Anthropic's Lydia Hallie rethinks Claude Code best practices for newer models
AI Engineer · youtube · 2026-11-12
In an AI Engineer Code Summit talk, Lydia Hallie, Member of Technical Staff at Anthropic working on Claude Code devrel, argues that most of what developers added to their Claude Code setups over the past year is now holding them back as models have gotten really good. She walks through what best practices look like for the newest Claude models — what's still worth keeping and what you no longer need to think about.
More from coding & agent
- Google AI Studio adds Maps Platform: build location-aware apps from a single prompt — RichardsonDx · 2026-11-12
- Seroter's reading list: Gemini Decisions API, agent memory components, enterprise AI adoption — rseroter · 2026-11-12
- Baidu's KOOKO turns a restaurant brief into a working booking system in one chat — socialwithaayan · 2026-11-12
- Why no wave of BYOAI SaaS? Building products on users' own GPT/Claude subscriptions — paul_cal · 2026-11-12
- SparseRead cuts agent reading tokens by up to 92.9% without quality loss — ict-cas · 2026-11-12
- 10 open-source repos that force, validate and repair JSON output from LLMs — victor_explore · 2026-11-12