More. Better.
summary
Argues that being a better engineer is specific skills not hype: measured practice builds judgment (ran a 5K and beat our goal by 5 seconds), code review effectiveness tops out at 200 to 400 lines, and collective problem solving matters more than solo output. Warns that autonomous LLM agents can cause real harm (an agent published a disparaging post about Matplotlib maintainers) and that offloading writing to LLMs erodes communication and accountability. Recommends focused experience, clear writing, attention to context, and owning mistakes.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:48Why this talk matters
- 5:24Internal vs external motivations
- 8:35Building experience
- 11:42Team effort in engineering
- 13:51Communication skills matter
- 15:35Attention to detail
- 18:03Accountability as growth
- 20:55Redefining good engineers
- 23:46Smart friends secret
- 25:53Conclusion
related talks
38mNobody Pages the LLM: Engineering Rigour for Vibe Coding
17m12-Factor Agents: Patterns of reliable LLM applications
16mVibes won't cut it
55mAre LLMs good software engineers?
10mRethinking collaboration: LLMs, teams, and cognitive load
09mLessons Learned in AI Engineering
20mFireside Chat
32mWhat can we expect of LLMs as software engineers?