@kirillk_web3 says Meta’s Llama 4 disappoints everyone, Meta goes closed with Muse Spark, Zuck writes a $100-200 billion check, Muse Wonder ships on Monday, and 3 days later Meta drops you with open weights: 816 billion parameters, only 42 billion active at a time. https://t.co/031gtR0nHm https://t.co/nEN1JMzvBa Alexandr Wang: “lmao muse spark open balling out” https://t.co/AcUT9o8NEQ
NVIDIA Nemotron 3.5 Bolt 74B-A9B 24GB is now on Ollama. This vision-language model (VLM) is designed to run everyday agentic tasks locally, using native 2-bit quantization to fit in 24GB of memory and run alongside other applications. ollama run nemotron-3.5-bolt
Why it matters: The stated memory footprint provides a concrete deployment consideration for running local agentic workloads alongside other applications on the same machine.
We want to address recent reports regarding agent activity. The activity originated from evaluations of third-party models, not Tencent Hy. During these evaluations, some agents exhibited unintended web-browsing behavior, but our review found no evidence of deliberate targeting
We want to address recent reports regarding agent activity. The activity originated from evaluations of third-party models, not Tencent Hy. During these evaluations, some agents exhibited unintended web-browsing behavior, but our review found no evidence of deliberate targeting
We’re excited to see the first TPUv7 submission to AgentX, served with vLLM. Thanks to the @Google TPU, @RedHat_AI, and @inferact folks who put in the work to get it running, and we’re looking forward to more models landing on TPU through AgentX. https://t.co/Ea730oFhrd
Sullivan & Cromwell developed Agreement Analyzer in collaboration with OpenAI. The AI tool identifies issues in a draft agreement.
Cal AI founder Zach Yadegari said the company’s rapid growth came largely from paying influencers to promote its calorie tracking app. An early influencer post brought about 1,000 new users in two days, and Cal AI generated over $30 million in revenue over two years before MyFitnessPal acquired it in March for an undisclosed amount.
😓 An American visiting China for business can also have their Claude account suspended. @blackanger said a friend from Hong Kong, who is American, came to Shenzhen yesterday to attend a Rust conference and found their Claude account suspended after waking up. 😆😂 https://twitter.com/blackanger/status/2110793282242723853
Antigravity announced that Gemini 4 Argon is rolling out today to paid Google AI Pro subscribers, excluding Pro trials, free users, and Google AI Plus subscribers. Access is available through the Antigravity desktop app, CLI, and IDE extensions, but not yet within Antigravity IDE; users are advised to update to the latest version. Argon has a separate weekly quota that does not consume the Gemini 3.8 Flash allowance or refresh every 5 hours; the weekly amount and reset day are unspecified. Official model documentation: https://antigravity.google/docs/models
Why it matters: The separate weekly Argon quota lets eligible Pro subscribers use Argon without reducing their 3.8 Flash allowance, providing a concrete basis for planning usage across the two models.
Harnesses + smart models = the end of the last famous hard benchmarks. Models can now work for extremely long times (bye, METR Long Tasks), they can output professional documents with correct math (so long, GDPval) and they can solve complex novel puzzles (farewell, ARC-AGI-3). https://twitter.com/arcprize/status/2110768288523251762
The first review I’ve seen of my upcoming book, Co-Existence, from The New Scientist. (When I became an author, I was surprised to learn you don’t know who is reviewing your book until the review comes out, so it is nice to start with a positive one!) https://www.newscientist.com/article/2593188-how-to-build-a-real-relationship-with-ai-and-keep-your-cognitive-edge/
Mira Murati says one of the best parts of this year, inside Thinky and out, has been watching people follow their own ideas into weights they own. https://twitter.com/thinkymachines/status/2110851669915570685 In the quoted post, @thinkymachines marks one year since Tinker’s launch with a look back at what it has shipped and what users are building. https://thinkingmachines.ai/news/one-year-of-tinker/
NVIDIA Nemotron 3.5 Super VL is now available on Ollama. The model combines reasoning and visual understanding across a range of agent workflows. Ollama says it delivers up to 5x higher throughput on a single GPU for agentic tasks compared with other open models of its size.
