The Phoenix Architecture

Generative AI coding demands what we've always known: modularity, clear boundaries, disposable components. Principles that scaled human teams are now table stakes. Here, we make the implicit explicit

292Readers

28Posts

15d agoLast active

The Spec Can Come Later

Sep 28, 2026

#ai#software

SaaS Isn’t Dead. Sameness Is.

Aug 24, 2026

#saas#business

Most of Your Architecture Was Just Expensive Code

Aug 21, 2026

#architecture#software

The Specification Is Not a Document

Aug 19, 2026

#ai#architecture

When Does a Specification Become a Program?

Aug 11, 2026

A spec can contain prose, tests, schemas, formal models, even executable code. The useful boundary is not the syntax but, rather, what survives: if you throw the implementation away, what has to remain so you can build the next one?

#software#architecture

The Implementation Remembers

Jun 14, 2026

Every mature system is carrying around lessons that were never written down. They’re encoded as validations, retries, timeouts, workflows, and exceptions. The implementation remembers. The organization forgets

Production Is a Compiler Input

Apr 20, 2026

#ai#software

The Phoenix Primitives

Apr 13, 2026

The architecture of a regenerative system is defined entirely by what you can't delete.

#coding#ai

The Generative Stack

Apr 7, 2026

Trying to find the best tool or platform for generative software in 2026 is a mistake that could haunt you for decades

#ai#coding

The Conversation Is the Commit

Mar 26, 2026

#ai#software

Compile to Architecture

Mar 6, 2026

For a long time we’ve treated frameworks as the target of software development. But if systems are meant to be regenerated and replaced safely, the real compilation target has to be the architecture itself.

#ai#coding

The Regenerative Grain

Feb 19, 2026

In 2014 I gave a talk called Tiny (keep things small enough to understand) In 2026, small means something different. Small means safe to delete. New in the Phoenix Architecture series: The Regenerative GrainI