Guys Who Only Say 'Just Use Claude' Starter Pack

Level 1: The Friend Everyone Quotes

Imagine a group of friends who all have the same imaginary smart friend named Claude. Ask any of them anything — what game to play, how to fix a toy, whether it might rain — and every single one says some version of "let me ask Claude" or "well, Claude told me…". They all think they're being helpful in their own special way, but from the outside they're nine people in matching shirts saying the exact same thing. It's funny because nobody in the group answers a question themselves anymore — and the picture catches them all smiling about it.

Level 2: Meet the New Rubber Duck (It Talks Back)

Claude is Anthropic's AI assistant, one of the large language models (LLMs) that developers now use for writing code, reviewing pull requests, explaining errors, and drafting basically anything. Each speech bubble maps to a real workflow you'll encounter in your first year on a modern team:

• "Let me throw that into Claude" — pasting an error, a stack trace, or a teammate's code into the chat window for instant analysis. The 2026 equivalent of googling the error message.
• "Claude cooked" — slang meaning the AI produced something impressively good. ("Cooked" = excelled.)
• "Did you ask Claude?" — the new first-line support question, replacing "did you try turning it off and on again?"
• "Just use Claude" — the universal answer to "how do I…?" questions, helpful and slightly dismissive at the same time.

The healthy way to use this: treat the AI like an extremely well-read colleague who is confident even when wrong. It's brilliant for unblocking yourself, generating first drafts, and explaining unfamiliar code. The trap the meme warns about is letting "ask Claude" replace understanding — if you can't explain why the suggested fix works, you've shipped someone else's guess. The old practice of rubber-duck debugging (explaining your problem aloud to a toy duck until you spot the flaw yourself) worked because you did the reasoning. The duck now answers back, which is more useful and more dangerous.

Level 3: Consensus by Chat Window

Nine stock-photo men in coordinated blue shirts beam up at the camera, and around them orbit nine phrases that, taken together, form a complete transcript of every engineering Slack channel circa 2026: "Let's see what Claude thinks", "I asked Claude", "You should use Claude for that", "Claude cooked here", "Just use Claude", "Did you ask Claude?", "Claude told me", "Claude's actually pretty good at this", "Let me throw that into Claude."

The format matters. This is the venerable "guys who all say X" stock-photo meme, whose entire mechanism is uniformity wearing the costume of individuality — nine distinct faces, one personality. Applied to Claude evangelism, it lands a precise hit: the phrases sound like nine different conversational moves (a suggestion, a citation, a recommendation, praise, a challenge), but they're all the same move. Every one of them is delegating the thinking and routing the conversation through the same external oracle.

What the meme is really documenting is a shift in epistemic authority inside engineering teams. The classic appeal-to-authority moves — "the docs say", "Stack Overflow says", "the senior staff engineer says" — have been collapsed into a single citation: Claude told me. Notice the grammar of the bubbles, because it's doing satirical work. "Claude told me" is testimony. "Did you ask Claude?" is a process gate — the new "did you read the error message?". "Claude cooked here" is aesthetic judgment outsourced. And "Let's see what Claude thinks" is the truly loaded one: thinks. The anthropomorphization is complete; the LLM has a seat at the design review.

The senior-engineer discomfort this meme monetizes isn't that Claude is wrong — the bubble "Claude's actually pretty good at this" concedes, a bit sheepishly, that it usually isn't. The discomfort is about what happens to a team when the verification step quietly evaporates. An LLM citation has no accountability trail: when "Claude said so" replaces "here's my reasoning," code review degrades from argument into vibes arbitration between competing chat sessions. The failure mode isn't dramatic; it's the slow atrophy of the muscle that lets you evaluate an answer without asking the oracle again. Teams have always had a "just use Postgres" guy and a "just use Rust" guy; the novelty is that this archetype's tool claims to be good at everything, so the catchphrase fits every conversation. That's why there are nine of them in the photo. It's contagious by design.

The Telegram caption — "Chat, is it admin?" — twists the knife from the streaming-culture side: the habit of asking an external collective ("chat") to adjudicate reality is the same reflex, just with an LLM instead of an audience.