Recently, I’ve seen posts from people who should know better, claiming things like “We’ve ditched Agile” or “AI lets us ship daily instead of in sprints.” Each time, I’m reminded how Agile, a guiding philosophy, is often confused with Scrum, a specific framework.

Here’s the Agile Manifesto in it’s entirity:

We are uncovering better ways of developing
software by doing it and helping others do it.
Through this work we have come to value:

Individuals and interactions over processes and tools
Working software over comprehensive documentation
Customer collaboration over contract negotiation
Responding to change over following a plan

That is, while there is value in the items on
the right, we value the items on the left more.

That’s it. Nine lines. There is nothing in there about standups, sprints, retros, planning, etc., because Agile is not a methodology.

Many methodologies come from Agile principles. XP was once prominent and still matters. Kanban remains popular, but Scrum now dominates. Scrum is often used without considering team workflows or dynamics. Organisations default to Scrum, but effective use is rare.

In Scrum, engineers often estimate time poorly. This causes incorrect ticket sizing, finishing work too early, or missing sprint deadlines. Even with effective Scrum delivery, the process feels like a hamster wheel, as teams deliver feature after feature under the stern scrutiny of the Scrum master.

The “Agile is dead!” crowd misses that even in the era of AI, the original manifesto remains relevant.

Switching to a JIT Kanban-esque methodology can’t happen overnight, even if your tools produce ten times as much code. You need an effective CI system tailored to your workflows, rigorous testing and quality checks, and resilient infrastructure that can roll back mistakes before customers are affected.

It’s one thing that I like about AI. Our industry must fix broken processes, follow best practices, and tighten up how we work to leverage it effectively. And if that means technical leaders having to learn a bit of history around process design, methodologies, etc., then so be it. AI also has the interesting side-effect of making some ideas that have fallen out of favour viable again. Think along the lines of Progressive Enhancement, HATEOAS, XSLT, etc. Something that may have been too rigid or time-consuming for humans to implement properly is now a good fit for AI-driven development.

Teams that embrace this shift will excel with AI-powered delivery, quality products and happy customers. Teams that don’t risk frequent outages and bugs.