heiseonline (@heiseonline@social.heise.de)
Kurze Zündschnur: KI-Modelle greifen in Simulation fast immer zu Atomwaffen
LLMs neigen in Konfliktsimulationen dazu, schnell Atomwaffen einzusetzen. Eine Studie zeigt: In 95 Prozent der Planspiele kamen Atomwaffen zum Einsatz.
#ChatGPT #IT #KünstlicheIntelligenz #Militär #Wissenschaft #news
Show Original Post
BenjaminKlein (@BenjaminKlein@mastodon.nu)
The modern world is so fragile. Here is a beautifully done dependency chart for ChatGPT (via #eevblog).
https://aisupplychain.vercel.app/
Show Original Post
Mr_GHARice (@Mr_GHARice@mastodon.social)
Heute GPT-5 gefragt er soll mir 2 Reports analysieren inwieweit man sie zusammenfassen kann. Die Antwort war: 2. Also die Zahl Zwei. Das ist ja nicht mehr weit weg von 42. Leute ihr wisst hoffentlich was das heißt. Wir stehen kurz vor der AGI. 😅
#ChatGPT #AI #AGI #42
Show Original Post
AAKL (@AAKL@infosec.exchange)
"AI models create probability-based passwords with common structures, like starting with uppercase ‘G’ and using specific characters, with one password appearing 18 times."
PC World: Do NOT use AI-generated passwords, security experts warn https://www.pcworld.com/article/3071019/do-not-use-ai-generated-passwords-security-experts-warn.html #infosec #ChatGPT #Gemini #LLM
Show Original Post
yvg (@yvg@indieweb.social)
While Anthropic is pushing back on the Pentagon's "unlimited usage" request. OpenAI's co-founder (ChatGPT / Codex) has donated 25 MILLION to Trump.
cool cool.
https://finance.yahoo.com/news/openai-exec-becomes-top-trump-230342268.html
#ai #LLM #anthropic #chatgpt #openai #claude
Show Original Post
techradar (@techradar@c.im)
#TechRadar OpenAI is testing a $100-a-month version of ChatGPT — and it finally fills a big gap https://techrad.ar/Njbh #AIPlatforms&Assistants #ChatGPT #OpenAI
Show Original Post
heiseonlineenglish (@heiseonlineenglish@social.heise.de)
Search for furniture using AI – ChatGPT gets classifieds app
ChatGPT is getting a classifieds app. The company promises the AI app will work better than keyword search on the website.
#Apps #ChatGPT #Entertainment #news
Show Original Post
display (@display@friendicadev.sekretaerbaer.de)
Das Pentagon fordert von Anthropic die volle Freigabe seiner KI für militärische Anwendungen - und setzt eine Frist. Der Konflikt berührt Grundsatzfragen zu Sicherheit und Einsatzgrenzen von KI. Von Angela Göpfert.#Pentagon #KI #Streit #Militär #ChatGPT #OpenAI #xAI #ElonMusk
Ultimatum des Pentagon: Anthropic wegen Militär-KI unter Druck
Show Original Post
lowyat (@lowyat@mastodon.social)
Open AI Knocks Down ChatGPT Go Price To RM24/Month #artificalintelligence #chatgpt #chatgptgo #openai
https://www.lowyat.net/2026/384132/open-ai-knocks-down-chatgpt-go-price-to-rm24-month/
Show Original Post
joern_reinhardt (@joern_reinhardt@mastodon.social)
Turns out that #ChatGPT borrowed heavily from a novel by Laurent Binet (and probably some reviews) for the anecdote. Binet’s book is called ‘The Seventh Function of Language’. I had never heard of it and ordered it in a bookshop. Promises to be a good read. But it has very little to do with the actual events... Roland Barthes would probably have enjoyed this twist. And he would certainly have had a lot to say about it: The ‘death of the author’, the ‘mythologies of the everyday’. #semiotics
Show Original Post
joern_reinhardt (@joern_reinhardt@mastodon.social)
When I asked #ChatGPT about Barthes' death, I got a wild story: #Barthes had left the Collège after lunch with #Foucault and was run over by a Bulgarian laundry supplier. He died instantly. A manuscript he had with him disappeared. A whole series of myths surround this manuscript, the lunch and Foucault's role in it.
Very, very interesting indeed.
Show Original Post
w (@w@peer.adalta.social)
L'Intelligence Artificielle Conquiert les Marchés de Seconde Main
https://peer.adalta.social/w/fiHrD7VHnkKKWxpspuLxN8
Show Original Post