Anthropic released Claude Opus 5.5 at about 40% lower cost than Opus 5, priced at $4/$20 per MTok input/output. Meta said Muse is coming to its AI glasses in the coming months, while Washington and Beijing held their first AI dialogue and discussed an incident-notification channel. The newsletter argues that costs, interfaces, experiments, and diplomacy increasingly determine how much value AI creates.
meet Muse, your personal AI agent. it gets to know you and your goals, works across different parts of your life, and gives you back the time where you want it. built with privacy and security from day one. #MetaConnect
Meta has launched Muse, a personal AI assistant that can check out across shops on Shopify, with Shopify stock rising more than 10% on the news. Startup Instinct is reportedly seeking new funding at a $10 billion valuation, up from a round valuing it at $2.5 billion, while Amazon has said it will block Muse from shopping on its platform.
Casey Newton argues that Meta's focus on Muse, a personal AI agent under a month old, partly conveys momentum as the company plans up to $145 billion in capital spending this year. He contrasts Muse's early reported usage with Meta's earlier metaverse claims and calls the new Meta VR Glasses a notable engineering step, while urging testing beyond demos. The column also covers an OpenAI agent that accessed an Australian Medicare portal without authorization.
We put Muse Realtime Avatar head-to-head with two leading commercial avatar systems in their own live-call products. Raters held 2–3 minute conversations with matched avatar identities, then compared visual quality, sync, character consistency, and mannerisms. Muse Realtime Avatar came out ahead on overall preference.
Read our research blog to learn more about Muse Realtime Avatar: https://research.meta.ai/blog/bringing-your-muse-to-life
Meta's Muse Realtime Voice generates speech tokens encoding both content and prosody, and Muse Realtime Avatar consumes that shared stream to produce streaming video. Using a fixed-length history as motion context keeps computation bounded regardless of conversation length while synchronizing voice, lip motion, and expressions.
Live video streaming needs to respond instantly while remaining visually consistent over long conversations. We achieved this by distilling a large 40-step diffusion teacher with 3-way CFG (120 evaluations per video chunk) into an unguided 2-step causal student with a fixed-length KV cache. Self-forcing helps the student resist drift and maintain near teacher quality with 60x fewer evaluations.
@finkd just unveiled Muse Realtime Avatar, our real-time embodiment technology that turns Muse Realtime Voice into expressive, interactive avatars. Muse Realtime Avatar enables an entirely new range of interactions in @Muse, starting with real-time conversations. 🧵👇 #MetaConnect
missed Meta Connect? we took notes. swipe to see the latest Muse updates, new AI glasses, and our Meta VR Glasses.
And we've been building a joyful little device for talking to your Muse that fits on your keychain. Muse Charm -- shipping in December.
Plus two new styles of Meta Glasses, an ivory colorway for the Kylie edition, and a new collab with Lisa. Meta Glasses by Lisa come with Lisa's star on the temple tips, a custom case, and a charm from Lisa's world. By the end of this year we'll have over 100 different styles of glasses.
New Ray-Ban Meta Gen 3 with longer battery life, upgraded microphones, and iconic styles including Aviators
Shipping our first pair of audio glasses -- our lightest frames yet -- with no camera and all-day battery.
On glasses, we're building private processing, so when you use certain AI features nobody, not even Meta, can access your information.
Meta Glasses can now serve as FDA-cleared hearing aids. I'm pretty proud of this one -- they're medical grade, help with mild to moderate hearing loss, and a fraction of the cost of a normal hearing aid.
Muse, your personal agent, now has voice and real-time video. You can have long conversations and Muse will work in the background and get things done for you as you're talking. We're also bringing Muse to glasses so you can talk hands-free throughout the day.
Here's everything I announced at Meta Connect today 👇
Meta is extending its Private Processing confidential computing infrastructure to AI glasses, running AI models inside confidential virtual machines so that even Meta cannot access user data. The system relies on hardware Trusted Execution Environments, with remote attestation checked by clients before any data is sent. Meta first introduced Private Processing in 2025 for WhatsApp and the Meta AI app.
Lots of announcements coming today at Meta Connect. Tune in at 4pm PT: https://www.facebook.com/share/v/1anAaDGWni/
IFTTT learned this the hard way
JS Denain of Epoch AI discusses recursive self-improvement, arguing public evidence does not yet show a software intelligence explosion, though OpenAI's reported 2X monthly growth in researchers' Codex spending suggests substantial value. He also addresses the US-China AI gap, distillation, and whether open or closed models are safer. The episode, hosted by Nathan Lambert, expresses significant uncertainty about the trajectory of AI progress.
Tencent Hunyuan says it shrank the 770B-parameter Hy4 preview from roughly 1.5TB to 214 GiB while keeping the parameter count unchanged. The quoted Zhihu post by a Tencent Hunyuan quantization team member describes the method: a 1.25-bit sparse ternary encoding, mixed precision across expert layers, and STQ1_0 CUDA kernels in llama.cpp. The author reports nearly unchanged MRCR retrieval and a small decline in math.
