Acceleration whiplash 📉

2026-06-24 · Source: Refactoring · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning · Depth: Intermediate, quick

Summary

The Faros report, based on a survey of over 22,000 developers across more than 4,000 teams, reveals a concerning trend in software development due to AI coding adoption. It indicates that despite an increase in activity, with developers touching +67% more PRs daily and completing +33% more tasks, the overall quality is declining, and shipping work has become harder. The report highlights a significant slowdown in the "last mile," showing a +26% increase in tasks stalled for 7+ days and a +14% rise in work restarts. Consequently, lead time has skyrocketed by +480%, driven by an average +80% waiting time between pipeline steps, with code reviews identified as a major bottleneck.

Key takeaway

For Directors of AI/ML or engineering leads evaluating AI coding tool adoption, recognize that increased developer activity does not equate to faster delivery or higher quality. Your teams may experience a +480% lead time increase and significant bottlenecks in code reviews, even if they are high-performing. Prioritize optimizing downstream SDLC stages, especially code review processes, before scaling AI-assisted code generation to prevent poor quality from reaching production.

Key insights

AI coding increases development volume but degrades quality and slows shipping across all team performance levels.

Principles

• AI amplifies existing development pipeline issues.
• Increased work-in-progress without throughput growth inflates lead time.
• High-performing teams are not immune to AI's downstream degradation.

Topics

• AI Coding
• Software Development Life Cycle
• Developer Productivity
• Code Reviews
• Lead Time
• Faros Report

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

Related on AIssential

• AI Coding Assistants Haven’t Sped up Delivery Because Coding Was Never the Bottleneck — AI coding tools shift bottlenecks from coding to specification and verification, demanding new team structures and workflows.
• AI Made Your Engineers 10x Faster and Your Product 10x Worse — AI coding assistants boost speed but degrade product quality, necessitating a rebuilt quality layer.
• Agentic Code Review — AI agents make code writing cheap, shifting the software engineering bottleneck to human verification and review.
• AI Is Writing More Code Than Ever. So, why is Software Quality Getting Worse? — The provided content presents an article titled "AI Is Writing More Code Than Ever.
• The Final Bottleneck — Accelerated code generation creates new bottlenecks in review and integration, leading to unsustainable backlogs.
• AI in the SDLC: Rethinking AI Coding Tools & AI Agents — True AI productivity in SDLC requires redesigning workflows around AI, not just augmenting coding.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Refactoring.