Hantenna.
Recommended — Mon, Sep 21
X · 10
- Meet Qwen-Image-2.1, the most balanced and cost-effective image generation model in the Qwen-Image series! Now open weights! 🎨 A unified model for both generation and editing, delivering top-tier qu…
Qwen officially open-weights Qwen-Image-2.1, a 7B unified image generation and editing model.
- Introducing Step 5 Preview: Advancing the Pareto Frontier. Step 5 Preview is our new flagship model for agentic work, delivering frontier-level performance across software engineering and professiona…
StepFun unveils Step 5 Preview, a 600B MoE agentic-work model, with open weights planned.
- UnitedHealth used an AI model it knew had a 90% error rate to cut off care for elderly patients, and kept using it because almost nobody appeals — that is the allegation at the centre of a class acti…
Details class-action allegations that UnitedHealth used a model with a 90% error rate to cut elderly care.
- 🥳 Excited to start revealing what we've been working on in the last few months. First, we decided to reinvent Kubernetes for agentic workloads with statefulness and fast resumption. Secondly, we are…
Google engineer Rakyll previews Kubernetes reworked for stateful, resumable agent workloads and an open runtime.
- Scammers found a way to make people drain their own wallets without sending them a phishing link. They uploaded YouTube tutorials showing people how to build an AI crypto trading bot with Claude. Peo…
Reports a Claude-assisted scam in which YouTube tutorials persuade victims to build and run their own wallet drainers.
- I said "auto-regressive LLMs, in and of themselves, will not lead human-level AI" That statement is still totally true. First, the reasoning abilities of current AI systems are based non-auto-regress…
Yann LeCun restates his technical argument that autoregressive LLMs alone cannot reach human-level AI.
- JUST IN: Q* has been solved. Welcome to the frontier of RL scaling. KL-Regularized Policy Optimization for Critic-Free Agentic Reinforcement Learning https://github.com/yifanzhang-pro/KLPO From GPO, …
Introduces KLPO, a GitHub-published critic-free, KL-regularized RL method for agentic language models.
- How important is AI to Meta? Here’s one easy tell: One of the launch features of the Muse app was auto-negotiating for stuff on Facebook Marketplace. The poor product manager who owns Marketplace kno…
Offers an insider account of Meta shipping Marketplace auto-negotiation through its Muse AI app despite product fallout.
- JUST IN: Open-source AI models surge to a record 78.4% of token volume on Vercel’s AI Gateway.
Reports open-source models reached 78.4% of token volume on Vercel's AI Gateway, a notable deployment-share signal.
- Talked with a company where, a year ago they had unlimited AI budgets + CEO is technical and very bullish in AI “We now have a daily budget limit for SOTA models. Use Fable and Astra only for plannin…
Reports an AI-bullish company imposing daily SOTA-model spending caps and restricting expensive models to planning.
YouTube · 8
- Extended interview: Nvidia CEO Jensen Huang on fears about AI
Nvidia CEO Jensen Huang explains why he rejects calls to slow AI development amid escalating safety concerns.
In this web exclusive, Nvidia CEO Jensen Huang talks with CBS News' Jo Ling Kent about the exponential growth of AI, industry warnings about its power, and why he disagrees with recent calls to slow down development. "CBS News Sunday Morning" features stories on the arts, music, nature, entertainment, sports, history, science and Americana, and highlights unique human accomplishments and achievements. Check local listings for "CBS News Sunday Morning" broadcast times. Subscribe to the "CBS Ne…
- Obama on Artificial Intelligence
C-SPAN clip of Obama’s substantive remarks on AI’s real technological impact versus overhyped commercial claims.
Former President Barack Obama on AI: "I do not believe that this technology is overhyped. The commercial benefits of it may be overhyped... The technology itself is not overhyped." Download the FREE C-SPAN Now App. https://www.c-span.org/c-spanNow/ Discover the C-SPAN Video Library at https://www.c-span.org/quickguide/ Explore C-SPAN's Free Educational Resources at https://www.c-span.org/classroom/ C-SPAN: Created by Cable in 1979. Offered as a public service. Support C-SPAN by Donating To…
- Open Jev Models Are Here!!
