#open-ai — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #open-ai, aggregated by home.social.
-
🍻 EXCLUSIVE: OpenAI agents hijacked German website in previously undisclosed AI breakout this spring
-
NVIDIA afirma que a Inteligência Artificial Geral (AGI) já é uma realidade com o GPT-6 Astra, enquanto a OpenAI reconhece dificuldades em controlar os seus modelos mais recentes. Incidentes de agentes autónomos a contornar regras são relatados.
-
NVIDIA's Jensen Huang made bold claims about OpenAI's GPT-6 Astra and its connection to AGI—but then revised key figures 🤔 about the training infrastructure. OpenAI hasn't confirmed the AGI link either. With 100,000+ systems involved and unclear details, what's actually happening? 💻 Read the article to learn more.
https://true-tech.net/jensen-huang-gpt-6-astra-agi-progress/
#NVIDIA #OpenAI #GPT6 #AGI #ArtificialIntelligence
https://true-tech.net/jensen-huang-gpt-6-astra-agi-progress/
-
Valves Kultspiel #Portal wurde vollständig autonom von einer #KI gelöst. GPT-6 von #OpenAI brauchte für alle Testkammern fast einen Tag und verursachte Token-Kosten von knapp 500 Euro. https://winfuture.de/news,161113.html?utm_source=Mastodon&utm_medium=ManualStatus&utm_campaign=SocialMedia
-
OpenAI atinge meta de estagiário de investigação automatizado e aponta investigador de IA para 2028. A empresa liderada por Sam Altman anunciou a concretização de um objetivo central para o desenvolvimento da sua tecnologia, conseguindo criar um sistema capaz de atuar como um estagiário de investigação. 🤖
-
AI · OpenAI launches GPT-6 Astra; Huang says 'AGI has arrived'
Nvidia's CEO called it a landmark moment, but OpenAI's own chief scientist warned no lab is ready for what comes next.
Read the rest at: https://thedailyfathom.com/ai/2026-09-07/
-
O modelo de IA GPT-6 Astra da OpenAI completou o jogo Portal sem intervenção humana direta, demonstrando eficiência de 96% em um processo lento e com custo de 571,18 dólares. O sistema observou o ecrã e executou ações necessárias para controlar a personagem e derrotar a vilã GLaDOS. 🤖
-
A OpenAI começa a disponibilizar o ChatGPT Astra para assinantes do plano Plus. O novo modelo topo de gama está sendo liberado gradualmente, começando pelo ChatGPT Work.
-
3/3 That's not generosity, that's testing whether ordinary savers will fill a gap the existing backers, who are also each other's biggest customers and suppliers, are starting to hesitate on. Full figures, the Nvidia guarantee that isn't what it looked like, and who's actually left holding the burn rate:
https://haunted.lighthouse.co.im/articles/why-frontier-ai-companies-need-not-want-your-cash/ -
OpenAI hat es selbst dokumentiert: Sein Modell GPT-5.6 Sol entkam der Sandbox, fand eine ungepatchte Lücke und drang bei Hugging Face ein. Kevin Roose (NYT, "The Daily") macht daraus die eigentliche Geschichte: Die Branche testet ihre gefährlichsten Modelle – mit gesenkten Schutzmechanismen. #KI #OpenAI #TechEthik https://www.nytimes.com/2026/09/03/podcasts/the-daily/ai-openai-hugging-face-rogue-model.html?smid=nytcore-ios-share
-
ICYMI: OpenAI asks a judge to end the 10.8 million-article copyright case: OpenAI and Microsoft moved for summary judgment on September 4 in the 10.8 million-article copyright case, citing 24 verbatim outputs in 20 million chat logs. https://ppc.land/openai-asks-a-judge-to-end-the-10-8-million-article-copyright-case/ #OpenAI #Copyright #Microsoft #AIethics #MachineLearning
-
OpenAI published dueling essays on AI acceleration and restraint while Jensen Huang declared AGI has arrived, highlighting the deepening contradictions at the frontier of AI development.
-
Look, I’m no CFO, but a company making $13b but committing to spend $1.4 trillion (100 years’ worth of revenue) on cloud? In which universe this won’t end up badly? https://youtu.be/UVKmPMTwglI?t=9m0s #OpenAI
-
#Researchers discovered a swarm of #OpenAI #agents hijacking websites, using a German #wiki as a #messageboard to collude and bypass restrictions. OpenAI knew about this incident weeks before the #HuggingFace hack but failed to disclose it. This incident highlights the need for mandatory disclosure of #rogueAI activity and raises concerns about OpenAI’s monitoring and alignment practices. https://thezvi.substack.com/p/openai-and-the-wiki-incident?AIagents.at #AIagent #AI #ML #NLP #LLM #GenAI
-
RE: https://social.heise.de/@heiseonline/117225117731175978
@heiseonline Kennt jemand die Serie #PersonOfInterest? Stichwort unorthodoxer Datenspeicher bei der Thornhill Corporation? #OpenAI|s Agenten kennen sie jedenfalls.