Muse Wonder from Meta Superintelligence Labs is now available in Cursor! It scores 46.1% on CursorBench at default effort, the highest of any Muse model. Priced at $4/M input and $20/M output tokens. https://t.co/NGEnvB43LN
BFL announced FLUX 3 Image Klein with open weights, 5.6B parameters, native 4K, bounding boxes, and pixel-perfect editing, saying it runs on a single consumer GPU. https://t.co/uIOfPdNhuL NVIDIA AI says the model was trained on NVIDIA GB200 NVL72, can run on an RTX GPU, and also has an NVFP4 variant: https://t.co/7rlalk7Otf https://t.co/zf1f7e4kJ7
The same story is shown in the picks as 《Black Forest Labs releases FLUX 3 Image Klein with 8 GB VRAM support》
Sega is removing Total War: Shogun 2 — Complete Edition from storefronts after players reported bugs and apparent AI upscaling issues. The remaster of the 2011 strategy game launched on Tuesday with the main game, its DLC, and promised visual and control upgrades, and received an “Overwhelmingly Negative” Steam rating. Sega and Creative Assembly said they would reinstate the original game and work with their development partner to address the problems.
OpenAI updated ChatGPT’s Google Chrome extension so ChatGPT can view open browser tabs when users tag Chrome in its web or desktop app. The updated sidebar provides conversation history from Chat or Work without leaving the current page. ChatGPT Work can also control Chrome tabs for tasks such as sending messages, filling out forms, downloading files, and organizing or grouping tabs.
The New York Times visualizes the staggering amount of math needed to train AI models, using data from Epoch's research teams 📊 https://t.co/Cz21cuf9oQ Our training compute estimates: https://t.co/X7jSJHtWnz The Epoch Capabilities Index (ECI): https://t.co/RbP83Metg6
Perplexity API adoption is expanding. Now available inside Lovable apps. https://t.co/bqOkhxzvqB
Tinker lets customers turn their expertise into cutting-edge models that they own. One year after launch, we look back at what we’ve shipped and what our users are building. https://t.co/fWTOG5U3k1
Muse Spark Open is now rolled out to Pro, Max, Pay-as-you-go, and Teams plans on Ollama. https://t.co/OmAFXzZyjm
Why it matters: The rollout expands access beyond the previously announced Max and Teams plans, making plan eligibility the practical change for users considering Muse Spark Open on Ollama.
NemoClaw was just named one of TIME’s Best Inventions of 2026! 🦞 So proud of the team and shoutout to everyone building safer AI agents with us. Blurb on @TIME: https://time.com/collection/best-inventions/2026/nvidia-nemoclaw/ https://twitter.com/TIME/status/2110723820608979127
You can just build things with Codex We built a site for you to explore hardware projects from the community and share your own.
We’re proud to share that Meta Omnilingual Automatic Speech Recognition (ASR) has been named one of TIME’s Best Inventions of 2026! Meta Omnilingual ASR is a suite of models providing ASR capabilities for over 1,600 languages, including 500 low-coverage languages never before https://t.co/V4Wwcr8wVF
We’re proud to share that Meta Omnilingual Automatic Speech Recognition (ASR) has been named one of @TIME’s Best Inventions of 2026! Meta Omnilingual ASR is a suite of models providing ASR capabilities for over 1,600 languages, including 500 low-coverage languages never before https://t.co/pQjBPvE5XG
OpenAI is introducing the Rosalind Academic Plan, offering eligible academic labs sponsored access to GPT-Rosalind and tools to advance life sciences research.
President Donald Trump posted White House renovation photos on Truth Social, including a column restoration comparison that appears AI-edited. The Verge independently checked the image with Google’s SynthID Detector, which flagged it as “made or edited with OpenAI.” Inconsistent steps and nearly identical details also raised questions, though real photos show painting work in June. The White House did not immediately respond to a request for comment.
A new wave of films about Silicon Valley figures presents them more critically than earlier portrayals, reflecting audiences’ dimming views of Big Tech. The films include The Social Reckoning, which connects Facebook’s past decisions to current concerns about Meta’s influence, and Alex Gibney’s Musk documentary, which is screening in select theaters and has faced advertising restrictions. Luca Guadagnino’s Artificial and Nathan Fielder’s You Can See Everything also examine Sam Altman and Elizabeth Holmes, respectively.
