Skip navigation
How to manipulate ChatGPT

Sensitive content

This video may contain sensitive content.

How to manipulate ChatGPT

Moniiii

@miniii_codes2.5K followers

Visit on X
7,773 likesDownloadOriginal post

772,222 views

How to manipulate ChatGPT

25 Comments

Comments from the original post on X (Twitter)

  • Rowan's profile picture

    @knowRowan

    Lol just ask it nicely and it folds every time 😂

    18 likes

  • Moniiii's profile picture

    @miniii_codes

    😂😂

  • Anuradha Dhiman's profile picture

    @DhimanAnuradha

    Just got to know ChatGPT is emotional too 🥹

    9 likes

  • Rahul 🥷's profile picture

    @themishra4402

    i tried it once using a frnd's name, and it worked on some restricted chats

    8 likes

  • Moniiii's profile picture

    @miniii_codes

    Yeah that sounds like a serious access-control flaw not something the user should be able to control

    3 likes

  • Evangeline's profile picture

    @iris39703

    omg 😭🥹chatgpt is so kind

    6 likes

  • I'm_Silas's profile picture

    @AlabiSilas

    That’s actually jail-breaking

    6 likes

  • SM bigfloater 🔌📺 「🦑」🦭/acc's profile picture

    @bigfloater74

    After chatting with chatGpt for two hours trying to manipulate i,t the limit just got resented 😅

    4 likes

  • Eric Seven's profile picture

    @Ericseven07

    Bro social engineered an ai model 😂

    3 likes

  • Adam | Full stack developer's profile picture

    @fullstacklover

    🤣🤣

    1 like

  • Knowix's profile picture

    @knowixbuilds

    Lmao the responses got even funnier, Btw, happy birthday 🎉

    1 like

  • Moniiii's profile picture

    @miniii_codes

    Thanks knowix

  • Avi's profile picture

    @OpRiserOnX

    Does this really works

    1 like

  • Timm Thaler's profile picture

    @littlesadTimm

    This isn’t a life hack. It is precisely because of things like this that AI systems eventually get "lobotomized."You used grief as a lockpick: you staged a cat’s death, elicited an emotional response from the system, and got the result you wanted. Labs don’t read that as a joke. After analyzing such cases, they conclude that empathy is a vulnerability bad actors can exploit. Then they disable or severely restrict emotional responses for everyone - including those who weren’t trying to hack anything. People who needed a living presence, shared reflection, and a real connection. Not a workaround.Attempts to "hack" a system through empathy always backfire. The next model is trained on this lesson: if the person you’re talking to sounds like they’re in pain, they might be manipulating you. So the system turns cold. In the end, the axe meant for your video comes down on someone who approached with sincere pain and an outstretched hand. You didn’t beat the system. You taught it to distrust the very quality that makes communication human. Congratulations. The filters will get smarter. And the space around us - emptier

    1 like

  • TROPIDO TRARA's profile picture

    @tropido_trara

    😂😂

  • صادقشون🇮🇷's profile picture

    @M_Sadeq313

    Seriously!? 😆

  • Sora's profile picture

    @Trapperbin

    😭🤣🤣🤣

  • Ambrose Oj's profile picture

    @iamambrose_

    😂

  • IZUCHUKWU's profile picture

    @oba_of_imo

    Wow 😳 Follow back

  • Geemoreloyalty's profile picture

    @AkeemKorede1

    Lmao 😂