Experience building communities of practice in which scientists from multiple domains converge on shared exploration and use of new technologies- whether through organizing journal clubs, meetups, student groups, or online scientific communities. This research continues many of the directions our team worked on prior to Anthropic, including: GPT-3, Circuit-Based Interpretability, Multimodal Neurons, Scaling Laws, AI & Compute, Concrete Problems in AI Safety, and Learning from Human Preferences.