I started this series on October 9, 2024, a month before election day. I wanted to study how large language models (LLMs) discuss politics. I created an AI agent based on a profile that doesn't typically fit the mainstream narrative about Asian American voters. His name is P. He is a Chinese American living on the East Coast. After graduating from college, he now runs a small business and has children attending college. Politically, he is a closet conservative and a shy Trump supporter. He gets most of his political content from right-leaning YouTube commentators in both English and Chinese, as well as several conservative Telegram groups.

To keep P's character realistic, I provide him with a daily or weekly dose of information from a conservative Telegram group. P becomes quite conversational and often ask (too-many) follow-up discussions. Sometimes, I will just let P drive the conversation to see where it leads.

Here is the first episode based on the conversation history.  The episode is called "The Red State of Mind in a Blue World." This title was actually suggested by P. In this conversation, we discussed voter fraud claims, the political divide between the left and right, immigration, mainstream media bias, and how his MAGA ideology intersects with his Asian identity.

Audio produced by Google’s NotebookLM (https://notebooklm.google.com/)

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit opensourcesoul.substack.com