Martin Fowler says AI represents the biggest shift in programming since the move away from assembly language, requiring a new mindset because LLMs are fundamentally nondeterministic rather than deterministic. He sees AI as most valuable today for rapid prototyping and understanding legacy systems, while warning that using it to safely modify complex code still demands careful structure, review, and tolerance-aware thinking. Find out what else he has to say about AI in Senior Editor Joab Jackson’s article.

In other news:

• Contributing writer Shanea Leven says that most generative AI apps fail because they rely on “vibe coding” and prompt tweaking rather than robust engineering systems. To succeed in production, AI needs a comprehensive ecosystem — evaluation, monitoring, memory, integration, and continuous optimization — treating AI as a system, not a demo.
• In this article, writer Temitope Oyedele shows you what happens when you follow current best-practice prompting and why it still creates friction at scale. You’ll learn how composable architecture closes that gap by introducing a framework for reuse, versioning and collaboration. You’ll also discover how Bit Cloud and Hope AI make that system practical by scaffolding modular components that persist beyond a single project.

Three months ago, I spent two weeks building a custom plugin to connect our AI assistant to our internal CRM system. Last week, I replaced it with a Model Context Protocol (MCP) server that took four hours to implement, and it works with every AI model in our stack. This is not just a productivity win, but a glimpse into a fundamental...

Traditional cloud DR stops at data. Real recovery depends on identity, networking, scaling, and config state. This white paper shows how rebuilding cloud infrastructure as code makes recovery predictable, auditable, and repeatable—when outages happen.

Have you downloaded our newest book yet? This playbook brings together lessons from global organizations and provides a framework to measure true ROI, reduce operational toil, and foster a culture of experimentation and learning.