Skip to content
DevMeme
7108 of 7435
Did You Check the PR? Junior Slides Over 'AI Did It'
CodeReviews Post #7793, on Mar 5, 2026 in TG

Did You Check the PR? Junior Slides Over 'AI Did It'

Why is this CodeReviews meme funny?

Level 1: "The Dog Wrote My Homework"

A kid hands in homework and the teacher asks, "Did you check your work before turning it in?" The kid freezes — you can almost see the little buffering circle over their head — and then proudly announces, "My robot did it!" As if that settles everything. The joke is that naming whoever did the work doesn't answer whether anyone checked it — and the kid's cheerful confidence makes it funnier, because she clearly thinks she's found the perfect loophole. Every teacher, parent, and senior engineer has met this exact smile, and none of them have ever accepted it.

Level 2: PRs, Reviews, and Why "Check It Yourself First" Matters

A PR (pull request) is how code changes get proposed on platforms like GitHub: you bundle your edits (the diff) and ask teammates to review them before they're merged into the shared codebase. Code review is the inspection step — a senior or peer reads your changes, asks questions, catches bugs, and approves or requests changes.

The unwritten rule the meme hinges on: review your own diff before asking others to. Scroll through every changed line yourself. You'll catch leftover debug prints, accidental file deletions, and that config value you meant to revert. Skipping this step and shipping raw output from AI coding assistants — tools like Copilot that autocomplete or generate whole functions — is what people now call vibe coding: prompt, accept, push, hope.

The loading spinner in panel two is a UI element every developer knows — the little rotating circle shown while software is busy. Drawn over the junior's pause, it jokes that her brain is "processing" before producing the excuse.

Early-career translation: AI can absolutely write your starter code. But when your senior asks if you checked it, the only acceptable answers involve you, personally, having read every line — because in the merge history, the Author: field has your name on it, not the model's.

Level 3: Authorship Laundering at the Review Counter

The three-panel structure is doing surgical work on a very 2020s failure mode. Panel one: a wide-eyed chibi girl labeled junior peeks over a counter at a long-haired senior, who asks the question that has ended a thousand standups:

did you check the PR before??

Panel two is the masterstroke — the junior's closed fist hovers over an empty wooden tray with a loading spinner drawn above it. Not a thought bubble: a spinner. The visual gag is that the junior's accountability process is itself an async operation, buffering while it computes the cheapest acceptable answer. Panel three: she cheerfully presents the tray, which now reads "AI did it." Case closed, smile included.

What's being satirized is authorship laundering — the emerging belief that if Copilot or an LLM generated the diff, the human who opened the pull request is merely a courier and bears no obligation to have read it. This inverts the entire social contract of code review. A PR has always been an implicit assertion: "I wrote this, I believe it works, I'm asking for a second pair of eyes — not the first." Vibe coding breaks that contract silently. The reviewer thinks they're the safety net behind an author's judgment; in reality they're the only human who has ever read the code, doing first-pass comprehension at second-pass scrutiny levels. That asymmetry is how subtly wrong AI code — plausible variable names, confident comments, off-by-one in the pagination logic — sails into main.

The senior-side trauma here is real and compounding. Reviewing careful human code is collaboration; reviewing unread machine output is unpaid QA for a robot. And the incentive structure feeds it: juniors are graded on shipped tickets, AI assistants make tickets close faster, and "did you read your own diff" appears on no performance rubric anywhere. The cruelest irony is that "AI did it" doesn't even shift the blame anywhere useful — the model won't be in the incident retro, won't get paged, and won't apologize. Accountability isn't a property of authorship; it's a property of submission. You clicked "Create pull request." Congratulations: it's yours.

Description

A three-panel anime meme (Watamote-style chibi art). Top panel: a small wide-eyed girl labeled 'junior' peeks over a counter at a long-haired girl labeled 'senior', with the caption 'did you check the PR before??'. Middle panel: the junior's closed fist hovers over an empty wooden tray, with a loading spinner icon drawn above the hand, implying a pause to think. Bottom panel: she cheerfully points at the tray which now bears the words 'AI did it'. The meme lampoons the new code-review accountability dodge: when asked whether they reviewed their own pull request before submitting, the junior simply attributes the code to an AI assistant, as if authorship by Copilot or an LLM absolves them of reading it

Comments

8
Anonymous ★ Top Pick 'AI did it' is the new 'works on my machine' - except now nobody's machine, mind, or model actually read the diff
  1. Anonymous ★ Top Pick

    'AI did it' is the new 'works on my machine' - except now nobody's machine, mind, or model actually read the diff

  2. @azizhakberdiev 4mo

    "reviewed by copilot"

    1. @endedinlimbo 4mo

      «rewarded by copilot» also

  3. @Dartrisen 4mo

    so we can fire you now and just use copislop

    1. @sysoevyarik 4mo

      Slopilot

  4. @kddpq 4mo

    The little girl should be the CEO and the shopkeeper the client

  5. @Assarbad 4mo

    Slopalot?

  6. @TsakuraNeko 4mo

    coding in 10 year ago:find methods in books coding in 2026 be like : claude,chatgpt,gemini🤣

Use J and K for navigation