80% of dev candidates now say they direct AI agents instead of writing code, sparking HN debate on post-AI hiring
mdwelsh · hn · 2026-09-20
An Ask HN thread highlights a hiring dilemma: roughly 80% of interviewed dev candidates now say they rarely write code themselves, instead directing AI agents. The poster wants to verify candidates can still code and reason about system design, but questions whether conventional leetcode and design interviews still make sense in a post-AI world.
Key discussion points:
- Should traditional coding and system design interviews be kept or abandoned?
- How to evaluate candidates who claim a fully agent-driven workflow?
- Or should interviewers just accept that everyone will use tools like Claude Code regardless?
The thread captures the growing mismatch between testing for hand-written coding skills and a job reality that increasingly amounts to orchestrating agents.
More from coding & agent
- Mnemos MCP nears launch after shared memory layer fix — RileyRalmuto · 2026-11-04
- Moving mobile dev to remote machines makes it multiplayer across product, design, engineering — rudrank · 2026-11-04
- Kuaishou's EvoRec: self-evolving agentic recommender lifts ad revenue 4.6% — _reachsumit · 2026-11-04
- TRIO: three specialized agents with aligned objectives refine retrieval-augmented generation — _reachsumit · 2026-11-04
- How Multica uses Multica to build itself: full human-agent dev workflow — jiayuan_jy · 2026-11-04
- Claude Code fixes game copyright refusal rule in 2.1.317: naming a game now works — apples_jimmy · 2026-11-04