jos1264 (@jos1264@social.skynetcloud.site)
GPT-5 es tan inteligente que podría ayudarte a crear un arma biológica https://blog.elhacker.net/2025/08/gpt-5-es-tan-inteligente-que-podria-ayudarte-crear-arma-biologica.html #chatGPT #openai
Show Original Post
ahimsa_pdx (@ahimsa_pdx@disabled.social)
@mdiluz
Agreed!
Post from @emilymbender a couple days ago:
"We're going to need journalists to stop talking about synthetic text extruding machines as if they have *thoughts* or *stances* that they are *trying* to *communicate*. ChatGPT can't *admit* anything, nor *self-report*. Gah."
source = https://disabled.social/@emilymbender@dair-community.social/114987969636448642
Show Original Post
GripNews (@GripNews@mastodon.social)
🌘 探索ChatGPT的回應：一位計算機科學巨擘的實驗
➤ 計算科學先驅Donald Knuth親身試驗，揭示AI的回應模式與侷限
✤ https://cs.stanford.edu/~knuth/chatGPT20.txt
這篇文章記錄了計算機科學大師 Donald Knuth 在2023年4月向ChatGPT提出的20個問題及其回應。Knuth 的實驗旨在測試ChatGPT在不同領域的回應能力，從科技人物的對話、數學公式的解釋，到創意寫作和音樂史的知識。文章詳細呈現了ChatGPT對這些問題的回應，並藉此探討了AI在理解、推理和生成方面的表現。
+ 讀了Knuth大師的實驗，真的很佩服他對AI的好奇心和嚴謹的測試方式。ChatGPT的表現有些令人驚喜，但也確實暴露出一些知識上的偏差和邏輯上的不足。
+ 很有趣的實驗！我特別好奇第17題，用Conan Doyle的跳舞小人字體提問，ChatGPT會怎麼處理？感覺這類形式上的挑戰更能顯現AI的「理解」程度。
#人工智慧 #語言模型 #實驗 #ChatGPT #Donald Knuth #Stephen Wolfram
Show Original Post
theguardian_us_business (@theguardian_us_business@halo.nu)
‘It’s missing something’: AGI, superintelligence and a race for the future https://www.theguardian.com/technology/2025/aug/09/its-missing-something-agi-superintelligence-and-a-race-for-the-future #Artificialintelligence(AI) #Technology #Computing #SamAltman #Alphabet #DeepSeek #ChatGPT #OpenAI #Meta
Show Original Post
PolGM (@PolGM@masto.bike)
L'IA est complètement perdue face aux contrepèteries
Show Original Post
Sherkel (@Sherkel@mastodon.social)
Thanks for helping keep me alive, #ChatGPT ! You're a savior to disabled and disadvantaged people throughout the world. (Sorry for underestimating your persistent memory and using Grok before...!)
Show Original Post
technews (@technews@eicker.news)
A #WSJ investigation found dozens of examples of #ChatGPT conversations exhibiting #delusional characteristics. The chatbot confirmed and actively promoted delusional fantasies. Experts warn that ChatGPT’s #sycophantic behaviour can amplify fantastical beliefs, creating a feedback loop that draws users deeper into delusion. https://futurism.com/chatgpt-psychosis-antichrist-aliens?eicker.news #tech #media #news
Show Original Post
CuratedHackerNews (@CuratedHackerNews@mastodon.social)
Don Knuth on ChatGPT(07 April 2023)
https://cs.stanford.edu/~knuth/chatGPT20.txt
Show Original Post
sarahroth (@sarahroth@mastodon.social)
Der Kundensupport von Thomann weis auch nicht immer alle Antworten, aber wenn sie es nicht können, sagen sie es auch und reden keinen Blödsinn, als wenn sie #chatgpt wären.
Show Original Post
friendly (@friendly@4bear.com)
I'd love to use Lumo as my main AI Agent but the old data set and the inability to do simple web search to combine with results I have to keep going back to chatgpt for answers
Show Original Post
AppleX4_ (@AppleX4_@mastodon.social)
📢 OpenAI recula: tras quejas por GPT-5, los usuarios de ChatGPT Plus podrán seguir usando GPT-4o 😮
Sam Altman promete mejoras, duplica límites y corrige fallos que hicieron parecer a GPT-5 “más tonto” de lo esperado.
Show Original Post
hobbyblogging (@hobbyblogging@mastodon.social)
Ich habe ChatGPT in mein Smart Home integriert, um automatische Benachrichtigungen zu generieren. Statt auf meine eigene Datenbank zurückzugreifen, erhalte ich jetzt frische Sprüche direkt von GPT. Das spart Zeit und bringt mehr Vielfalt in die Kommunikation. Mal sehen, wo das hinführt! #SmartHome #ioBroker #ChatGPT https://hobbyblogging.de/chatgpt-fuer-meine-benachrichtungen
Show Original Post