James G. Beldock (@jbeldock) profile picture

Day Job: Meta Superintelligence Lab, formerly @ShotSpotter. Night Job: SF Bay Area Immigrant. Life Job: New Yorker and lesser half of @TatyanaDB.

San Francisco, CA
Joined March 2009
Photo by James G. Beldock (@jbeldock)Photo by James G. Beldock (@jbeldock)Photo by James G. Beldock (@jbeldock)Photo by James G. Beldock (@jbeldock)Photo by James G. Beldock (@jbeldock)Photo by James G. Beldock (@jbeldock)Photo by James G. Beldock (@jbeldock)Photo by James G. Beldock (@jbeldock)Photo by James G. Beldock (@jbeldock)Photo by James G. Beldock (@jbeldock)
James G. Beldock retweeted
Muse Wonder, Muse Spark 1.3, and Muse Ember are coming soon to Microsoft Foundry on @Azure.
375
236
43
4,439
502,448
James G. Beldock retweeted
A must have for Muse users :)
Want to take a stab at guessing how popular is muse. This Chrome extension (which they released a week ago!?) allows your muse to browser things on your own browser, has been installed 100 thousand times! (and has a 4.5 ★ rating) BTW, it is pretty great!
16
2
89
9,901
James G. Beldock retweeted
Muse is now available in Australia. Looking forward to seeing what people build and get done with @Muse. Proud of our team and the care they've put into building Muse with privacy, safety and security first. Because Muse acts for you, getting this right matters. You can read more about that work in our tech post. security.muse.ai/
Muse, brekkie, and beach, in that order. Muse has arrived in Australia: muse.ai 🇦🇺
10
6
114
5,647
James G. Beldock retweeted
We built Muse Realtime Voice and Muse Realtime Avatar to keep the voice, expressions and gestures in sync. It responds right away, and you can jump in while it’s talking. Proud of the team that built this!
Introducing Voice. Talk to your Muse and Muse talks back, so you can go from “what should I cook?” to cooking, just like that ✨
23
16
4
256
16,259
James G. Beldock retweeted
1/ starting today, muse and meta ai will run on new specialized models trained using muse wonder. ACTIVE MEMORY: models can reason across time and better remember what matters ACTIVE FOCUS: enabling muse to continue working towards a goal even when met with roadblocks
168
169
50
2,714
231,762
James G. Beldock retweeted
MUSE FEATURE ALERT: muse lil guy is coming to all meta glasses. just say his name to wake him up. coming soon!!!!
🤖 Made with AI
91
41
15
850
77,014
James G. Beldock retweeted
konnichiwa! we are launching muse in japan today! 🇯🇵 it is now available—download it from your favorite app store! (i hope you're ready for the anime memes 😇)
210
187
90
2,442
188,878
James G. Beldock retweeted
it feels way more exciting when your personal agent gets smarter vs a chatbot getting smarter. instead of “i might get better answers” you think “oh my life is about to get better. my muse can handle more things for me now”
Introducing Muse Wonder, our new family of frontier models with state of the art performance for personal and professional agentic work.
