Ars Technica reports that modern AI coding assistants and agents can produce huge amounts of working code, but programmers know better than to trust its accuracy, so substantial effort goes into reviewing the output.
A recent study of actual coding practices, which Ars discusses, finds that more code does not translate into more finished software.
Source: arstechnica.com, reported 10-09. https://arstechnica.com/ai/2026/10/ai-coding-agents-generate-more-code-but-not-more-software/