Eighteen Years In, We Still Can’t Agree on How to Build Software

2026-06-21 · Source: Artificial Intelligence on Medium · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning, Project & Product Management · Depth: Intermediate, medium

Summary

The article critiques the persistent challenges in software development processes, specifically critiquing the effectiveness of frameworks like Scrum over 18 years. It highlights that while Scrum organizes communication, it often fails to address underlying technical debt, weak testing, and design rot, a phenomenon Martin Fowler termed "Flaccid Scrum" in 2009. The author notes these fundamental engineering issues persist in 2026, often exacerbated by management pushing dogmatic processes without technical understanding. AI tools, while useful for tasks like testing (62% of respondents in GitLab's 2026 German DevSecOps survey), documentation (63%), and code reviews (49%), are presented as amplifiers of existing discipline, making good bases better and bad bases worse. The core problem remains a lack of agreement on basic craft standards and team autonomy in defining their own processes, leading to a cycle of new frameworks failing to fix foundational engineering problems.

Key takeaway

For Software Engineering Managers evaluating new process frameworks or AI integration, recognize that process alone cannot compensate for technical debt. Your focus should be on establishing a strong engineering foundation, including robust testing, continuous refactoring, and a meaningful "Definition of Done." Empower your teams to define their own workflows and protect them from unsustainable crunch, as AI tools will only amplify existing discipline, not create it.

Key insights

Software development frameworks organize talk, but engineering fundamentals like testing and refactoring determine code health.

Principles

• Frameworks organize communication, not code quality.
• Technical debt undermines process effectiveness.
• AI amplifies existing engineering discipline.

In practice

• Prioritize real tests and refactoring.
• Establish a "Definition of Done with teeth."
• Grant teams decision-making power over process.

Topics

• Software Development Process
• Agile Methodologies
• Scrum
• Technical Debt
• AI in Software Development
• Engineering Discipline
• Team Autonomy

Best for: Software Engineer, Director of AI/ML, Consultant

Related on AIssential

• Questions to rate your team, business impact on unhealthy code, and weekly readings 💡 — Optimizing software development requires structured team assessment, prioritizing code quality, and strategic AI integration for efficiency and career adaptation.
• don't let AI rob you — AI should augment, not replace, the intrinsic joy and learning derived from the software engineering journey.
• It Doesn’t Help To Push AI Into A Crappy Process — AI amplifies existing software development process quality; good processes benefit, poor ones worsen.
• 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.
• Software Engineering: Beyond Coding — A Comprehensive Exploration of the Discipline, Skills, Career… — Software engineering is a systematic discipline focused on the entire lifecycle of robust, scalable, and ethical software systems.
• Build to Last — Chris Lattner talks with Jeremy Howard — Durable software requires deep understanding, robust architecture, and craftsmanship, resisting AI hype and superficial productivity metrics.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence on Medium.