2
5
39
2,384
Meta has released Muse Wonder and Muse Wonder Full, achieving Artificial Analysis Intelligence Index v4.3 scores of 52 and 56 at max reasoning effort - each a material gain over Muse Spark 1.3 (max) at 48, with a release gap of only 6 weeks Muse Wonder is available for users on Meta’s first-party API, while Muse Wonder Full is currently only available for Muse Code users. Meta notes that each model is designed for personal and professional agent use Congratulations to @AIatMeta, @finkd, and @alexandr_wang on the release! Key Takeaways: ➤ Both models exhibit a material jump in overall Intelligence Index score. Muse Wonder Full (max) improves on Muse Spark 1.3 (max) across all ten Intelligence Index evaluations, gaining 8 points to 56. The largest gains were seen in agentic evaluations: Terminal-Bench 4.0 (+24 points, 33% to 57%), GDPval-AA v2.1 (+176 Elo, 1681 to 1857), and AA-Briefcase (+148 Elo, 1581 to 1729). Muse Wonder (max) gains 4 Intelligence Index points and lands at 52, with notable improvements vs. Muse Spark 1.3 (max) in Terminal-Bench 4.0 (33% to 52%) and AutomationBench-AA (58% to 68%) ➤ Muse Wonder Full (max) places in the top three on six of the ten Intelligence Index evaluations, and is particularly strong in agentic knowledge work and scientific reasoning. In agentic knowledge work, it narrowly leads on GDP.pdf (33%), ranks #2 on GDPval-AA v2.1 (1857 Elo), and #3 on both AA-Briefcase (1729 Elo) and AutomationBench-AA (71%). In scientific reasoning, it tops the CritPt leaderboard at 33% and ranks #3 on SciCode with 63%. This builds on the relative strengths of Muse Spark 1.3 (max), which also showed particular ability in these two domains ➤ Both models use fewer output tokens than Muse Spark 1.3 (max). Muse Wonder (max) and Muse Wonder Full (max) each use around 54k output tokens per Intelligence Index task, down ~10% from Muse Spark 1.3 (max, 60k). This usage is lower than similarly intelligent models Gemini 4 Argon (high, 62k) and Claude Fable 5.1 (max w/ fallback, 78k), but higher than GPT-6.1 Sol (max, 38k) and GPT-6 Astra (max, 27k). Muse Wonder Full (max) uses 16% fewer input tokens per Intelligence Index task than Muse Wonder (max) (6.44M vs 7.69M), meaning it is both more intelligent and more token-efficient ➤ Cost per Task for Muse Wonder (max) lands higher than most similarly intelligent models. Despite efficient output token use, premium API pricing ($4 per million input tokens, $20 per million output tokens) and high input token use (7.69M per task) push Muse Wonder (max)’s cost to $4.52 per Intelligence Index task. This is lower than Claude Fable 5.1 (max w/ fallback, $7.63), Claude Opus 5.5 (max w/ fallback, $5.98), and Claude Sonnet 5.5 (max w/ fallback, $5.46), but higher than GPT-6 Astra (max, $3.26), Gemini 4 Argon (high, $1.99), GPT-6.1 Sol (max, $0.72), and its predecessor Muse Spark 1.3 (max, $1.60). Per-token pricing is not currently available for Muse Wonder Full, so no Cost per Task figure can be estimated at this stage ➤ Both models gain significantly in Coding Agent performance. Muse Wonder Full achieves 68 on the Artificial Analysis Coding Agents Index at max effort in Muse Code, alongside Claude Code with Sonnet 5.5 (max, 68) and Opus 5.5 (max, 66). Muse Wonder (max) scores 63. These represent material gains over Muse Code with Muse Spark 1.3 (max), which scores 54 Other details for the two new models: ➤ Context windows: 1M tokens each, unchanged from Muse Spark 1.3 ➤ Pricing: for Muse Wonder, $4 input / $0.20 cached input / $20 output per 1M tokens. Pricing per token is not yet announced for Muse Wonder Full ➤ Input modalities: text, image, video, and audio ➤ Output modalities: text ➤ Availability: Muse Wonder is available on Meta's first-party API and Muse Code at launch, while Muse Wonder Full is only available to Muse Code users
158
308
179
2,680
472,918
James G. Beldock retweeted
Muse Wonder, Meta's Fable-class model, is out. It scores 56 on AA's Intelligence Index, putting it 2 points below SOTA (Opus 5.5), and overall the 2nd best model in the world per the Index!
86
90
64
2,247
267,006
James G. Beldock retweeted
This model is very very good. I was struggling to find anything to throw at it that it couldn’t do, even when other top models failed. Meta is back.
introducing muse wonder our new frontier model. state-of-the-art at personal agents and professional agents. our largest and most capable model yet—rivals the big bois
126
206
40
5,610
446,449
Big day! Muse Wonder Full is here!
We’re excited to introduce the Muse Wonder series with Muse Wonder and Muse Wonder Full, our most capable models yet. Both are frontier models, built to deliver state-of-the-art performance for personal and professional agentic tasks. Muse Wonder is available today via the Meta Model API and Muse Code. Muse Wonder Full is available today for Muse Code subscribers. Muse Wonder will also soon power experiences in our personal agent product, Muse, rolling out first to subscribers.
