--- title: AI coding got faster. Why didn’t engineering? | daily.dev description: Despite AI investment increasing 28x for most companies, engineering velocity metrics are stagnant or declining. The DX State of AI Impact in Engineering... canonical: https://daily.dev/posts/ai-coding-got-faster-why-didn-t-engineering--epmtgpec2 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: AI coding got faster. Why didn’t engineering? | daily.dev og:description: Despite AI investment increasing 28x for most companies, engineering velocity metrics are stagnant or declining. The DX State of AI Impact in Engineering... og:url: https://daily.dev/posts/ai-coding-got-faster-why-didn-t-engineering--epmtgpec2 og:image: https://api.daily.dev/og/posts/EpMtgPeC2.png og:image:alt: AI coding got faster. Why didn’t engineering? 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. # AI coding got faster. Why didn’t engineering? **[The New Stack](https://daily.dev/sources/newstack)** · [@jkriggins](https://daily.dev/jkriggins) · 8 min read · 0 upvotes · 0 comments ## Summary Despite AI investment increasing 28x for most companies, engineering velocity metrics are stagnant or declining. The DX State of AI Impact in Engineering report reveals that while AI helps individual developers understand and modify code faster, change confidence has dropped into negative territory — engineers are less confident releasing code to production. The Developer Experience Index (DXI) fell by two points industrywide, the first-ever decline, with each point representing ten hours per engineer per year. PR sizes have grown from a median of 42 lines in July 2025 to 72 lines a year later, with LinearB data showing elite AI adopters keep PRs under 100 lines while laggards exceed 228. Only 31% of teams are measuring AI's impact at all, and of those reporting productivity gains above 25%, most are going on instinct rather than confirmed data. Larger organizations face structural barriers — legacy processes designed when code-writing was expensive — that prevent them from realizing AI's benefits, while small teams can eliminate coordination overhead entirely. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How much did median pull request size grow after teams adopted AI coding tools? Median pull request size grew from 42 lines of code in July of one year to 72 lines of code a year later, according to DX's State of AI Impact in Engineering report. Larger PRs correlate with reduced incremental delivery and lower change confidence, undermining the productivity gains AI tools provide to individual developers. _Track how PR size and review overhead shift as your team leans on AI coding tools with daily.dev._ ### What percentage of engineering teams actually measure AI's impact on productivity? Only 31% of teams measure AI's impact at all, per LeadDev's AI Impact Report 2026. Separately, 70% describe AI adoption as wide or complete, and only 26% report productivity gains above 25%, though that figure includes people going on instinct rather than confirmed data, meaning most who report gains cannot prove them. _Compare real measurement practices before trusting anecdotal AI productivity claims, a recurring daily.dev discussion topic._ ### Why hasn't AI coding assistance improved overall engineering velocity despite a 28x increase in AI spending? Velocity has stagnated or declined because organizational coordination costs, legacy processes, and larger batch sizes offset individual speed gains from AI tools. DX's Developer Experience Index dropped industry-wide for the first time, driven by falling change confidence even as code maintainability improved, meaning engineers understand code better but trust releasing it less. _Engineers weighing AI tool ROI against real velocity data discuss findings like these on daily.dev._ ## Similar posts on daily.dev - [AI in engineering: Q2 2026 benchmarks & research readout](https://daily.dev/posts/ai-in-engineering-q2-2026-benchmarks-research-readout-a5bsn2lal) · Engineering Enablement · 10 upvotes · 0 comments - [The current impact of AI on engineering velocity](https://daily.dev/posts/the-current-impact-of-ai-on-engineering-velocity-lic0nalsa) · Software Engineering Research · 0 upvotes · 0 comments - [The State of AI Impact in Engineering: Q2 2026](https://daily.dev/posts/the-state-of-ai-impact-in-engineering-q2-2026-ynv5kqw0q) · Engineering Enablement · 5 upvotes · 1 comments - [AI productivity gains: More modest than expected](https://daily.dev/posts/ai-productivity-gains-more-modest-than-expected-iqt7heemt) · Engineering Enablement · 7 upvotes · 0 comments --- Tags: [#productivity](https://daily.dev/tags/productivity), [#ai-coding](https://daily.dev/tags/ai-coding), [#dora-metrics](https://daily.dev/tags/dora-metrics) [View this post on daily.dev](https://daily.dev/posts/ai-coding-got-faster-why-didn-t-engineering--epmtgpec2) ```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":"AI coding got faster. Why didn’t engineering?","url":"https://daily.dev/posts/ai-coding-got-faster-why-didn-t-engineering--epmtgpec2","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/ai-coding-got-faster-why-didn-t-engineering--epmtgpec2"},"datePublished":"2026-08-09T14:01:36.200Z","dateModified":"2026-09-14T09:00:41.129Z","description":"Despite AI investment increasing 28x for most companies, engineering velocity metrics are stagnant or declining. The DX State of AI Impact in Engineering...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/43149c3a7483ba63cc81632a1a0159ff?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/43149c3a7483ba63cc81632a1a0159ff?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"The New Stack","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":"Person","name":"Jennifer Riggins","url":"https://daily.dev/jkriggins","image":"https://lh3.googleusercontent.com/a-/AOh14Gi73gNjeWuGfr5F27hqtAdGKRmfEEUKNGT-qRG8N6g=s100","interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"EndorseAction"},"userInteractionCount":80}},"commentCount":0,"discussionUrl":"https://daily.dev/posts/ai-coding-got-faster-why-didn-t-engineering--epmtgpec2","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"productivity,ai-coding,dora-metrics","timeRequired":"PT8M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"The New Stack","item":"https://daily.dev/sources/newstack"},{"@type":"ListItem","position":3,"name":"AI coding got faster. Why didn’t engineering?"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/ai-coding-got-faster-why-didn-t-engineering--epmtgpec2#faq","mainEntity":[{"@type":"Question","name":"How much did median pull request size grow after teams adopted AI coding tools?","acceptedAnswer":{"@type":"Answer","text":"Median pull request size grew from 42 lines of code in July of one year to 72 lines of code a year later, according to DX's State of AI Impact in Engineering report. Larger PRs correlate with reduced incremental delivery and lower change confidence, undermining the productivity gains AI tools provide to individual developers. Track how PR size and review overhead shift as your team leans on AI coding tools with daily.dev."}},{"@type":"Question","name":"What percentage of engineering teams actually measure AI's impact on productivity?","acceptedAnswer":{"@type":"Answer","text":"Only 31% of teams measure AI's impact at all, per LeadDev's AI Impact Report 2026. Separately, 70% describe AI adoption as wide or complete, and only 26% report productivity gains above 25%, though that figure includes people going on instinct rather than confirmed data, meaning most who report gains cannot prove them. Compare real measurement practices before trusting anecdotal AI productivity claims, a recurring daily.dev discussion topic."}},{"@type":"Question","name":"Why hasn't AI coding assistance improved overall engineering velocity despite a 28x increase in AI spending?","acceptedAnswer":{"@type":"Answer","text":"Velocity has stagnated or declined because organizational coordination costs, legacy processes, and larger batch sizes offset individual speed gains from AI tools. DX's Developer Experience Index dropped industry-wide for the first time, driven by falling change confidence even as code maintainability improved, meaning engineers understand code better but trust releasing it less. Engineers weighing AI tool ROI against real velocity data discuss findings like these on daily.dev."}}]} ```