The Augmented Engineer: Evolving Sensible Defaults through Human-Centric AI
summary
Argues that AI should be a disciplined assistant, not an owner: teams keep humans in the loop, retain architectural decisions and responsibility, and treat generated code as assume code is wrong. Uses concrete practices: a one‑day‑per‑week one day per week AI‑first experiment (finds it slower until humans stay involved), create an agent.md with project context, prune community GitHub skills (React example) for Kotlin projects, and do red‑green‑refactor with agent TDD. These measures reduce plausibility traps and keep ownership visible.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 6:48Sensible defaults in software
- 11:02AI as tool not replacement
- 17:52Ownership with AI
- 24:17What to delegate to AI
- 27:29Security with AI code
- 31:30AI collaboration techniques
- 35:43Summary
related talks
27mAgentic Engineering: Working With AI, Not Just Using It
01h 00mKenote: AI Didn’t Replace You. It Promoted You!
27mEngineering in the Age of AI: Making AI Work for Us, Not Replace Us
23mHuman-Centered AI: Supporting, Not Replacing, Frontend Developers
40mHuman Centric Software Development Augmented by AI
20mHuman above AI
15mThe AI Shift: Rethinking the Developer’s Role in a Machine-Driven World
36mEngineering Intent: Solving the AI Productivity Paradox