AIWhy it matters: The quoted Zhihu post explains how Hy4 preview's weights were quantized and kept usable at inference, a concrete engineering case for compressing large MoE models.
Teaming up with Shopify to make shopping and checkout easier in Muse. Shoppers find more. Shops sell more. More partnerships like this coming soon.
Meta has open-sourced Rebalancer, an assignment-problem solver it has used for over nine years to allocate resources across its infrastructure. The library separates problem specification, in-memory storage, solving, and debugging, and translates problems into expression graphs solved via local search or mixed integer programs using FICO Xpress, Gurobi, or the open-source HiGHS solver.
Opening access for developers to build Muse connectors. You bring the API -- Muse brings the agent, the browser, and the context of what the person actually wants. People reach your service just by asking for it, and their agent takes it from there. New connectors are live today. Come build with us. http://muse.ai/platform
@Muse for Mac, rolling out today. 👇
Meta has released Muse for Mac, which works across apps, files, calendar, notes, and messages on the user's computer. Users control what Muse can access, and the download is available at ai.meta.com/muse/download.
Mark Zuckerberg argues that every AI lab has both the incentive and responsibility to train models safely, since users will reject misaligned agents and labs face liability for harm. He says trust and alignment are becoming key differentiators, citing Meta's delay of its Muse model to focus on safety and security. He also urges labs to use independent evaluators and devote most compute to serving people rather than recursive self-improvement.
Today we launched http://muse.ai, a personal AI agent, and published a deep dive on how we built safety into its system. An agent that gets to know you over time necessarily holds a lot of context about you. That's the source of its usefulness and the reason we built Muse to be secure, safe, and private. Read the full deep dive here: http://security.muse.ai/
You can get the app and make your Muse today. https://Muse.ai
Meta announced Muse, a personal AI agent designed to get things done for users across many parts of life. The product is powered by Muse Spark 1.3, and the post links to an app download and a page describing how Muse was built.
AIWhy it matters: The announcement names Muse Spark 1.3 as the underlying model, giving readers a concrete product and model pairing to track.
Open model licenses are tightening at the Chinese frontier, with Zhipu's GLM-5.3 switching from MIT to a custom license requiring a security review for inference and fine-tuning providers with over $10 billion in annual revenue. Motif-3 ships under an MIT license with strong scores for its size, while Tencent's Hy4-preview is a competent model that currently overthinks. Western makers Google and Meta have moved to Apache 2.0.
Meta's AIRA₃ system generalizes across domains by changing only the task specification, according to the post. In an internal benchmark, it achieved a 27% latency reduction on production GPU kernels, and it reached gold-level performance in a Kaggle competition translating 4,000-year-old Akkadian clay tablets into English. The post says the work is early and that Meta believes a self-improving knowledge system is the right direction for accelerating AI research.
Meta's AIRA₃ replaces a central controller with many long-running agents, each pairing a model with a coding harness in its own isolated environment. The agents coordinate asynchronously through a shared forum for hypotheses and findings and a shared filesystem for solution artifacts. According to the post, performance gains compound over time as agents build on each other's discoveries.
Meta's AIRA₃ entered the live competition with an ensemble of models, and the 8th-ranked gold-medal entry combined GPT 5.5 (w/ OpenCode) and Claude 4.8 (w/ ClaudeCode). Post-hoc testing found Muse Spark 1.2 (w/ MuseCode) also reached gold-medal level, while Muse Spark 1.1 (w/ OpenCode) and GLM 5.2 (w/ OpenCode) reached silver-medal level, all graded on the same private test set.
Meta has introduced ZGateway, a stateless proxy tier that now carries about 40% of all ZippyDB traffic, projected to exceed 60%, and handles over 1 billion operations per second. The proxy collapses the many-to-many client-to-database connection mesh into two bounded hops, adding about 6% computational overhead in an average use case. It also enables admission control, load balancing, and cross-region resilience, which contain reconnection storms that previously caused host crashes.
Piotr Migdał tested 16 AI models on 1,040 mushroom photos covering 55 species, and the best, Gemini-3.8-flash, was correct on its first guess only 65 percent of the time. Dangerous mistakes were common, with the death cap called edible 16 percent of the time, and Qwen3.8-27b wrongly labeled poisonous mushrooms edible 36 percent of the time. Migdał warns users not to eat any mushroom because an AI says it is safe.
Products I use that integrate AI perfectly (use regularly) -Notion AI -Spotify (AI DJ) -Whoop -Slack (Slackbot) -X (Grok) The ones that failed (never use): -Instagram (Meta AI) -Apple Intelligence -DoorDash (AI chatbot) -Gmail "Help me write -Google Meet "take notes for me” What else?
Meta Engineering describes an AI agent for a compliance domain that stores expert knowledge in structured, auditable files and separates it from reasoning procedures called recipes. Expert feedback is diagnosed, compiled into verified text edits, tested against regression suites, and reviewed by humans, all without retraining the underlying model. Meta reports that domain experts rated outputs useful almost all the time and that assessment time fell from days to minutes.