{"slug": "does-human-involvement-actually-make-agent-code-worse", "title": "Does Human Involvement Actually Make Agent Code Worse?", "summary": "Jellyfish analyzed AI telemetry and engineering workflow data across 300,000 developers at 1,300 companies and found that agent-written pull requests merged with more human involvement were reverted roughly twice as often at each step of involvement. The company attributes the pattern to triage rather than worse code: engineers wave easy PRs through and focus on riskier ones, and reviewed agent PRs showed no statistically meaningful difference in reverts, follow-up fixes, or bugs versus unreviewed ones. Jellyfish also reported that the typical developer now does 26% more code reviews per week since early 2025, almost all of it agent-assisted work, and that switching on an AI reviewer did not reduce human review load.", "body_md": "Whether you’re trying to build a software factory or already running one, you’ve probably heard some version of the pitch: agents can write the code now, and humans are the bottleneck. Get people out of the loop and let the agents cook.\n\nAnd for most teams, review is exactly where things get stuck. At the top 10% of companies, agents are now writing a majority of merged PRs, but a person still signs off on the vast majority of them before they merge.\n\nIf people are the bottleneck, it’s fair to ask what all that human attention is actually buying us. Does it make agent code better?\n\nAt Jellyfish we analyze AI telemetry and engineering workflow data across 300,000 developers at 1,300 companies. When we looked at agent-written PRs that merged and checked how often they got reverted within a month, the answer depended on how involved a person had been along the way:\n\nThe more a human got involved, the more often the code got rolled back, roughly doubling at each step. So are humans making agent code worse?\n\nWell… not quite. What you’re actually seeing is the effect of triage, powered by good human engineering judgment.\n\nEngineers are good at spotting risk, so they wave the easy PRs through, approve the middling ones, and roll up their sleeves on the perilous ones. And since hard things are hard, the riskiest PRs get reverted much more often, even with a person helping. (These gaps hold up when you compare PRs of the same size within the same company, so this isn’t just true for big PRs or messy orgs.)\n\nThe triage is working, too. When we compare agent PRs that nobody reviewed against similar ones that were reviewed, there’s no statistically meaningful difference in reverts, follow-up fixes, or bugs. For the work they’re choosing to let through, engineers are holding the line on quality.\n\nThe problem is that at most orgs, engineers are indeed holding the line, but at great effort and expense. This triage that keeps quality in check is maintained by hand, one PR at a time, at the very end of the process (which is the most expensive place to do it). Since early 2025, the typical developer is doing 26% more code reviews per week, and essentially all of that growth is work a coding agent helped write. Review has become a second job stacked on top of the first one.\n\nYou might expect AI review to pick up the slack, but so far it hasn’t. When a repo switches on an AI reviewer, people still do about two reviews per PR, and it still takes about three hours to merge. The bot is typically one more voice on the PR, not a decision that offloads work from a human.\n\nThe good news is that the hard part, knowing which changes are risky, already exists. It just lives in your engineers’ heads. So get it out:\n\nThe fastest way past the review bottleneck probably isn’t fewer humans or more bots. It’s getting the judgment your engineers already have out of their heads and into the process.", "url": "https://wpnews.pro/news/does-human-involvement-actually-make-agent-code-worse", "canonical_source": "https://dev.to/dailycontext/does-human-involvement-actually-make-agent-code-worse-74d", "published_at": "2026-11-11 01:17:28+00:00", "updated_at": "2026-11-11 01:21:28.812046+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "developer-tools", "artificial-intelligence"], "entities": ["Jellyfish"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/does-human-involvement-actually-make-agent-code-worse", "markdown": "https://wpnews.pro/news/does-human-involvement-actually-make-agent-code-worse.md", "text": "https://wpnews.pro/news/does-human-involvement-actually-make-agent-code-worse.txt", "jsonld": "https://wpnews.pro/news/does-human-involvement-actually-make-agent-code-worse.jsonld"}}