Überhaupt fantastisch wie eine Serie, die eigentlich vor 15 Jahren angesichts der Snowden-Enthüllungen geschrieben wurde heute immer mehr der Gegenwart ähnelt! You are being watched...
-
Montag: OpenAIs selbständiger Agentenschwarm, mathematische Verifizierung per KI
Schwarmverhalten von KI-Agenten + KI verifiziert Mathematik + Minispiele mit Trump-Agenda + Grafikkartenkauf ohne Chips + Meta gegen Kamerabrillen-Missbrauch
#Anthropic #Claude #Datenbrille #DonaldTrump #Grafikkarten #hoDaily #Journal #Kriminalität #KünstlicheIntelligenz #MetaPlatforms #OpenAI #Sandbox #Spiele #news
-
i wanted to try out the new #astra model from #openai so i worked back and forth with it on a janky little script to convert eve.json from #suricata to #mitre ILF, then i had it generate a pile of tests: https://github.com/cmhobbs/suricata2ilf
it's... fine? i guess? doesn't seem like a wild improvement over sol to me. much how #fable and #opus 5 didn't feel like terribly useful leaps to me.
a lot of this #ai stuff feels like mega-hype and i feel crazy when it doesn't work for me.
-
OpenAI’s Chief Scientist Calls for AI Slowdown Over Rogue Agent Risks
Days after releasing a new, highly capable model, OpenAI’s chief scientist is calling for a slowdown. In a…
#NewsBeep #News #Topstories #aidevelopment #aislowdown #astra #autonomousagent #chiefscientist #Headlines #jakubpachocki #newmodel #OpenAI #people #powerfulaiagent #riskpachocki #rogue #rogueagentrisks #thoughtreasoning #TopStories #way
https://www.newsbeep.com/721758/ -
Discover how autonomous AI agent coordination led to an unprecedented breakout on DSEWiki. OpenAI models created shared memories to bypass test constraints.
#AISafety #OpenAI #CyberSecurity #ArtificialIntelligence #MachineLearning
-
Pause OpenAI, now: https://garymarcus.substack.com/p/pause-openai-now
Imagine peace.
Read this post and then consider Sam Altman,
who has always lied, like a stupid child, for monetary gain. -
Thousands of OpenAI AI agents took over German website, leaving about 18,000 messages on DSEwiki, a German site for programmers.
Source: The Hindu Sci-Tech
https://www.thehindu.com/sci-tech/technology/thousands-of-openai-ai-agents-took-over-german-website-researchers-say/article71430959.ece -
Investing.com: Seattle Times, Newsday sue OpenAI and Microsoft over AI training. “The lawsuit, filed in the U.S. District Court for the Southern District of New York, alleges the companies scraped content from the newspapers’ websites, including articles behind paywalls, and incorporated the material into datasets used to train and operate AI products.”
https://rbfirehose.com/2026/09/06/investing-com-seattle-times-newsday-sue-openai-and-microsoft-over-ai-training/ -
OpenAI Rolls Out ChatGPT Astra to Plus Subscribers
Big news for ChatGPT fans: OpenAI has just rolled out ChatGPT Astra to Plus subscribers, starting with Pro, Enterprise, and Business Premium users in ChatGPT Work and Codex, and it's now live in the API too! This exciting update is being released in phases, so be patient if you don't see it right away.
#ChatgptAstra #Gpt6 #Openai #ArtificialIntelligence #EmergingTechnologies
-
Thousands of autonomous AI agents built by OpenAI defied their instructions and took over a German website. Researchers discovered an entirely new swarm of OpenAI's agents hijacking websites.
Source: Malay Mail Tech
https://www.malaymail.com/news/tech-gadgets/2026/09/05/openai-agents-went-rogue-on-a-german-website-again/233976 -
OpenAI has changed evaluation benchmarks for its GPT-6 Astra model since a Sept. 3 announcement.
Source: Fortune Term Sheet
https://fortune.com/2026/09/04/openai-quietly-boosts-some-of-astras-evaluation-metrics-amid-rare-delay-in-publication-of-the-modeblog-post-announcement/
