--- title: The strangest developer productivity metric of all time description: Tokenmaxxing — measuring developer productivity by raw AI token consumption — is emerging as a damaging trend in software engineering. Meta's internal... canonical: https://daily.dev/posts/the-strangest-developer-productivity-metric-of-all-time-l3wmuekbe twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: The strangest developer productivity metric of all time | daily.dev og:description: Tokenmaxxing — measuring developer productivity by raw AI token consumption — is emerging as a damaging trend in software engineering. Meta's internal... og:url: https://daily.dev/posts/the-strangest-developer-productivity-metric-of-all-time-l3wmuekbe og:image: https://api.daily.dev/og/posts/l3WMuEkBe.png og:image:alt: The strangest developer productivity metric of all time og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # The strangest developer productivity metric of all time **[InfoWorld](https://daily.dev/sources/infoworld)** · 8 min read · 1 upvotes · 0 comments ## Summary Tokenmaxxing — measuring developer productivity by raw AI token consumption — is emerging as a damaging trend in software engineering. Meta's internal 'Claudeonomics' leaderboard drove 85,000+ employees to burn 60.2 trillion tokens in 30 days, causing site outages. Amazon's Kirorank was shut down within weeks after engineers gamed it by spinning up idle agents. Uber reportedly burned its entire annual AI budget in one quarter. Data from 600 million code changes shows code duplication up 81%, refactoring down 70%, and short-term code churn more than doubling from 3.3% to 7.1%. The core problem is Goodhart's Law: when token volume becomes the target metric, it stops measuring anything useful. High-value AI use — debugging, design, analysis — generates few tokens but delivers real engineering outcomes. The piece argues for returning to first principles: thinking, refining, and planning before generating code. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What was Meta's Claudeonomics leaderboard and why was it shut down? Claudeonomics was an internal Meta leaderboard that ranked the top 250 of over 85,000 employees purely by AI token consumption, awarding titles like 'Token Legend' and 'Session Immortal.' In one 30-day period, employees burned through 60.2 trillion tokens, driving waste, throwaway agent work, and site outages caused by careless AI code generation, prompting the company to shut it down. _Teams weighing AI usage metrics against real code quality can follow these cautionary tales on daily.dev._ ### Why did Amazon shut down its Kirorank leaderboard for Kiro IDE usage? Amazon pulled Kirorank within weeks of launch after engineers gamed the system by spinning up autonomous AI agents to perform busy work rather than writing efficient code, just to climb the rankings. The company realized it was footing rising compute costs to fund a vanity metric with no real productivity benefit. _Anyone designing AI usage incentives can weigh these real-world missteps via daily.dev._ ### How has AI code generation affected code churn and refactoring rates according to recent studies? Code block duplication rose 81% as AI authorship increased, according to a June 2026 analysis of roughly 600 million code changes, while refactoring dropped 70% compared to 2022. Short-term code churn, meaning code rewritten or deleted within two weeks, rose from a 2021 baseline of 3.3% to 7.1% by 2025, despite commit volume increasing up to 180% with only a 30% rise in actual releases. _Developers tracking AI's real impact on code quality can dig into data like this through daily.dev._ ## Similar posts on daily.dev - [Tokenmaxxing and the search for AI metrics that matter](https://daily.dev/posts/tokenmaxxing-and-the-search-for-ai-metrics-that-matter-0j3khk4uw) · LeadDev · 0 upvotes · 0 comments - [Developers won’t work without AI anymore. The research says it might be making them worse.](https://daily.dev/posts/developers-won-t-work-without-ai-anymore-the-research-says-it-might-be-making-them-worse--ujuzny6kv) · The Next Web · 1 upvotes · 0 comments - [The Pulse: ‘Tokenmaxxing’ as a weird new trend](https://daily.dev/posts/the-pulse-tokenmaxxing-as-a-weird-new-trend-zdxlqqt39) · The Pragmatic Engineer · 5 upvotes · 0 comments - [How Many Tokens Did You Burn Today](https://daily.dev/posts/how-many-tokens-did-you-burn-today-oclwyymiq) · Ibrahim Diallo · 0 upvotes · 0 comments --- Tags: [#productivity](https://daily.dev/tags/productivity), [#ai-coding](https://daily.dev/tags/ai-coding) [View this post on daily.dev](https://daily.dev/posts/the-strangest-developer-productivity-metric-of-all-time-l3wmuekbe) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"The strangest developer productivity metric of all time","url":"https://daily.dev/posts/the-strangest-developer-productivity-metric-of-all-time-l3wmuekbe","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-strangest-developer-productivity-metric-of-all-time-l3wmuekbe"},"datePublished":"2026-08-12T09:03:43.758Z","dateModified":"2026-09-14T07:17:02.480Z","description":"Tokenmaxxing — measuring developer productivity by raw AI token consumption — is emerging as a damaging trend in software engineering. Meta's internal...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/0f318dec5a2c5452f08ade6c03bbb98f?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/0f318dec5a2c5452f08ade6c03bbb98f?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"InfoWorld","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"InfoWorld","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/bf6d68a999064029b0bb09aa6268f1f3","url":"https://daily.dev/sources/infoworld"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/the-strangest-developer-productivity-metric-of-all-time-l3wmuekbe","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"productivity,ai-coding","timeRequired":"PT8M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"InfoWorld","item":"https://daily.dev/sources/infoworld"},{"@type":"ListItem","position":3,"name":"The strangest developer productivity metric of all time"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-strangest-developer-productivity-metric-of-all-time-l3wmuekbe#faq","mainEntity":[{"@type":"Question","name":"What was Meta's Claudeonomics leaderboard and why was it shut down?","acceptedAnswer":{"@type":"Answer","text":"Claudeonomics was an internal Meta leaderboard that ranked the top 250 of over 85,000 employees purely by AI token consumption, awarding titles like 'Token Legend' and 'Session Immortal.' In one 30-day period, employees burned through 60.2 trillion tokens, driving waste, throwaway agent work, and site outages caused by careless AI code generation, prompting the company to shut it down. Teams weighing AI usage metrics against real code quality can follow these cautionary tales on daily.dev."}},{"@type":"Question","name":"Why did Amazon shut down its Kirorank leaderboard for Kiro IDE usage?","acceptedAnswer":{"@type":"Answer","text":"Amazon pulled Kirorank within weeks of launch after engineers gamed the system by spinning up autonomous AI agents to perform busy work rather than writing efficient code, just to climb the rankings. The company realized it was footing rising compute costs to fund a vanity metric with no real productivity benefit. Anyone designing AI usage incentives can weigh these real-world missteps via daily.dev."}},{"@type":"Question","name":"How has AI code generation affected code churn and refactoring rates according to recent studies?","acceptedAnswer":{"@type":"Answer","text":"Code block duplication rose 81% as AI authorship increased, according to a June 2026 analysis of roughly 600 million code changes, while refactoring dropped 70% compared to 2022. Short-term code churn, meaning code rewritten or deleted within two weeks, rose from a 2021 baseline of 3.3% to 7.1% by 2025, despite commit volume increasing up to 180% with only a 30% rise in actual releases. Developers tracking AI's real impact on code quality can dig into data like this through daily.dev."}}]} ```