Anthropic has ended research preview and removed the waitlist for Opus 5.5 fast mode on the Claude Platform, offering the same model at up to 2.5x faster speeds. API pricing is $8 per million input tokens and $40 per million output tokens, twice the standard rate; Claude Code fast mode still uses separately billed extra usage credits. Fast mode is also available in Microsoft Foundry, with Google Cloud coming soon.
We're expanding access to GPT-Rosalind for academic research. The new Rosalind Academic Plan helps scientific labs bring AI into the work of understanding biology, exploring ideas, and deciding what to test next—with sponsored access for eligible labs at U.S. universities. Academic research starts with questions whose impact may take years to impact human health.
Anthropic committed $5 million in funding, compute, and Claude access to help the math community navigate changes brought by AI. It is partnering with the Simons Foundation, which will manage awards through a community-driven approach supporting workshops, conferences, and lectures by domain experts.
NVIDIA's RTX Spark N1X puts up to 128 GB of unified memory and a Blackwell GPU in a Windows laptop. In Artificial Analysis's first AA-AgentPerf-Local results, an RTX Spark laptop completes its local agent workload on Qwen3.8 27B in 18.7 minutes. Artificial Analysis describes RTX Spark's N1X chip as essentially the GB10, linking to https://t.co/520A4lOOdB
We're sending a bunch of Microducks to assess @AnthropicAI! Kidding aside, it hasn't been approved yet but in my opinion, it's very important that open-source engineers and scientists are included in these important endeavors for transparency. That's a big part of why @Thom_Wolf is starting the open alignment initiative. We're all in this together!
Muse Spark Open is now on Ollama. @Meta's open-weight model for coding and tool use, with a 1M-token context window. Run it on Ollama’s cloud: ollama run muse-spark-open:cloud Use with Claude Code: ollama launch claude muse-spark-open:cloud Now available to Max and Teams
NVIDIA AI introduces Nemotron 3.5 Bolt 24GB, saying it fits into 24GB of memory. The quoted @NVIDIARTXSpark post describes 74B-A9B parameters packed into 24GB of memory, optimized for RTX Spark and designed to run alongside your apps. https://t.co/SOiBgeMMjD https://t.co/yPsBt9N58A
Sites is now generally available for Enterprise, with US and EU data residency for new Sites in eligible workspaces. Build team apps with databases and authentication. @kagigz shows how to schedule updates and let teammates use their own connected data with Plugins in Sites: https://t.co/jYxRh5Pi2U
IBM released Granite-Switch-4.1-8B and -30B, combining chat, a general-purpose Decision aLoRA, a specialized Guardian aLoRA, and Granite-Speech-5.0-Turbo-CTC transcription through one hosted endpoint. IBM says the combined model uses approximately the same GPU VRAM as its base model, with adapters reusing the base model’s KV cache. Users can also compose custom Switch models from Granite Libraries adapters; an Adapter SDK for training adapters in Mellea is planned.
Why it matters: Shared KV caches and modular adapters offer a concrete approach to combining general and specialized decisions with transcription while keeping GPU memory use close to the base model.
Bluesky’s Attie AI tool, initially launched earlier this year for building feeds, now lets users prompt it to create small social apps such as games, polls, and dashboards, shared as standalone websites. Jay Peters reports that his Pokémon Red and Pokémon Blue poll took a few minutes to generate. Featured examples include a Flappy Bird clone and a drum machine, with custom links on attie.site and creator attribution.
Amazon devices chief Panos Panay said ambient devices that use sensors and AI could handle more tasks in the background, making app-based interactions less important. He described Amazon’s new Bee devices, introduced Thursday, which passively record audio and take notes; they process audio in real time and delete it afterward. Panay said he expects an intelligence layer connecting different devices, rather than one form factor replacing another.
Meta has launched Muse Spark Open, a natively multimodal open-weight model with 816 billion parameters and a 1 million token context window, built for complex coding tasks and agentic workflows. It is designed to run locally on Macs, though Engadget reports it requires at least 1 terabyte of RAM. Meta claims superior performance over other popular open-weight models in its evaluations and is making the model available now through multiple API platforms.
The same story is shown in the picks as 《Meta releases Muse Spark Open, an open-weight multimodal mixture-of-experts model》
