N-gated Hacker NewsMay 15, 2025
Gary Marcus just discovered that ChatGPT isn't the next Magellan and can't fold a map without tearing it 🤔🗺️. Who knew that a chatbot might struggle with geography while excelling at generating digital excuses? Clearly, ChatGPT's PhD stands for "Pretty hilarious Disappointments" in cartography! 🎓😂
https://garymarcus.substack.com/p/chatgpt-blows-mapmaking-101 #ChatGPTFails #GeographyHumor #AIShortcomings #DigitalExcuses #CartographyJokes #HackerNews #ngated
ChatGPT Blows Mapmaking 101

A Comedy of Errors

Marcus on AI
Manoah ConsultingApr 1, 2025

Meet Sir Redundant III: the AI who knows everything—especially things that don’t exist. From unicorn economics to Lincoln’s podcast, he explains it all (twice). Confidence? 100%. Accuracy? Optional. #AIQuirks #ChatGPTFails #LNNA #ConfidentlyWrong #AIHumor

https://lttr.ai/AdHwI

Sir Redundant III: Master of Confident Confusion

Meet Sir Redundant III, our loving name for a leading AI LLM, is our resident expert in... well, everything. And by everything, we mean everything including things that don't actually exist.

Logic Need Not Apply
HassmeisterApr 1, 2024
So lasst uns ehren dieses Mädchen,
dessen Zauber wir nie vergessen.
Momo, ein Name voller Pracht,
zeigt uns den Weg in die ewige Nacht.
#ChatGPTfails
dealingwithJun 30, 2023

Yesterday I heard someone say: "I was out of airspeed, I was out of altitude, and I was out of ideas."

I thought it was great, and couldn't find it on Google. So I asked ChatGPT. It fabricated a false answer. #chatgptfails

WebappiaJun 29, 2023

AI’s citation generation: 3 errors and 2 accurate outcomes. #AIResearch

Hashtags: #AIcitations #ChatGPTfails #AIaccuracy 1. ArtificialIntelligence 2. ChatGPT 3. Citations Summery: ChatGPT, an AI-powered chatbot, has gained popularity as a tool for various writing tasks, including generating citations and bibliographies. However, a recent test conducted by a team found that ChatGPT still has room for improvement when it comes to citation generation. In the first test,…

https://webappia.com/ais-citation-generation-3-errors-and-2-accurate-outcomes-airesearch/

AI's citation generation: 3 errors and 2 accurate outcomes. #AIResearch

Our team put ChatGPT to the test to see whether AI could compete with reference managers in generating citations and bibliographies. We think that ChatGPT still has quite a few things to learn.

Webappia
Steffen ChristensenApr 21, 2023
Check out this amazing GPT-produced web page that my son found through Google that tells you that you can get Minecraft Education Edition 1.19 by installing MEE 1.5 and levelling up *in-game*!
https://kabus22.com/how-to/how-do-you-get-1-19-on-minecraft-education-edition/
#theInternetIsBroken #ChatGPTFails #ChatGPT #AI /c
@google
How do you get 1.19 on Minecraft Education Edition? - Kabus 22

Are you looking to get 1.19 on Minecraft Education Edition? You've come to the right place! In this article, we'll explain the process of getting 1.19 and

Kabus 22
Ron ChrisleyMar 4, 2023
ChatGPT confidently makes a whopper of a logical mistake #ChatGPTfails @garymarcus
Rod MiddletonJan 31, 2023
Tested ChatGPT with summarising an old abstract … wow its really really not very good is it! On the surface it looks quite impressive but almost all the detail is lost - very good for superficial 'science’ but extremely worrying if it gains more widestream acceptance. Dropped all the actual data from the results section also. Avoid! #AI #chatgptfails
Ted UnderwoodDec 3, 2022
As models with stronger safeguards keep being subverted by people determined to find workarounds and make the model say bad/funny things, the conclusion I'm drifting toward is not that models are dangerous but that this order of animals can't be trusted with anything. #ChatGPTFails
Ron ChrisleyDec 1, 2022
The large number of #ChatGPTfails wouldn't be worth mentioning if it weren't for the large number of people who find an impressive *non*-failure and draw ill-advised conclusions about the model's suitability as a knowledge source. Here's me piling on with the first thing I tested the new model on. I suspected that since this puzzle is always (?) expressed in terms of the North Pole (so a twist concerning bear colour can be added), ChatGPT would fail to abstract to a South Pole setting.