Do developers still need to review every line of AI-written code?
Ruby on Rails creator David Heinemeier Hansson (DHH) posted that believing humans must still review every line of AI-generated code is itself a form of 'AI psychosis,' prompting a wide debate among developers about where human code review still matters.
Side A
Line-by-line human review is outdated
This camp argues that AI coding agents now produce code too quickly for humans to read every line. DHH says that thinking AI is 'just another tool' or that humans must still review every line is a mistake people need to snap out of. Others say manual review has become the main bottleneck in shipping software, and that checking all agent output by hand is futile outside specific industries.
Evidence posts
Overlap
Where they overlap
- Both camps say code review has become the main bottleneck now that AI agents write much of the code.
- Several voices on both sides say high-stakes or irreversible changes still warrant some form of human check.
Side B
Human review is still essential
This camp argues that people must keep reviewing AI-written code before it ships. Some say they do not believe human code review can be replaced if the goal is well-written software. Others say they still review every line touching money, auth, secrets, dependencies or the database, that anyone who cannot review AI code should not ship it, and that irreversible or financially consequential changes need a human in the loop.
Who took this side
No public figures have taken this side on X yet
Evidence posts
Reader vote
Be the first to weigh inWhere do you land? Anonymous, one vote per reader. Sign in with X to be counted as a verified reader.
Free newsletter
Both Sides Daily
One email a day with the day's debates: what each side says, where they meet, and how readers voted. No spam, unsubscribe with one click.
This page was compiled by AI and published after automated checks. Something wrong?Report a correction