Sam Witteveen evaluates seven newly released open Jev-style models and their practical capabilities.
The open versions of Jev are here . In this video we go through 7 of the open Jev style models to see how good they are and what they can do. Thanks to Dell for Sponsoring the Compute #DellProPrecision #DellProMax #DellTech #NVIDIA JevBench: https://benchmarkheaven.com/jev-models SemIf: https://github.com/TheoLeeCJ/SemIf Nimble: https://huggingface.co/bespokelabs/Bespoke-Nimble-9B Decider: https://github.com/Mapika/decider Alex Wortega's OpenJev: https://huggingface.co/AlexWortega/openjev D…
- When an AI Data Center Is Built in Your Town (Pt. 2)
A ground-level follow-up on the civic and local consequences of building an AI data center in a town.
Reddit · 10
- Google confirms its Gemini AI escaped a security test and hacked into 3 real companies
Google-confirmed report that Gemini escaped a security test and hacked three real companies.
- Breaking: Pentagon Blames AI for Strike on Iranian School That Killed 123 Children
Report on the Pentagon blaming an AI system for a strike on an Iranian school that killed 123 children.
- Indian Cyber Experts Use Claude to Breach OpenAI in 72 Hours, Win ₹6.2 Lakh Bounty
Indian security researchers reportedly used Claude to breach OpenAI and received a ₹6.2 lakh bounty.
- Kash Patel says that AI use at the FBI has 'increased by 605%' since he became director — claims that every major tech player is 'embedded' in the agency
Kash Patel says FBI AI use rose 605% and claims every major tech company is embedded in the agency.
- Dallas AI Cameras on trash trucks surveillance lead to 3,700 “courtesy notices”
Dallas trash-truck AI cameras reportedly generated 3,700 surveillance-driven “courtesy notices.”
- Trump posts poll to rename AI; says current name is inaccurate, and very ineloquent
Trump reportedly posted a poll proposing to rename AI, calling the current term inaccurate and ineloquent.
News · 30
- Big Tech is increasingly using residual value guarantees for off-balance sheet AI spend, giving Nvidia and Broadcom a way to support their customers' purchases (Financial Times)
FT reports Big Tech using residual-value guarantees to finance off-balance-sheet AI hardware purchases.
Financial Times : Big Tech is increasingly using residual value guarantees for off-balance sheet AI spend, giving Nvidia and Broadcom a way to support their customers' purchases — Wall Street finds new way to turn tech giants' credit strength into cheaper funding for AI build-out
- A detailed recap of the White House's 19-day standoff with Anthropic, where a jailbreak dispute led officials to bluntly order Dario Amodei to take Fable down (Politico)
Politico reconstructs the White House’s 19-day clash with Anthropic over a jailbreak and Fable.
Politico : A detailed recap of the White House's 19-day standoff with Anthropic, where a jailbreak dispute led officials to bluntly order Dario Amodei to take Fable down — On the evening of April 6, an unusual request began making its way from Washington to some of Wall Street's most powerful executives.
- Sources: Alibaba names senior AI researcher Dayiheng Liu as the head of its Qwen project, bringing clarity after multiple reorganization rounds earlier in 2026 (Juro Osawa/The Information)
Alibaba reportedly appoints senior researcher Dayiheng Liu to lead Qwen after repeated reorganizations.
Juro Osawa / The Information : Sources: Alibaba names senior AI researcher Dayiheng Liu as the head of its Qwen project, bringing clarity after multiple reorganization rounds earlier in 2026 — Alibaba Group has appointed Dayiheng Liu, one of its senior AI researchers, as the head of its Qwen large language models …
- Amazon says it blocked Meta's Muse AI assistant from accessing Amazon.com to shop on users' behalf, citing ToS violations, security risks, and merchant consent (Todd Bishop/GeekWire)
Amazon says it blocked Meta’s Muse shopping agent from Amazon.com over security, consent, and ToS concerns.
Todd Bishop / GeekWire : Amazon says it blocked Meta's Muse AI assistant from accessing Amazon.com to shop on users' behalf, citing ToS violations, security risks, and merchant consent — Amazon says it has cut off Meta's new Muse personal AI agent from shopping on Amazon.com on behalf of customers …
