Leading Through Change: Culture, Feedback & AI (Fireside Chat)
summary
Discussion centers on concrete practices: a recently published experiential reinforcement learning approach that has models analyze their own mistakes and overnight train on generalized patterns, using that to reframe personal error; treating simple clicks as failure once instant answers appear and then digging deeper into why signals break; internal log comparison showing Microsoft employees less likely to say “thank you” than Bing users; and clear actions — build psychological safety, give quicker in-the-moment feedback, and externalize a lot when prompting models.
AI-written summary · about
chapters (18)
- 0:00Leadership and change
- 1:07Chastity Melvin’s career
- 2:12Championship teams and culture
- 3:55AI-driven change at Microsoft
- 6:22Jessi Melvin’s resilience
- 7:45Overcoming failure
- 9:11AI insights from reinforcement learning
- 10:15Curiosity in leadership
- 11:41Analytics versus human skills
- 13:05Feedback in AI and teams
- 15:10Immediate feedback in teams
- 17:15Learning from AI feedback
- 19:20Adapting to AI
- 20:43Safety for risk-taking
- 22:05Teams feel valued
- 23:10Healthy conflict in teams
- 24:12AI advice next six months
- 26:15Closing thoughts
related talks
26mCulture at Speed Scaling Without Losing People in the AI era
23mLive Fireside Chat
47mEngineering Leadership in the Age of AI
26mFireside chat: Challenges with AI implementation in organizations
20mFailing Can Be Good! Bringing a Growth Mindset to Every Project
34mLead Like a Bot, Think Like a Human: Navigating Leadership in the Age of AI
06mIgnite: Failure and the Third Way
31mFireside Chat – Sam Ghods