Why “programmer laziness” matters more in the LLM era

A recent article by Bryan Cantrill revisits Larry Wall’s “laziness” as an engineering constraint that drives cleaner abstractions and simpler systems. He argues LLMs can amplify bloat without human taste and limits.

Why “programmer laziness” matters more in the LLM era

TL;DR

  • Revisits Larry Wall’s virtues, focusing on laziness as a long-term software engineering discipline
  • Laziness as constraint: hard thinking to form clean abstractions and keep systems simple (but no simpler)
  • Benefits: less repetition, lower cognitive load, easier composition and extension over time
  • Critiques “false industriousness” culture prioritizing output volume over design quality
  • LLMs misalign incentives: effectively free work, little pressure to reduce scope, delete code, or tighten design
  • Example cited: Garry Tan’s 37,000 LOC/day claim; teardown found redundant assets, harnesses, generated-like baggage

“The peril of laziness lost” is a sharp (and very timely) meditation on an old programming cliché: Larry Wall’s trio of virtues—laziness, impatience, and hubris—with a particular focus on what “laziness” is supposed to mean when software is built to last.

Laziness as an engineering constraint, not a vibe

The piece leans on the Camel Book’s tongue-in-cheek framing to make a serious point: real programmer laziness isn’t about avoiding work, it’s about doing the hard thinking needed to build clean abstractions and keep systems simple (but no simpler). That effort pays future dividends—less repetition, lower cognitive load, and software that’s easier to compose and extend.

It also draws a line from hammock-driven development to a much less flattering modern counterpart: a culture of “false industriousness” that prizes output volume over design quality.

Where LLMs amplify the wrong incentives

The most pointed section lands on what happens when LLMs enter that environment. The argument isn’t that LLMs are useless—it’s that they inherently don’t share the core constraint that makes human laziness valuable. For a model, work is effectively free, and there’s no built-in motivation to reduce scope, delete code, or converge on a tighter design.

A viral example gets cited: Garry Tan publicly claiming a pace of 37,000 lines of code per day, followed by a teardown from another engineer that found a mess of baggage in the resulting app—extra harnesses, redundant assets, and other artifacts that read like “generated” in the least flattering sense.

LLMs as tools under human taste and limits

The author’s closing idea is more grounded than reactionary: LLMs can still be an “extraordinary tool” in software engineering, but only if used in service of human constraints and judgment—the instinct to simplify, to pay down debt, and to keep systems understandable for whoever inherits them next.

Original source: https://bcantrill.dtrace.org/2026/04/12/the-peril-of-laziness-lost/

Continue the conversation on Slack

Did this article spark your interest? Join our community of experts and enthusiasts to dive deeper, ask questions, and share your ideas.

Join our community