Netlify CTO Dana Lawson: Writing Code Is No Longer the Job

Netlify CTO Dana Lawson made a provocative claim: "Writing code is no longer the job." In a recent interview, she argues that the role of developers is shifting from writing code to orchestrating AI agents and designing user experiences. Engineers become "agent experience engineers" who curate agent outputs, set boundaries, and manage system reliability—not write every line.
Lawson emphasizes that developers will still need deep technical understanding to verify agent-generated code, but the daily activity changes to prompt engineering, feedback loops, and integration testing. She draws a parallel to how cloud abstraction shifted ops work: "Just like we stopped managing servers, we'll stop writing boilerplate code."
The interview touches on Netlify's internal use of AI coding agents. Lawson says they already use agents for routine tasks like writing tests and configuration files, freeing engineers to focus on architecture and user-facing features. She warns that teams who resist this shift will struggle to keep velocity: "The bottleneck is no longer writing code—it's deciding what to build and how to orchestrate."
Key takeaways for developers using AI agents
- Shift mindset: Treat agents as junior engineers that need supervision, not replacements. Review their output as you would a pull request.
- Design for agent interaction: Build clear APIs, documentation, and feedback loops so agents produce correct code without manual handholding.
- Focus on experience: With agents handling implementation, developers spend more time on system design, user research, and edge-case handling.
The full discussion on Hacker News (49 comments) debates whether this vision is realistic or hype. Some commenters note that agents still require heavy debugging for complex logic, but agree that the trend toward orchestration is inevitable.
📖 Read the full source: HN AI Agents
👀 See Also

Anthropic restricts Claude subscription use with third-party harnesses including OpenClaw
Anthropic announced that starting April 4 at 12pm PT/8pm BST, Claude subscription limits can no longer be used with third-party harnesses like OpenClaw. Users will need to enable extra usage with separate pay-as-you-go billing for these integrations.

NYC Hospitals End Palantir Contract as UK Expansion Faces Scrutiny
New York City's public hospital system will not renew its $4 million contract with Palantir in October, transitioning to in-house systems. Meanwhile, Palantir faces privacy concerns over its £330 million NHS deal and new UK financial regulation contract.

Claude-Code v2.1.74 Release: Memory Leak Fixes, Context Optimization, and Plugin Improvements
Claude-Code v2.1.74 fixes a critical memory leak in streaming API responses that caused unbounded RSS growth on Node.js/npm code paths. The update adds actionable suggestions to the /context command and introduces the autoMemoryDirectory setting for custom auto-memory storage.

SDL Project Bans AI-Written Commits in Response to GitHub Issue
The SDL project has implemented a policy banning AI-generated commits after a GitHub issue raised concerns about Copilot usage in code reviews. The issue specifically mentions reviews #13277 and #12730 as examples where AI assistance was detected.