Every team with vibe coders working alongside traditional coders is struggling with the same thing today. Engineering teams are trying to run a relay race in which some of the runners are going at a snail’s pace and others are lapping them ten times a second — and each has a part to play. How do you coordinate that?
“Vibe coding” is barely a year old as a phrase. Andrej Karpathy coined it last February, describing a way of working where you “fully give in to the vibes, embrace exponentials, and forget that the code even exists.” By November, Collins had made it Word of the Year, defined as “the use of artificial intelligence prompted by natural language to write computer code.” A phrase goes from a post to the dictionary in nine months when it names something a great many people are suddenly doing.
Speed, meanwhile, is harder to measure than it looks from the outside. METR ran a randomized study last summer with 16 experienced open-source developers and 246 real issues and found that “when developers are allowed to use AI tools, they take 19% longer to complete issues.” The finding that should worry every manager is the perception gap: “Developers expected AI to speed them up by 24%, and even after experiencing the slowdown, they still believed AI had sped them up by 20%.” The relay team can’t even agree on who’s fast.
And the volume is real regardless. Google’s DORA report in September found 90% of respondents using AI at work, and its central claim was that “AI doesn’t fix a team; it amplifies what’s already there.” More change, faster — and “without robust control systems, like strong automated testing, mature version control practices, and fast feedback loops, an increase in change volume leads to instability.” That’s the relay race in a report’s language: the baton arrives faster than the hands can close on it.
All of these teams are seeking solutions. Many are developing them. I very much hope something sticks. How to coordinate the two speeds is still the open question.
