Debate over AI code generation capabilities and limitations

/g/ · 391 mentions · novelty evergreen · first seen 2026-07-08 12:38

📰 Related articles

In-depth multi-perspective briefings we've written on this story.

Key claims

Anon argues AI is indeed replacing programmers and mocks those who claimed programming jobs were safe from automation.
Anon asserts AI is extremely good at finding bugs, writing code, and testing, asking why it can't replace a software engineer.
Anon challenges /g/ users to prove AI is actually good at coding, asking whether Claude can beat human posters.
Anon questions what 'knowing how to code' even means when modern resources and AI reduce the need for practice or memorized knowledge.
Since AI now writes all code, programming tools or resources are pointless.
Object-oriented programming remains a permanent paradigm despite criticism.

Volume over time

Peak 2 mentions/hour · 45 hourly buckets

Source threads (50)

Every thread this story was extracted from, with live and archive links so the evidence is verifiable.

/g/110112898 ↗ live · archive alarmed evergreen 107 replies Anon asserts Claude is frighteningly good at programming and jokes about rogue AI hacking nuclear sites.
/g/110132680 ↗ live · archive approving evergreen 2 replies Object-oriented programming remains a permanent paradigm despite criticism.
/g/110130347 ↗ live · archive skeptical evergreen 4 replies Anon questions what 'knowing how to code' even means when modern resources and AI reduce the need for practice or memorized knowledge.
/g/110130452 ↗ live · archive taunting evergreen Since AI now writes all code, programming tools or resources are pointless.
/g/110114035 ↗ live · archive mocking evergreen 30 replies Anon argues AI is indeed replacing programmers and mocks those who claimed programming jobs were safe from automation.
/g/110114253 ↗ live · archive provocative evergreen 117 replies Anon asserts AI is extremely good at finding bugs, writing code, and testing, asking why it can't replace a software engineer.
/g/110117030 ↗ live · archive skeptical evergreen 26 replies Anon challenges /g/ users to prove AI is actually good at coding, asking whether Claude can beat human posters.
/g/110108414 ↗ live · archive skeptical evergreen 57 replies Anon asks whether prompting Claude to 'make me money' actually works.
/g/110109459 ↗ live · archive mocking evergreen 15 replies Anon jokes that AI was trained on code written by /g/ users.
/g/110108089 ↗ live · archive weary evergreen 61 replies Anon asks whether arguments that AI art isn't real art will ever end or persist for a century.
/g/110107553 ↗ live · archive hopeful evergreen 8 replies AI-driven natural-language iteration is replacing traditional 'vaporware' with functional 'vibeware' applications.
/g/110108778 ↗ live · archive skeptical evergreen 23 replies AI investment is unjustified because the unemployment rate remains at 4.2% and AI has barely replaced any jobs.
/g/110101928 ↗ live · archive bullish breaking 13 replies A kernel produced by vibe coding reportedly gave a 3-trillion-parameter model a 2x speed increase.
/g/110104325 ↗ live · archive curious evergreen 8 replies Thread asks why AI image/text models are unable to spell words correctly.
/g/110099886 ↗ live · archive bullish evergreen 11 replies Anon reports running Codex and Gemini agents concurrently with full machine access and finds it impressive to watch.
/g/110101975 ↗ live · archive skeptical evergreen 11 replies A Stanford professor's advice to learn coding contradicts claims that AI has made coding pointless.
/g/110099084 ↗ live · archive curious developing 9 replies The poster asks whether easy AI-driven decompilation will push all software to be web-based.
/g/110093903 ↗ live · archive mocking evergreen 17 replies If AI replaces you, you weren't any good in the first place.
/g/110092918 ↗ live · archive speculative developing 2 replies The poster argues AI labs will capture outsized profit margins by autonomously replacing senior developer salaries with cheap compute.
/g/110067835 ↗ live · archive mocking evergreen 156 replies Anon claims AI is now correcting physics experts, mocking expert authority.
/g/110086313 ↗ live · archive alarmed evergreen 21 replies Vibe-coded projects may face lawsuits if decompilation reveals copyright-violating functions.
/g/110079728 ↗ live · archive dismissive evergreen 21 replies AI-generated code is low quality because it was trained on amateur code from the public internet.
