E-Ink News Daily

Back to list

Conceptual integrity and counting lines of code

Simon Willison argues that lines of code can be a legitimate productivity metric when AI coding agents are involved, since agents can produce far more debugged code than humans. However, he identifies cognitive capacity as the new bottleneck, explaining why teams remain necessary despite agents dramatically increasing individual output.

Background

Simon Willison is a well-known developer and blogger who frequently writes about AI's impact on software development. This excerpt comes from his Talking Postgres podcast conversation with Claire Giordano on how AI is changing software development.

Source
Simon Willison
Published
Aug 20, 2026 at 06:46 AM
Score
6.0 / 10