Willison argues lines of code matter for measuring coding agent output
Simon Willison explores a counterintuitive argument about measuring coding agent productivity through lines of code, challenging conventional wisdom that dismisses LOC as a metric. In a podcast discussion on AI's impact on software development, Willison articulates when code volume becomes a legitimate signal of agent capability rather than a vanity metric. This reframes how teams should evaluate LLM-powered coding tools, shifting focus from abstract efficiency claims to concrete output measurement. The insight matters for practitioners deploying agents in production, as it clarifies what productivity signals actually matter when humans and AI collaborate on codebases.
Modelwire context
Skeptical readWillison doesn't just defend LOC as a signal; he argues it becomes legitimate only when agents operate within constraints that preserve conceptual integrity. The qualifier buried here is crucial: raw code volume without architectural coherence is still noise.
This connects directly to Jeremy Morrell's point about extensible software economics (from our August 19 coverage). Morrell showed that LLMs collapse the cost of building user-facing customizations within sandboxed boundaries. Willison's argument about measuring agent output through code volume only makes sense in that same bounded context. If agents are generating thousands of lines of unvetted core logic, LOC becomes a liability signal, not a capability signal. The two pieces together suggest the real productivity win isn't 'more code faster' but 'more targeted code in the right architectural layer.'
If teams deploying LLM coding agents report that their best productivity gains came from agents working on plugin layers or isolated modules (rather than core system changes), that validates Willison's framing. Conversely, if the highest-value agent contributions turn out to be refactoring or deletion rather than net new lines, the LOC metric collapses again.
Coverage we drew on
- Quoting Jeremy Morrell · Simon Willison
This analysis is generated by Modelwire’s editorial layer from our archive and the summary above. It is not a substitute for the original reporting. How we write it.
MentionsSimon Willison · Claire Giordano · Talking Postgres · Claude
Modelwire Editorial
This synthesis and analysis was prepared by the Modelwire editorial team. We use advanced language models to read, ground, and connect the day’s most significant AI developments, providing original strategic context that helps practitioners and leaders stay ahead of the frontier.
Modelwire summarizes, we don’t republish. Simon Willison originally reported this story as “Conceptual integrity and counting lines of code”. The full content lives on simonwillison.net. If you’re a publisher and want a different summarization policy for your work, see our takedown page.