Hantenna.
Recommended — Mon, Sep 14
X · 10
- Watch a company being built live with @Grok @Bot!
Elon flags xAI’s planned three-day live Grok Bot demonstration of agents building a company from scratch.
- By now, it's amply clear that the big bosses of US frontier AI labs are really, really scared of China. And, fascinatingly, part of the answer why might be a tiny place in Inner Mongolia that you've …
Reports on China’s Ulanqab data-center expansion and its potential scale in the frontier-compute race.
- Latham & Watkins, the 2nd largest US law firm, is buying Nvidia hardware to fine-tune open weights in-house. Open weights + proprietary data + local compute = enterprise sovereign AI stack.
Reports that Latham & Watkins is buying Nvidia hardware to fine-tune open-weight models on proprietary data in-house.
- astra was setting up OBS, and it saw I didn't respond so it looked through camera to see if i was looking (which i wasn't) it made my mac beep which made me lookup and it flashed the question on scre…
First-hand account of GPT-6 Astra using a camera, Mac alert, and screen prompt to regain an inattentive user’s attention.
- 🚨EXCLUSIVE: China has now replied to Dario Amodei’s call to slow AI. China's state-backed Global Times says Dario Amodei's proposal to slow frontier AI is really a 'Cold War playbook' targeting Chin…
Documents Global Times’ response framing Amodei’s frontier-AI slowdown proposal as a Cold War bid to preserve US dominance.
- First time I’ve seen this one Chinese chain of thought in official Claude app, Fable 5.1 Reminds me of my friend GLM 5.3 flash, he thinks in Chinese a lot too
Shows Chinese-language chain-of-thought appearing in Anthropic’s official Claude Fable 5.1 app.
- Setting my subagent model to Opus has helped me save so much usage on Fable 5.1. It really likes to spawn subagents but mine rarely need Fable-level reasoning `CLAUDE_CODE_SUBAGENT_MODEL=opus`
Offers a concrete Claude Code cost-control configuration: route spawned subagents to Opus via CLAUDE_CODE_SUBAGENT_MODEL.
- A lot of people have asked, here's how I setup my auxiliary models in Hermes Agent. Gemini Flash saves a lot of dough, and astra gives me a second perspective when I run /review
Nous Research’s Teknium shares a specific Hermes Agent auxiliary-model setup using Gemini Flash and Astra for review.
- In the near term (definitely not in the long term), more capable models should mean safer models (maybe paradoxically). Current models are unsafe not because they're too smart, but because they take …
François Chollet argues near-term model capability gains can improve safety by reducing failures caused by weak reasoning.
- One of the most worrying risks linked to frontier AI is extreme power concentration. The only way to avoid extreme power concentration is to ensure we have multiple independent providers of frontier …
François Chollet identifies frontier-AI power concentration as a core risk and argues for multiple independent providers.
YouTube · 10
- Inside Bracket22, A Trading Firm Powered By Al Agents
CNBC profiles Bracket22, a trading firm using AI agents to trade crypto, stocks, and bonds with its founder’s own capital.
Brian Kelly previously ran a cryptocurrency hedge fund, but closed that shop in early 2025. Later that year, he began testing out uses of artificial intelligence. Bracket22 invests only Kelly's own capital and trades cryptocurrencies, stocks and commodities. It's powered entirely by AI agents — a stark example of a growing reality on Wall Street as firms test the benefits and limits of AI in finance. Chapters: 00:00 - 00:31 Introduction 0:32 - 3:11 AI Agents 3:12 - 4:05 Future for Wall Stree…
- New Deepseek, human genome map, Navier Stokes, GPT finance, Suno v6, YuE2: AI NEWS
Roundup covers DeepSeek V4.1 Flash, AlphaGenome Atlas, YuE2 and other newly released AI models and research.
HUGE AI NEWS: Deepseek v4.1 Flash, AlphaGenome Atlas, AuK, YuE2, Navier Stokes solved, Lingbot World 2 & more #ai #ainews #aitools #aivideo #agi #singularity Thanks to our sponsor Higgsfield. Try it today https://higgsfield.ai/s/theaisearch-MAFGvE Marigold v2 https://huggingface.co/spaces/huawei-bayerlab/marigold-v2-web Unimate https://linzhanmou.com/unimate/ AlphaGenome Atlas https://blog.google/innovation-and-ai/models-and-research/google-deepmind/alphagenome-atlas/ Lingbot World 2 https:…
- I Ran A 27B Model On A Hand-Sized PC… Didn't Expect This
Hands-on test runs a 27B-parameter model on a hand-sized PC, probing practical local inference constraints.
