(news.css@g873MbjYKKvPFKy21lu5.css) (y18.svg) (item@id=49420048.html) You Don't Need AI to Generate Code | Hacker News  (https://news.ycombinator.com)   (https://news.ycombinator.com/news) Hacker News  (https://news.ycombinator.com/newest) new | (https://news.ycombinator.com/front) past | (https://news.ycombinator.com/newcomments) comments | (https://news.ycombinator.com/ask) ask | (https://news.ycombinator.com/show) show | (https://news.ycombinator.com/jobs) jobs | (https://news.ycombinator.com/submit) submit   (https://news.ycombinator.com/login?goto=item%3Fid%3D49420048) login          (https://news.ycombinator.com/vote?id=49420048&how=up&goto=item%3Fid%3D49420048) (upvote)     (item@id=49420048.html) You Don't Need AI to Generate Code     4 points by (https://news.ycombinator.com/user?id=Lozybug) Lozybug (2026-08-24T14:10:19 1787580619) (item@id=49420048.html) 1 day ago   | (https://news.ycombinator.com/hide?id=49420048&goto=item%3Fid%3D49420048) hide | (https://hn.algolia.com/?query=You%20Don%27t%20Need%20AI%20to%20Generate%20Code&type=story&dateRange=all&sort=byDate&storyText=false&prefix&page=0) past | (https://news.ycombinator.com/fave?id=49420048&auth=dfd3a4af67dc850dc6c0c48573f6d2016b372b15) favorite | (item@id=49420048.html) 6 comments     Hi Guys, I'm going to make a heretical statement. It's this: you don't need AI to generate code and you shouldn't use AI to generate code.Instead I believe AI should be used to elicit the user requirements and the AI can then be used to generate a specification from which code can be mapped to produce code deterministically. We were doing this in a simplified form in the 1980s. If you think this is relevant as an alternative to vibe coding and agentic engineering. Let me know I can show you how all of this works and why it is important in 2026 and beyond. Yes, I am actually that old, but I have a view on the fundamental basics of software engineering. I feel like a survivor of a lost civilisation that has been washed away and forgotten. My approach brings together formal methods and AI in a way that was impossible back in the day. But... now that we have the compute and proper AI, not just IKBS, we can focus on generating a specification and not the code. A specification that can persist across the entire SDLC. All a bit like all the other branches of engineering that have drawings, specifications and circuit diagrams and things. Cheers me dears, Loz     (49420048) (item?id=49420048) (7da7796b816c38c6985dda455a54454470fafcc5)  (https://news.ycombinator.com/formatdoc) help  (add comment)      (https://news.ycombinator.com/vote?id=49438100&how=up&goto=item%3Fid%3D49420048) (upvote)     (https://news.ycombinator.com/user?id=aristofun) aristofun (2026-08-25T18:06:08 1787681168) (https://news.ycombinator.com/item?id=49438100) 1 hour ago   | (item@id=49420048.html#49437973) next [–]     > mapped to produce code deterministicallythat means your "spec" is essentially a code and the point you're trying to make is empty (https://news.ycombinator.com/reply?id=49438100&goto=item%3Fid%3D49420048%2349438100) reply           (https://news.ycombinator.com/vote?id=49437973&how=up&goto=item%3Fid%3D49420048) (upvote)     (https://news.ycombinator.com/user?id=mirmor23) mirmor23 (2026-08-25T17:57:01 1787680621) (https://news.ycombinator.com/item?id=49437973) 2 hours ago   | (item@id=49420048.html#49438100) prev | (item@id=49420048.html#49424604) next [–]     > now that we have the compute and proper AI, not just IKBS, we can focus on generating a specification and not the code. A specification that can persist across the entire SDLC. All a bit like all the other branches of engineering that have drawings, specifications and circuit diagrams and things.I'm not clear what this is supposed to achieve; if engineers can use AI to generate good specs, what's the issue with using llm in tdd/codegen cycle.currently, for complex projects with big teams the strategy seems to let llm loose and shoot for impossible deadlines which will likely backfire as llm could generate good code for a small subsystem, it fails frequently in brownfield, large projects resulting in engineering rework;by considering human effort to code in the release schedule, and then using llm to generate/refactor/tdd a good quality software could be built, without frustrating engineers having to deal with ai slop. (https://news.ycombinator.com/reply?id=49437973&goto=item%3Fid%3D49420048%2349437973) reply           (https://news.ycombinator.com/vote?id=49424604&how=up&goto=item%3Fid%3D49420048) (upvote)     (https://news.ycombinator.com/user?id=fedupofslop) fedupofslop (2026-08-24T19:19:03 1787599143) (https://news.ycombinator.com/item?id=49424604) 1 day ago   | (item@id=49420048.html#49437973) prev | (item@id=49420048.html#49420238) next [–]     Part of me really wants to agree, part of me just thinks it's too late to go back to the old ways and some slop will always exist. Do you have any good ways to evaluate AI Slop? Made a fun project (could call it a campaign) with a similar PoV, I'm spiritually old, check it out - dontshipslop.com, and let me know if you have any ideas on how to build a slop detector that evaluates things from a deterministic spec to code program synthesis approach... (https://news.ycombinator.com/reply?id=49424604&goto=item%3Fid%3D49420048%2349424604) reply           (https://news.ycombinator.com/vote?id=49420238&how=up&goto=item%3Fid%3D49420048) (upvote)     (https://news.ycombinator.com/user?id=wseqyrku) wseqyrku (2026-08-24T14:23:15 1787581395) (https://news.ycombinator.com/item?id=49420238) 1 day ago   | (item@id=49420048.html#49424604) prev | (item@id=49420048.html#49424060) next [–]     It's not even about generating code now. Good luck getting "agentic programming" out of people's head. (https://news.ycombinator.com/reply?id=49420238&goto=item%3Fid%3D49420048%2349420238) reply           (https://news.ycombinator.com/vote?id=49420504&how=up&goto=item%3Fid%3D49420048) (upvote)     (https://news.ycombinator.com/user?id=Lozybug) Lozybug (2026-08-24T14:42:28 1787582548) (https://news.ycombinator.com/item?id=49420504) 1 day ago   | (item@id=49420048.html#49420238) parent | (item@id=49420048.html#49424060) next [–]     Hi Wseqyrku, I'm curious. What is it about? Genuinely interested in your view. Cheers Loz (https://news.ycombinator.com/reply?id=49420504&goto=item%3Fid%3D49420048%2349420504) reply           (https://news.ycombinator.com/vote?id=49424060&how=up&goto=item%3Fid%3D49420048) (upvote)     (https://news.ycombinator.com/user?id=colesantiago) colesantiago (2026-08-24T18:35:12 1787596512) (https://news.ycombinator.com/item?id=49424060) 1 day ago   | (item@id=49420048.html#49420238) prev | (item@id=49420048.html#49433315) next [–]     great ted talk! (https://news.ycombinator.com/reply?id=49424060&goto=item%3Fid%3D49420048%2349424060) reply                (https://news.ycombinator.com/newsguidelines.html) Guidelines | (https://news.ycombinator.com/newsfaq.html) FAQ | (https://news.ycombinator.com/lists) Lists | (https://github.com/HackerNews/API) API | (https://news.ycombinator.com/security.html) Security | (https://www.ycombinator.com/legal/) Legal | (https://www.ycombinator.com/apply/) Apply to YC | (mailto:hn@ycombinator.com) Contact  Search: