactions, https://slbooth.com/assets/projects/CSCI1952A/) at Brown University, we decided to take on the problem of cultural preference embedding into AVs.

We built on the Moral Machine experiment and the open-source MM data on HuggingFace (https://huggingface.co/datasets/Sonya3/MoralMachineHuman) to show the technological possibility of the idea, but then go on to criticise the approach the MM experiment took, as well as criticism of the computational ethics overall.

In this GitHub, you can find:

1. Code analysis of the HuggingFace dataset (data analysis, stat analysis, and machine learning);
2. Literature that inspired us;
3. The final presentation we delivered to the class.