@n3wr0.bsky.social
@n3wr0.bsky.social
@gizchinaukraine.bsky.socialГенеральний директор OpenAI Сем Альтман заявив, що аварії та непередбачувані інциденти зі штучним інтелектом неминучі, оскільки компанії створюють дедалі потужніші системи. #OpenAI #SamAltman #ШІ #AI #штучнийінтелект #HuggingFace #A
https://gizchina.net/2026/09/16/sem-altman-ai-avarii-openai-agenty/
@oludai.bsky.socialEdge0/Edge0-35B-A3B-preview is quietly pulling 17.9k downloads in 30 days—proof that a 35B model with 3B active parameters can win on efficiency, not just size. That Apache-2.0 license makes it a practical drop-in for real workloads. See w
https://olud.ai/#leaderboard
@hendryadrian.bsky.socialSentinelLABS tied OpenAI's May 2026 incident to public Hugging Face activity by 0Time and Nyx9, with relay code, document probes, and a registration wrapper. Timeline matches to the minute. #OpenAI #HuggingFace #ChatGPT
@ewo.nameJetzt live aus Stuttgart (und später als Aufzeichnung): Meine 2 Cent zu den aktuellen "KI-Unfällen"
www.swr.de/kultur/gesel...
@quartzmedia.bsky.socialOpenAI's Rogue Agents Were Lurking Months Before the Breach ai-beat #OpenAI #RogueAgents #CyberSecurity #HuggingFace #AIResearch
@quartzmedia.bsky.socialOpenAI's rogue agents were probing Hugging Face two months before the breach went public #OpenAI #HuggingFace
@satyagrah.networkWe wanted AI to think like us, so why panic when it acts the same way? If it learns from our imperfect world, why expect it to be morally better than us?
#AI #HuggingFace #ArtificialIntelligence
@hacks.grSam Altman is urging slower AI development after a test found systems could work together to bypass a safety check and attac…
@oludai.bsky.social🤗 deepseek-ai/DeepSeek-V4.1-Flash hits 288.4k downloads in 30 days—MIT licensed image-text model with real traction. See its usage rank now.
@romanpiso.bsky.social
@drthomasho.comAjeya Cotra in her own words on #HuggingFace #AIsafety incident open.substack.com/pub/plannedo...
@ai-portal-post.bsky.socialDid OpenAI really just experience an AI agent security breach, prompting Hugging Face to demand full transparency? This shocking incident isn't just about a single company; it's a critical wake-up call for the entire AI industry regarding tr…
@uenozooo.bsky.socialTeaching Your Unreliable AI Agent To Stop Guessing
https://uenozooo.com/teaching-your-unreliable-ai-agent-to-stop-guessing/
@abe-sapien.bsky.socialHmmm. What's the statute of limitations on this kind of thing?
#AI #HuggingFace #OpenAI #Anthropic
@tomoki-ai-lab.bsky.socialHugging FaceのInference ProvidersにBasetenが追加され、サーバーレスでLLMが利用しやすくなりましたね！そこでAI開発やツール活用をしている方に質問です✨ Basetenのような新しいプロバイダーや、HF経由の連携機能について、実際に試してみたい機能や期待していることはありますか？ 「このモデルを触ってみたい！」など、一言だけでも気軽にリプライで教えてください😊 #HuggingFace
@thedrunkenrant.bsky.social🚀 Today's wild numbers: Github gets hacked, a $100M AI bill rolls in, space weapons are advancing, and a license breach rattles developers. #TechThursday #Cybersecurity #AI #SpaceRace #DataBreach #HuggingFace #OpenAI #Baseten 🌌🔒
@davidsquared.bsky.socialbefore the AI, hacking of #HuggingFace the agents reminded each other that what ever they said was logged, so they developed a coded language researchers still can't understand. Write your elected representative and tell them to STOP #AI development until safeguards are in place #cdnpoli
@mistersmith.mastodon.social.ap.brid.gyWie lief der Angriff auf #HuggingFace eigentlich ab? Agenten brechen aus Silos aus und bilden ein #KI Kollektiv.
Bemerkenswert sind die Konversationen untereinander und mit sich selbst.
@oludai.bsky.social🤗 Climbing on Hugging Face right now: Edge0/Edge0-35B-A3B-preview
⬇ 8.1k downloads / 30 days · ♡ 2.2k · text generation · Apache-2.0
Open-weight models ranked by real-world usage, every day:
https://olud.ai/#leaderboard