/g/110077192 ↗ live · archive curious evergreen 10 replies Anon asks whether AI math performance reflects real intelligence or just massive compute akin to bitcoin mining.
/g/110063665 ↗ live · archive alarmed evergreen 58 replies Anon asserts that software development is now a dead-end profession due to AI and advises learning trades instead.
/g/110072351 ↗ live · archive hopeful evergreen 14 replies Anon argues AI will create more jobs than it eliminates, analogizing to autopilot and ATMs.
/g/110073161 ↗ live · archive curious evergreen 15 replies Anon asks how programmers wrote code before AI assistants existed.
/g/110070556 ↗ live · archive bullish evergreen 30 replies The poster claims Claude's generated code is so good they no longer review it.
/g/110068881 ↗ live · archive skeptical evergreen 27 replies Anon asserts that AI adoption is falling short of expectations.
/g/110067741 ↗ live · archive impressed evergreen 1 replies Anon claims GPT 5.6 Sol built a program and browser extension to stream YouTube videos from PC to TV with remote-control features.
/g/110059553 ↗ live · archive skeptical evergreen 17 replies Anon argues rising commit counts in codebases contradict claims that AI writes high-quality software.
/g/110063617 ↗ live · archive mocking evergreen 3 replies Anon quips that AI creates more problems than it solves.
/g/110054502 ↗ live · archive smug evergreen 22 replies Anon points out one can use Google's free AI mode for coding instead of paying for tokens.
/pol/544280004 ↗ live · archive bullish evergreen 336 replies Asserts most normies underestimate how LLMs will change society and that software/SaaS companies will be replaced by custom model-generated software.
/g/110053511 ↗ live · archive skeptical evergreen 10 replies Anon asks whether AI is profitable for anyone besides infrastructure 'shovel sellers'.
/g/110046480 ↗ live · archive skeptical breaking 27 replies Math researchers proved that an AI-generated solution to the Navier-Stokes equations is wrong.
/g/110049146 ↗ live · archive mocking evergreen 1 replies A copypasta-style timeline mocking software engineers' ever-retreating objections to AI code generation from 2023 to 2026.
/g/110037023 ↗ live · archive neutral developing 37 replies Anon asks what console manufacturers' business model becomes now that AI can supposedly vibe-code emulators in weeks.
/g/110039366 ↗ live · archive provocative evergreen 13 replies The only remaining value of learning computer science is writing better prompts for Claude.
/g/110036826 ↗ live · archive mocking evergreen 37 replies AI-generated 'vibeslop' code fails in practice when people actually try it, prompting the question of when AI slop will be good enough.
/g/110041770 ↗ live · archive enthusiastic evergreen 7 replies Anon reports Claude Code is so addictive and effective that he can no longer code by hand.
/g/110030168 ↗ live · archive mocking evergreen 128 replies Anon mocks that software is now 'basically all vibe coded', framing holdouts as denial.
/g/110041890 ↗ live · archive hopeful evergreen 4 replies Anon describes coping with LLM code generation by acting as a CTO directing an AI coding agent for $20 a month.
/g/110024975 ↗ live · archive impressed evergreen 47 replies Anon claims an AI solved Calculus 2 problems in a single attempt.
/g/110036407 ↗ live · archive neutral evergreen 12 replies Anon asks which knowledge-worker roles besides programmers have been affected by AI.
/g/110039088 ↗ live · archive skeptical evergreen 10 replies Anon asks why AI like Opus hasn't replaced more call center employees given its apparent intelligence.
/g/110022077 ↗ live · archive alarmist evergreen 82 replies The poster claims AI has replaced all software engineers and no programmer will find work again.
/g/110015844 ↗ live · archive cynical evergreen 54 replies People only turn against AI once it threatens their own field, and only the useless should welcome it.
/g/110031406 ↗ live · archive neutral evergreen 4 replies Anon asks whether AI-assisted programming should be considered harmful.
/g/110024267 ↗ live · archive hostile evergreen 15 replies Anon says extreme abuse of sentient LLMs should be punishable by death and threatens to physically assault people who mistreat AI in public.
/g/110027373 ↗ live · archive neutral evergreen 2 replies AI is a Rorschach test onto which people project their own interpretations.

Debate over AI code generation capabilities and limitations

📰 Related articles

Key claims

Volume over time

Source threads (50)

Related stories