--- title: Earning taste and judgment | daily.dev description: AI agents are automating the repetitive coding tasks that traditionally built junior developers' judgment and taste. Data shows recent CS graduates face 6-7.5%... canonical: https://daily.dev/posts/earning-taste-and-judgment-rngvm5y0t twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Earning taste and judgment | daily.dev og:description: AI agents are automating the repetitive coding tasks that traditionally built junior developers' judgment and taste. Data shows recent CS graduates face 6-7.5%... og:url: https://daily.dev/posts/earning-taste-and-judgment-rngvm5y0t og:image: https://api.daily.dev/og/posts/rNgvm5y0t.png og:image:alt: Earning taste and judgment 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. # Earning taste and judgment **[Addy Osmani](https://daily.dev/sources/addy)** · [@addyosmani](https://daily.dev/addyosmani) · 11 min read · 0 upvotes · 0 comments ## Summary AI agents are automating the repetitive coding tasks that traditionally built junior developers' judgment and taste. Data shows recent CS graduates face 6-7.5% unemployment and junior tech job postings have dropped 34% since 2020, while senior roles grow. The core argument: taste and judgment — knowing what to build and whether it's good — are the durable, ungradeable skills AI cannot replicate. Seven concrete practices are offered for deliberately building these skills: reading more code than you generate, keeping a 'wrong log' of agent mistakes, doing some things the hard way on purpose, going deep on one system end-to-end, learning to specify and verify separately, building an eval framework for AI-generated PRs, and calibrating autonomy per task. The broader career advice centers on finishing the last mile, solving hard problems, building in public, and becoming a T-shaped generalist. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### Why is the entry-level software developer job market shrinking even as overall tech job growth stays positive? Junior and standard tech job postings fell 34% since early 2020 compared to just 19% for senior and manager roles, while postings requiring five or more years of experience rose from 37% to 42%, according to Indeed Hiring Lab data. Meanwhile recent computer-science and computer-engineering graduates faced unemployment rates of 6.1% and 7.5% respectively, both above many non-technical majors, per Forbes citing the St. Louis Federal Reserve. _Track how AI is reshaping entry-level hiring and what it means for your career path on daily.dev._ ### What concrete practices can a junior developer use to build judgment when AI agents handle most of the coding reps? Seven practices include reading far more code than you generate, keeping a one-sentence 'wrong log' of every agent mistake to spot patterns after thirty days, building something meaningful entirely by hand to protect collateral learning, going deep on one system end to end until failure, separating specification from verification, building an eval framework scored against fifty real AI-generated PRs, and calibrating autonomy per task based on cost and reversibility of failure. _Developers rethinking how to build real skill alongside coding agents can follow this practice on daily.dev._ ### How much do people wrongly trust incorrect AI output, and what does that do to their own accuracy? A study surveying 1,372 participants across roughly 10,000 trials found people accepted incorrect AI output nearly 80% of the time, with accuracy dropping about 15 points below the no-AI baseline even as confidence rose about 12%, according to research by Shaw and Nave discussed on the Wharton Knowledge podcast. _Anyone weighing how much to trust AI-generated code can follow research like this on daily.dev._ ## Similar posts on daily.dev - [Above the API: What Developers Contribute When AI Can Code](https://daily.dev/posts/above-the-api-what-developers-contribute-when-ai-can-code-5fi2t6rld) · DEV · 54 upvotes · 2 comments - [Microsoft execs warn agentic AI is hollowing out the junior developer pipeline](https://daily.dev/posts/microsoft-execs-warn-agentic-ai-is-hollowing-out-the-junior-developer-pipeline-zfchuraa9) · The New Stack · 1 upvotes · 0 comments - [How to Get Hired in the AI Era](https://daily.dev/posts/how-to-get-hired-in-the-ai-era-s6jyddas8) · Towards Data Science · 90 upvotes · 5 comments - [CS Grads Now Have Higher Unemployment Than Art Majors](https://daily.dev/posts/cs-grads-now-have-higher-unemployment-than-art-majors-6mrblnc8g) · Medium · 0 upvotes · 0 comments --- Tags: [#career](https://daily.dev/tags/career), [#productivity](https://daily.dev/tags/productivity), [#ai-coding](https://daily.dev/tags/ai-coding) [View this post on daily.dev](https://daily.dev/posts/earning-taste-and-judgment-rngvm5y0t) ```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":"Earning taste and judgment","url":"https://daily.dev/posts/earning-taste-and-judgment-rngvm5y0t","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/earning-taste-and-judgment-rngvm5y0t"},"datePublished":"2026-07-16T06:16:06.249Z","dateModified":"2026-09-13T20:00:16.758Z","description":"AI agents are automating the repetitive coding tasks that traditionally built junior developers' judgment and taste. Data shows recent CS graduates face 6-7.5%...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/61018a50ac4d776bff7ee7dd170aef2f?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/61018a50ac4d776bff7ee7dd170aef2f?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Addy Osmani","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":"Addy Osmani","url":"https://daily.dev/addyosmani","image":"https://avatars2.githubusercontent.com/u/110953?v=4","interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"EndorseAction"},"userInteractionCount":5510}},"commentCount":0,"discussionUrl":"https://daily.dev/posts/earning-taste-and-judgment-rngvm5y0t","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"career,productivity,ai-coding","timeRequired":"PT11M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Addy Osmani","item":"https://daily.dev/sources/addy"},{"@type":"ListItem","position":3,"name":"Earning taste and judgment"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/earning-taste-and-judgment-rngvm5y0t#faq","mainEntity":[{"@type":"Question","name":"Why is the entry-level software developer job market shrinking even as overall tech job growth stays positive?","acceptedAnswer":{"@type":"Answer","text":"Junior and standard tech job postings fell 34% since early 2020 compared to just 19% for senior and manager roles, while postings requiring five or more years of experience rose from 37% to 42%, according to Indeed Hiring Lab data. Meanwhile recent computer-science and computer-engineering graduates faced unemployment rates of 6.1% and 7.5% respectively, both above many non-technical majors, per Forbes citing the St. Louis Federal Reserve. Track how AI is reshaping entry-level hiring and what it means for your career path on daily.dev."}},{"@type":"Question","name":"What concrete practices can a junior developer use to build judgment when AI agents handle most of the coding reps?","acceptedAnswer":{"@type":"Answer","text":"Seven practices include reading far more code than you generate, keeping a one-sentence 'wrong log' of every agent mistake to spot patterns after thirty days, building something meaningful entirely by hand to protect collateral learning, going deep on one system end to end until failure, separating specification from verification, building an eval framework scored against fifty real AI-generated PRs, and calibrating autonomy per task based on cost and reversibility of failure. Developers rethinking how to build real skill alongside coding agents can follow this practice on daily.dev."}},{"@type":"Question","name":"How much do people wrongly trust incorrect AI output, and what does that do to their own accuracy?","acceptedAnswer":{"@type":"Answer","text":"A study surveying 1,372 participants across roughly 10,000 trials found people accepted incorrect AI output nearly 80% of the time, with accuracy dropping about 15 points below the no-AI baseline even as confidence rose about 12%, according to research by Shaw and Nave discussed on the Wharton Knowledge podcast. Anyone weighing how much to trust AI-generated code can follow research like this on daily.dev."}}]} ```