Software Craftsmanship in the Age of AI

2026-03-13 · Source: AI & ML – Radar · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Intermediate, medium

Summary

The O'Reilly AI Codecon, scheduled for March 26, will explore the evolving concept of "software craftsmanship" as AI agents increasingly write code. The event features speakers addressing two main perspectives: the "dark factory" approach, where AI agents handle most implementation with human oversight for direction and review, exemplified by Ryan Carson's Antfarm and OpenClaw, and the "craftsmanship-means-oversight" position, championed by Addy Osmani and Cat Wu from Anthropic, which emphasizes active human involvement, coordination patterns, and "context engineering." Speakers like Nicole Koenigstein and Hila Fox will discuss the hidden costs and failure modes of agentic AI in production. Wes McKinney will present on "The Mythical Agent-Month," arguing that agents introduce new accidental complexity and that design talent remains the bottleneck. The conference also covers new architectures for AI-native engineering, including Juliette van der Laarse's "AI Flower" framework and Mike Amundsen's focus on augmentation over automation. Aaron Levie of Box will discuss how agents augment enterprise software by unlocking previously unaffordable work.

Key takeaway

For CTOs and VPs of Engineering navigating AI integration, your teams should prioritize developing expertise in system design, agent orchestration, and "context engineering" over traditional coding. The shift from hands-on coding to managing AI-generated code means that conceptual integrity and design judgment are now the most critical skills. Invest in frameworks and tools that enable effective human oversight and robust failure mode analysis, rather than solely pursuing maximum agent autonomy, to ensure long-term system reliability and maintainability.

Key insights

Software craftsmanship is shifting from manual coding to designing, orchestrating, and overseeing AI-driven development systems.

Principles

• Code is becoming a liquid, not a craft.
• Taste and design judgment are the new bottlenecks.
• Augmentation amplifies human expertise, automation replaces it.

Method

Implement a "dark factory" model where planning agents decompose features, separate agents implement and test stories, and humans review video-recorded output for quality assurance.

In practice

• Explore agent-to-agent collaboration for complex tasks.
• Focus on context engineering for reliable LLM performance.
• Design AI systems for human oversight and steerability.

Topics

• AI Agents
• Software Craftsmanship
• Code Generation
• Multi-Agent Systems
• AI Workflow Orchestration

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Software Engineer, AI Engineer, MLOps Engineer

Related on AIssential

• The Month AI Woke Up — AI agents have matured to enable autonomous task execution, fundamentally altering programming and impacting market valuations.
• FOD#141: What Happens to Software Engineering When Anyone Can Build? — AI agents are redefining software engineering, splitting it into harness engineering and human judgment manufacturing.
• Why AI is raising the stakes of the software engineering craft — AI elevates software craftsmanship by shifting focus to human orchestration, auditing, and deep system understanding to prevent cognitive debt.
• Steve Yegge Wants You to Stop Looking at Your Code — AI agents will transform coding into orchestration, shifting human roles to managing complex problems and leveraging taste.
• Judgment and creativity are all you need. — Coding agents move software development bottlenecks from implementation to judgment and creativity.
• Sea's View on the Future of Agentic Software Development with Codex — Agentic AI coding tools fundamentally shift software development by enhancing contextual awareness and enabling higher-level cognitive work.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI & ML – Radar.