--- title: The LLM Critics Are Right. I Use LLMs Anyway. | daily.dev description: A developer reflects on the cognitive dissonance of agreeing with most LLM criticisms — environmental cost, copyright issues, slop generation, harm to junior... canonical: https://daily.dev/posts/the-llm-critics-are-right-i-use-llms-anyway--02xhbt7xl twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: The LLM Critics Are Right. I Use LLMs Anyway. | daily.dev og:description: A developer reflects on the cognitive dissonance of agreeing with most LLM criticisms — environmental cost, copyright issues, slop generation, harm to junior... og:url: https://daily.dev/posts/the-llm-critics-are-right-i-use-llms-anyway--02xhbt7xl og:image: https://api.daily.dev/og/posts/02XHbt7xL.png og:image:alt: The LLM Critics Are Right. I Use LLMs Anyway. 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 LLM Critics Are Right. I Use LLMs Anyway. **[Hacker News](https://daily.dev/sources/hn)** · 19 min read · 4 upvotes · 1 comments ## Summary A developer reflects on the cognitive dissonance of agreeing with most LLM criticisms — environmental cost, copyright issues, slop generation, harm to junior developers, geopolitical risk — while still spending ~$10k/month on tokens. The post explores how LLMs can amplify existing human thinking rather than replace it, and shares concrete patterns: the 'grill-me' technique for forcing structured thinking, writing tight three-sentence problem statements before coding, using subagents to adversarially critique work, and leveraging hallucination as a UX design signal. The key thesis is that LLMs amplify what you already have — good thinking becomes sharper, but empty thinking produces fluent slop — and that trust and credibility remain the only reliable signals of quality in an LLM-saturated world. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is the 'grill me' skill for using LLMs before coding? The grill-me skill is a prompt technique, adapted from Matt Pocock's 'grill me' method, that instructs an LLM to interview you relentlessly about every aspect of a task, walking down each branch of a decision tree one question at a time, before acting. It asks the LLM to look up facts itself but require you to make actual decisions, forcing shared understanding before any code is written. _Developers refining AI-assisted coding habits can find workflow techniques like this through daily.dev._ ### Why are open source projects like Zig and Gentoo rejecting AI generated pull requests? Zig and Gentoo have adopted policies refusing LLM-generated pull requests because maintainers can no longer tell whether a contributor spent real effort on a submission or simply let an AI agent generate it unsupervised. Before LLMs, writing a proper PR required enough human time to naturally filter out low-effort or bad-faith contributions; that filter no longer holds, threatening trust in open source review processes. _Maintainers weighing AI contribution policies can track how projects like Zig and Gentoo handle it via daily.dev._ ### How did Armin Ronacher's coding agent project Pi handle a flood of AI-generated pull requests? Armin Ronacher, creator of Flask and an early Sentry team member, said his company Earendil auto-closes almost all pull requests and issues submitted to Pi, an open-source coding agent harness, because the volume generated by LLMs became unmanageable. He noted contributors shouldn't be discouraged from opening PRs anyway, since genuine human effort still stands out. _Anyone building or maintaining AI coding tools can follow real-world moderation approaches like this on daily.dev._ ## Community discussion Top comments from developers on daily.dev. **@kartiknvj** · 0 upvotes > The point about using subagents to adversarially critique work maps closely to how I run LLM-as-judge in my own pipelines, since a second model grading against an explicit rubric catches far more than a single pass ever does. I have also started treating hallucination as a signal rather than pure noise, usually it means my context was underspecified. Do you find the 'grill-me' technique holds up on longer multi-step agent tasks, or does the critique drift once the context window fills? ## Similar posts on daily.dev - [Why LLMs Rock \(or: BooLLMean Logic for the Uninitiated\)](https://daily.dev/posts/why-llms-rock-or-boollmean-logic-for-the-uninitiated--66elcilfw) · Beam Bloggers Webring · 0 upvotes · 0 comments - [Why talking to LLMs has improved my thinking – Vallified](https://daily.dev/posts/why-talking-to-llms-has-improved-my-thinking-vallified-nagkqxprh) · Hacker News · 1 upvotes · 0 comments - [The analogue programmer in the land of LLMs](https://daily.dev/posts/the-analogue-programmer-in-the-land-of-llms-fupyefolu) · Ruby Flow · 2 upvotes · 1 comments --- Tags: [#open-source](https://daily.dev/tags/open-source), [#llm](https://daily.dev/tags/llm), [#ai-coding](https://daily.dev/tags/ai-coding), [#claude-code](https://daily.dev/tags/claude-code) [View this post on daily.dev](https://daily.dev/posts/the-llm-critics-are-right-i-use-llms-anyway--02xhbt7xl) ```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 LLM Critics Are Right. I Use LLMs Anyway.","url":"https://daily.dev/posts/the-llm-critics-are-right-i-use-llms-anyway--02xhbt7xl","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-llm-critics-are-right-i-use-llms-anyway--02xhbt7xl"},"datePublished":"2026-07-16T12:35:50.592Z","dateModified":"2026-09-14T08:28:49.807Z","description":"A developer reflects on the cognitive dissonance of agreeing with most LLM criticisms — environmental cost, copyright issues, slop generation, harm to junior...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/06ef1ac4bc847252bb5594b4fcaa16f8?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/06ef1ac4bc847252bb5594b4fcaa16f8?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Hacker News","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":"Hacker News","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/hn","url":"https://daily.dev/sources/hn"},"commentCount":1,"discussionUrl":"https://daily.dev/posts/the-llm-critics-are-right-i-use-llms-anyway--02xhbt7xl","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":4},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":1}],"keywords":"open-source,llm,ai-coding,claude-code","timeRequired":"PT19M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Hacker News","item":"https://daily.dev/sources/hn"},{"@type":"ListItem","position":3,"name":"The LLM Critics Are Right. I Use LLMs Anyway."}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/the-llm-critics-are-right-i-use-llms-anyway--02xhbt7xl","comment":[{"@type":"Comment","text":"The point about using subagents to adversarially critique work maps closely to how I run LLM-as-judge in my own pipelines, since a second model grading against an explicit rubric catches far more than a single pass ever does. I have also started treating hallucination as a signal rather than pure noise, usually it means my context was underspecified. Do you find the ‘grill-me’ technique holds up on longer multi-step agent tasks, or does the critique drift once the context window fills?","datePublished":"2026-07-16T13:40:18.397Z","url":"https://daily.dev/posts/02XHbt7xL#c-BnxoaQS1R","author":{"@type":"Person","name":"kartik-nvjk","url":"https://daily.dev/kartiknvj","image":"https://media.daily.dev/image/upload/s--3gGgsVCw--/f_auto/v1781456774/avatars/avatar_TvTVeiMdkRCqWUDullFmy?_a=BAMAMiWQ0"}}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-llm-critics-are-right-i-use-llms-anyway--02xhbt7xl#faq","mainEntity":[{"@type":"Question","name":"What is the 'grill me' skill for using LLMs before coding?","acceptedAnswer":{"@type":"Answer","text":"The grill-me skill is a prompt technique, adapted from Matt Pocock's 'grill me' method, that instructs an LLM to interview you relentlessly about every aspect of a task, walking down each branch of a decision tree one question at a time, before acting. It asks the LLM to look up facts itself but require you to make actual decisions, forcing shared understanding before any code is written. Developers refining AI-assisted coding habits can find workflow techniques like this through daily.dev."}},{"@type":"Question","name":"Why are open source projects like Zig and Gentoo rejecting AI generated pull requests?","acceptedAnswer":{"@type":"Answer","text":"Zig and Gentoo have adopted policies refusing LLM-generated pull requests because maintainers can no longer tell whether a contributor spent real effort on a submission or simply let an AI agent generate it unsupervised. Before LLMs, writing a proper PR required enough human time to naturally filter out low-effort or bad-faith contributions; that filter no longer holds, threatening trust in open source review processes. Maintainers weighing AI contribution policies can track how projects like Zig and Gentoo handle it via daily.dev."}},{"@type":"Question","name":"How did Armin Ronacher's coding agent project Pi handle a flood of AI-generated pull requests?","acceptedAnswer":{"@type":"Answer","text":"Armin Ronacher, creator of Flask and an early Sentry team member, said his company Earendil auto-closes almost all pull requests and issues submitted to Pi, an open-source coding agent harness, because the volume generated by LLMs became unmanageable. He noted contributors shouldn't be discouraged from opening PRs anyway, since genuine human effort still stands out. Anyone building or maintaining AI coding tools can follow real-world moderation approaches like this on daily.dev."}}]} ```