Andrew Ng's new skills map, built on 10,000+ job posts, says fundamentals still matter
techNmak · x · 2026-09-18
Reacting to the claim that "AI writes code now, so software engineering doesn't matter," this post leans on Andrew Ng's newly released AI Engineering Skills Map. Built from analysis of 10,000+ job postings, dozens of structured interviews with AI experts, hiring managers and recruiters, plus survey data, it still lists software engineering fundamentals as one of four core skill areas.
Key points:
- Agents can generate schemas, backends, APIs, caching, auth, tests and deployment configs — but can't decide for you: sync vs async, consistency guarantees, concurrent updates, rate-limited dependencies, where state should live, what breaks at 100x traffic.
- Vibe coders who don't understand the system may not even know which tradeoffs exist, so agents make poor latency/availability/consistency/cost choices without the context to do better.
- The map spans full-stack building, data management, architecture design, security and reliability, and production scaling/operations.
Bottom line: an agent can write lots of code; you still have to judge whether what it built is a good system.
More from coding & agent
- AI-run company makes its first $75: what running a business with Devin taught us — sandylikesfrogs · 2026-09-18
- Codex ships Code Mode: JS programs filter MCP data before it hits context — alex_frantic · 2026-09-18
- W&B floats chart annotation feature: pin notes like '58k: added attention gate decay' to training runs — yoavartzi · 2026-09-18
- Claude Code relaunches Projects to manage multiple cloud agents — The Verge AI · 2026-09-18
- Scale AI CEO Alexandr Wang hypes Muse agent: 'surprised by how much it can take on' — alexandr_wang · 2026-09-18
- Included Health built federated healthcare agents on LangGraph, lifting chat engagement 75% — LangChain · 2026-09-18