Is the way AI talks to you actually helping you grow in your career?
In this episode, Gretchen and Bre'Ana get practical about training your AI assistant. Providers are working to make AI more human-centered, but we still have to use custom instructions so these tools talk to us in a way that's useful. They trade real examples: an assistant that wants to do too much of the thinking, flattery that won't quit, answers with no sources, and a career coach agent that positions itself as the only intelligence in the room. Rather than focusing on prompts, productivity hacks, or the latest AI product news, Coworkers with AI is about making sense of how AI is changing our work, and how we find our way through it, together.
RESOURCES
Believing Anthropomorphism: Examining the Role of Anthropomorphic Cues on Trust in Large Language Models
Invisible Saboteurs: Sycophantic LLMs Mislead Novices in Problem-Solving Tasks
Training better humans with AI, by Louis Tay, Karina Vold, and Ashton Anderson (Purdue University): https://www.purdue.edu/hhs/news/2025/...
How to tone down the sycophancy in Claude, ChatGPT, and other AI chatbots (Transparency Coalition): https://www.transparencycoalition.ai/...
Your AI Is Designed to Flatter You, by George Kao: https://georgekao.substack.com/p/your...