I Ran a 27B Model on This Tiny PC. I Was Fixing the Wrong Thing. 🔹 Try out ChatLLM - http://chatllm.abacus.ai/ltf and Abacus AI DeepAgent - http://deepagent.abacus.ai/ltf 💻 Run Windows on a Mac: https://prf.hn/click/camref:1100libNI (affiliate) Use COUPON: ZISKIND10 🛒 Gear Links 🛒 🎧⚡ The mini plus dock deal: https://link.amazon/B04w9tltf 🛠️🚀 The little dock: https://link.amazon/B0g6vWx6K ⚡ Thunderbolt 5 cable: https://amzn.to/40u9M3b 🖥️ Thunderbolt 5 dock: https://amzn.to/4fsyN35 🎧⚡…
Reddit · 10
- Iran used Claude to target US Navy in Middle East, Anthropic says
Anthropic says Iran used Claude in operations targeting the US Navy in the Middle East.
- Anthropic tells investors it will be profitable for second straight quarter
Anthropic reportedly told investors it expects a second consecutive profitable quarter.
- House Speaker Johnson says Congress shouldn't lead AI safety efforts
House Speaker Johnson says Congress should not lead US AI-safety efforts.
- Obama urges Democrats to have a ‘clear plan’ for AI safeguards
Obama urges Democrats to develop a clear plan for AI safeguards.
- Trump rejects call by CEOs of Anthropic, OpenAI and xAI to slow AI down: 'Whoever wins with AI wins'
Trump rejects frontier-lab CEOs' call to slow AI, framing AI leadership as a winner-take-all race.
- Bernie Sanders proposes 20 year prison sentence for AI devs who plow ahead with Artificial Superintelligence plans - penalty on par with illegally developing rogue nuclear weapons
Bernie Sanders reportedly proposes prison penalties for developers pursuing artificial superintelligence.
News · 30
- Source: Anthropic signed a $13.7B, six-year deal to lease compute from Rum Group's upcoming Georgia data center; Rum operates Rumble and hosts Truth Social (The Information)
Anthropic reportedly signed a $13.7B, six-year compute lease for Rum Group's planned Georgia data center.
The Information : Source: Anthropic signed a $13.7B, six-year deal to lease compute from Rum Group's upcoming Georgia data center; Rum operates Rumble and hosts Truth Social — Anthropic has signed a computing deal worth $13.7 billion over six years with Rum Group, a company with roots in social media …
- Current and former NSA officials: the US NSA plans major restructuring to form five new units for AI, China, cybersecurity, warfighting, and global intelligence (Washington Post)
NSA reportedly plans a major reorganization into five units, including dedicated AI and China divisions.
Washington Post : Current and former NSA officials: the US NSA plans major restructuring to form five new units for AI, China, cybersecurity, warfighting, and global intelligence — Gen. Joshua M. Rudd is retooling the massive spy agency to focus on priorities including AI, China and cybersecurity. — Summary
- In its first statement on AI, China's Ministry of State Security warns AI poses risks to the nation's political and social security, including cyber defenses (Eleanor Olcott/Financial Times)
China's Ministry of State Security issued its first AI statement, flagging political, social, and cyber-defense risks.
Eleanor Olcott / Financial Times : In its first statement on AI, China's Ministry of State Security warns AI poses risks to the nation's political and social security, including cyber defenses — Beijing signals growing concern over how rapid advances in technology could be used against the country
- Z.ai says it plans to raise $5B+ via equity and debt to fund its AI investments, including ~$2B through a share placement, after raising $4B in July (P.R. Venkat/Wall Street Journal)
Z.ai plans to raise more than $5B in equity and debt for AI investments after a $4B July raise.
P.R. Venkat / Wall Street Journal : Z.ai says it plans to raise $5B+ via equity and debt to fund its AI investments, including ~$2B through a share placement, after raising $4B in July — The company plans to raise $2.0 billion through a share placement and another $3.01 billion via zero-coupon convertible bonds
The rest of the sweep — 78 items
- China's Foreign Ministry criticizes warnings about AI risks, saying "fearmongering, confrontation, and vicious competition ... serve the interests of no one" (Bloomberg)
Bloomberg : China's Foreign Ministry criticizes warnings about AI risks, saying “fearmongering, confrontation, and vicious competition ... serve the interests of no one” — China criticized dire warnings about AI's risks after leading US technology executives backed calls to slow the development …
