Are shared libraries on their way out?
My latest for LeadDev reviews a 3-year retrospective study from GitClear that sheds light on the effects of AI on the shape of code.
Three years after the AI takeover began in coding, we can now see clear before-and-after effects:
Code duplication is up 81%.
Code reuse is down 70%.
Legacy refactoring is down 74%.
To get these figures, GitClear's latest study analyzed 623 million code changes from 2023 to 2026.
My latest for LeadDev looks at the report, and I also synced with the author, Bill Harding on what the findings mean for engineering teams.
Another interesting signal is what he calls "obfuscation." AI will go through hoops to avoid throwing errors, often masking them instead of addressing the underlying cause. It's up 47%.
Signals point to difficulties maintaining codebases at this rate... AI helps teams move faster, but at the cost of duplicated code, legacy systems gathering dust, and hidden bugs.
Is the era of shared libraries fading?











