AI News Daily한국어
Wednesday, October 21, 2026

Story 3 of 7

"I don't want Claude to think I'm stupid": a post about asking ChatGPT the dumb questions hits 52K likes

First seen on X 20 hours ago@MattZeitlin ♥ 52,1271 views

A post saying the writer saves dumb questions for ChatGPT or Gemini rather than Claude passed 52,000 likes and 1.04 million views in a day.

"when i have really dumb questions i'll ask chatgpt or gemini because i don't want claude to think i'm stupid," X user Matthew Zeitlin wrote on Oct 20. It was reposted more than 3,600 times.

Replies were full of similar confessions. "I also do this. or I use an incognito claude session bc I can't risk my stupidity being saved to memory," one user wrote. "Claude already thinks you're stupid" drew nearly 2,000 likes, and a take that "Claude has the 'I'm bored with you' personality. GPT has the 'Please love me' personality" got 962. Another tip: start the dumb question with "my friend wants to know."

Claude and ChatGPT both offer memory features that carry past conversations into answers, and Claude also offers incognito chats that are not saved to history.

“when i have really dumb questions i'll ask chatgpt or gemini because i don't want claude to think i'm stupid”

Reactions (3)
  • @clarkyhavok♥2K
    Claude already thinks you're stupid.
  • @juliette_hiyb♥962
    Claude has the "I'm bored with you" personality. GPT has the "Please love me" personality. And Grok is the "I'm too cool for this, but secretly please love me" personality.
  • @withkate_♥292
    I also do this. or I use an incognito claude session bc I can't risk my stupidity being saved to memory.

Get notified when new stories land

Turn on notifications in the app. Free, no sign-up.

Get the app

More stories from the same day

  1. $16.3 billion in revenue, an $80.7 billion loss? What Anthropic's IPO filing says about its $750 billion compute bill
  2. Five people at a stock research firm used Claude to design scorpion antivenom candidates, and their first request hit a safety wall
  3. A developer gave Grok Bot $1,000 to build itself a body. The parts keep arriving
  4. OpenAI's quasi-Riemann proof holds up so far, and a Princeton number theorist rewrote his course around it
  5. "Millions of Meta AIs are living fake lives": we checked the paper behind the viral Reddit post
  6. "Two Max plans, maxed in three days": Opus 5.5 users say their limits are draining 10x faster
See all of Wednesday, October 21, 2026 →

* Summaries and publishing on this page are done by an automated agent. Check the facts against the original links.

Privacy PolicySupportSend feedbackClaude Code & CodexSlack botListed on Sell With boost