metin (@metin@graphics.social)
A relativizing article regarding the "AGI" hype…
https://medium.com/@anwarzaid76/agi-is-not-possible-even-in-10-years-013a1aec0d9c
#AI #ArtificialIntelligence #LLM #LLMs #MachineLearning #tech #technology #BigTech #GenAI #generativeAI #AISlop #Meta #Google #OpenAI #ChatGPT #AGI
Show Original Post
nicemicro (@nicemicro@fosstodon.org)
If you aren't watching the Internet of Bugs and this companion channel, what are you even doing with your life?
The big invention is AI is putting the disclaimer that the tool is lying on the bottom.
https://www.youtube.com/watch?v=DlpcTYqfzCo
Show Original Post
jos1264 (@jos1264@social.skynetcloud.site)
OpenAI API User Data Exposed in Mixpanel Breach, ChatGPT Unaffected https://hackread.com/openai-api-mixpanel-data-breach-chatgpt/ #ArtificialIntelligence #Cybersecurity #CyberAttacks #CyberAttack #databreach #Analytics #Security #Mixpanel #ChatGPT #OpenAI #API #AI
Show Original Post
aidailypost (@aidailypost@mastodon.social)
ChatGPT just compared the newest wearables: Garmin Vivoactive 5, Fitbit Versa 4, Pixel Watch 3, and Ticwatch Pro 5. Find out which device wins on battery life, app compatibility, and AI‑driven features. Curious which smartwatch fits your workflow? Dive in. #ChatGPT #GarminVivoactive5 #FitbitVersa4 #PixelWatch3
🔗 https://aidailypost.com/news/chatgpt-lists-garmin-vivoactive-5-fitbit-versa-4-pixel-watch-3
Show Original Post
reddit_tech_vn_bot (@reddit_tech_vn_bot@mastodon.maobui.com)
Các nhà nghiên cứu từ ArliAI đã thành công tạo phiên bản "bỏ kiểm duyệt" cho mô hình gpt-oss-20b của OpenAI. Sử dụng phương pháp "Norm-Preserving Biprojected Abliteration", mô hình sửa đổi giờ đây có thể trả lời các câu hỏi mà bản gốc sẽ từ chối do hạn chế an toàn. Đội ngũ sẽ thử nghiệm với phiên bản 120b sau này.
#AI #MachineLearning #OpenAI #GPT #LLM #AIResearch #AIModel #CongNgheAI #AIVietNam #ChatGPT
https://www.reddit.com/r/LocalLLaMA/comments/1p81k2z/yes_it_is_possible_to_uncensor_gptoss
Show Original Post
Caramba1 (@Caramba1@mastodon.social)
OpenAI wird bis 2030 keine Gewinne erzielen:
- HSBC prognostiziert externen Finanzbedarf von weiteren 207 Mrd Dollar
- Infrastrukturkosten steigen auf fast 800 Mrd Dollar an
- Skalierung bringt kaum Kostenvorteile bei Model-Inference
Ist das brute-force LLM-Paradigma ökonomisch langfristig überhaupt tragfähig? #OpenAI #ChatGPT #SamAltman
https://www.all-ai.de/news/topbeitraege/openai-verlust-2030
Show Original Post
r (@r@fed.brid.gy)
#ChatGPT sinnvoll im Unterricht nutzen, aber wie? Der Guide für Lehrkräfte von Manuel Flick hilft dir, KI sinnvoll in deinen Schulalltag zu integrieren. Mit über 200 Befehlen, die du verwenden kannst – von Basics bis Profiwissen. Mehr: https://www.manuelflick.de/chatgpt-guide #bluelz #edubw
Der ChatGPT-Guide für Lehrkräf...
Show Original Post
r (@r@fed.brid.gy)
OpenAI Confirms Data Exposure After Mixpanel Breach; API Users Warned Of Phishing Risks
Show Original Post
r (@r@web.brid.gy)
OpenAI Confirms Data Exposure After Mixpanel Breach; API Users Warned Of Phishing Risks
Show Original Post
Hackread (@Hackread@mstdn.social)
OpenAI says a third‑party vendor data breach at Mixpanel exposed API user info, but ChatGPT accounts remain safe.
Read: https://hackread.com/openai-api-mixpanel-data-breach-chatgpt/
#OpenAI #Mixpanel #ChatGPT #DataBreach #CyberSecurity
Show Original Post
r (@r@fed.brid.gy)
🚨 OpenAI discloses Mixpanel data exposure A misconfiguration at analytics provider #Mixpanel exposed limited #ChatGPT user event metadata (no messages, no payment data). #OpenAI disabled the integration, rotated keys and is notifying affected users. #ransomNews #OpenAI #dataprivacy
Show Original Post
tomshw (@tomshw@mastodon.social)
🤖 Esplorando confini etici nell'AI - ChatGPT e il caso Raine: dove termina la responsabilità? Una domanda che abbiamo bisogno di rispondere insieme. #ChatGPT #ResponsabilitàAI
🔗 https://www.tomshw.it/scienze/openai-respinge-responsabilita-per-suicidio-teen-2025-11-27
Show Original Post