Beyond the code: Rediscovering the timeless skills of software engineering in the AI era
summary
AI coding demos (Codex / "C code", Cursor, cloud code) produce runnable pages but fail when specs grow: large context makes LMs lose focus — context degradation — and chaining many subtasks drops end‑to‑end success (example math: 99% per step → ~36% overall). A generated e‑commerce product page embeds price and stock so it cannot support variants or layered promotions. Effective engineering skills are breaking big problems into small, context‑free tasks and choosing architectures by weighing tradeoffs so solutions remain maintainable and extensible.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:13Breaking down problems
- 4:10AI’s context limits
- 9:35Problem breakdown skill
- 13:34Trade-offs in structure
- 17:49Solution trade-offs skill
- 20:42Trade-off thinking scale
- 24:18Key takeaways
- 26:50Value in AI era
related talks
26mBeyond the code: Rediscovering the timeless skills of software engineering in the AI era
26mBeyond the code: Leading engineering teams in the GenAI era
56mFundamentals of Software Engineering In the age of AI
39mBack to Basics: Cutting Through the AI-Generated Code Noise
09mLessons Learned in AI Engineering
15mUsing AI effectively in large Java codebases
45mThe Rebirth of Software Craftsmanship in the AI Era
01h 01m10 tips to level up your ai-assisted coding