1
44
James G. Beldock retweeted
Alexandr Wang (Chief AI Officer at Meta) explains how each Muse runs inside its own "secure VM", a sandboxed, isolated mini computer, and the data you share with it is stored only there. > Whenever the agent tries to send anything out, like filling a form or making a call, a separate "sentinel" agent checks it and blocks sensitive data such as credit card or Social Security numbers. > Muse also asks the user to approve every new website it wants to contact or send data to. > A more encrypted "confidential VM", is coming later with the goal of WhatsApp-level guarantees, though it may be slower. ---- Full video on "Cleo Abram" YouTube channel, (link in comment)
33
27
9
255
32,049
James G. Beldock retweeted
cleo is brilliant and we one of my favorite conversations about ai safety also i have a subtle diss during the interview to some well-known peeps 🙃 you’ll have to find it!
What can Muse actually do? When can we trust it? And what should we do to make sure this goes right? The HUGE* Conversation with @alexandr_wang out now!
64
50
4
1,286
131,728
James G. Beldock retweeted
we launched muse a month ago today for the past month, reality felt like a dream. we watched the world catch fire with muse. every day we read stories of people saving money and precious time. every day it made people’s lives better in small and big ways. every day i watched the muse team pour their hearts into bringing a little more wonder in the world. the response to muse has surpassed anything i could have imagined. i know the muse team will remember it for the rest of our lives. it is easy to become jaded in tech. products launch every day but they very rarely make any dent in the world. but i have seen muse make a real difference for moms, for grandpas, for students, for coworkers. many of you using it in ways we never expected. thank you to the muse team. thank you to everyone who’s tried muse. this is only the beginning. we have so much more to build.
1/ today we're rolling out Muse, our new personal ai assistant. Muse is always-on, wicked fast, can use a browser, connect to your apps, and is designed to be secure. try it now: muse.ai.
205
116
28
2,074
174,433
James G. Beldock retweeted
4 hours since Muse Gadgets was announced - the response has been amazing. Welcome to everybody joining our Discord and hanging out. Thank you for your pull requests (we merged some already), and for sharing the gadgets you built! Some of my favs 👇
🚨 MUSE ECOSYSTEM ALERT 🚨 today we are announcing Muse Gadgets! this is an open-source ESP32 firmware and Linux SDK for anyone to make hardware that works with Muse we are also releasing our own gadget—Muse Home Link—to enable your muse to work with your smart devices (TV, speakers, etc.)
9
9
2
144
22,899
James G. Beldock retweeted
mathematicians and muse spark collaborated to solve 6 open problems in math: 1. The Strict Threshold for Gaussian Ellipsoid Fitting 2. Finite-Time Blow-Up of Radial Negative-Energy Solutions for the Mass-Critical Biharmonic Nonlinear Schrödinger Equation 3. Semiabelian Groups Need Not Be Monomial 4. Tightness of the Cycle-Based Relaxation for Completed Length-Three Alpha-Cycles 5. String Two-Point Function = Height Function on a Curve 6. On Solvable Evolution Algebras and a Conjecture by García-Martínez and Pérez-Rodríguez
Following gold-medal-level performance from our AI models across five competitions in mathematics, physics, and chemistry, we asked a harder question: can AI contribute when a problem is genuinely open and without an existing solution path? Over the past several months, mathematicians worked with Muse Spark 1.1 and Muse Spark 1.2 in Thinking Mode through the regular meta.ai chat interface, with no custom research scaffold, to find solutions to such problems. Our goal wasn't to mass-produce papers, but empower researchers. Every collaboration followed the same principles: mathematicians guided the research, a second group of mathematicians then reviewed the work, each paper marks which passages were drafted primarily by humans or AI, and each credits the prior research it builds on. Where other teams independently announced solutions to the same problems, we acknowledge their work as well. Today, we're sharing six papers from that collaboration. 🧵👇
164
208
59
3,327
488,595
James G. Beldock retweeted
Excited to open source the Muse Gadgets SDK today! You can connect your DIY electronics like ESP32 boards and Raspberry Pi to Muse. Build an e-ink morning briefing, a push-to-talk pocket device, a sysadmin for your homelab, and much more. Details at gadgets.muse.ai
🚨 side project alert 🚨 Announcing Muse Gadgets, an open source ESP32 firmware and Linux sdk so that you can make hardware devices that work with Muse. Grab an API token from gadgets.muse.ai and point your favorite coding agent at the github repo to build your own peripherals for Muse.
28
40
10
592
35,752