Craig Sproule — WritingEssays on enterprise software, no-code systems, AI coding, and abstraction.https://craigsproule.pages.dev/The Ladder That Climbed Itself: Software’s Progressive Removal of the Humanhttps://craigsproule.pages.dev/articles/the-ladder-that-climbed-itself/https://craigsproule.pages.dev/articles/the-ladder-that-climbed-itself/Software history is often told as a sequence of languages. A more useful view is a sequence of decisions humans no longer have to express.Wed, 12 Aug 2026 00:00:00 GMTNo-Code’s Ceiling and the Sideways Delivery of Its Promisehttps://craigsproule.pages.dev/articles/no-codes-ceiling-and-the-sideways-delivery-of-its-promise/https://craigsproule.pages.dev/articles/no-codes-ceiling-and-the-sideways-delivery-of-its-promise/No-code did not eliminate software development. It relocated much of the promised benefit into platforms, components, and AI-assisted workflows.Wed, 05 Aug 2026 00:00:00 GMTBottleneck Migration: What a Decade of No-Code Reveals About AI Coding Claimshttps://craigsproule.pages.dev/articles/bottleneck-migration/https://craigsproule.pages.dev/articles/bottleneck-migration/When software gets easier to generate, the constraint does not vanish. It moves—from implementation toward definition, verification, and control.Wed, 29 Jul 2026 00:00:00 GMTWhen Abstractions Are Extracted, Not Received: LLMs and the Break in the Software Ladderhttps://craigsproule.pages.dev/articles/when-abstractions-are-extracted-not-received/https://craigsproule.pages.dev/articles/when-abstractions-are-extracted-not-received/Traditional abstractions are intentionally designed and handed down. Language models infer usable patterns from examples, changing how software layers are formed.Wed, 22 Jul 2026 00:00:00 GMTFrom Machine Code to Prose: How Coverage Stopped Scaling with Headcounthttps://craigsproule.pages.dev/articles/from-machine-code-to-prose/https://craigsproule.pages.dev/articles/from-machine-code-to-prose/As software moved from machine instructions toward natural language, the amount one person could express grew faster than the teams available to implement it.Wed, 15 Jul 2026 00:00:00 GMT