Saurabh Baji

Sr Director @AIatMeta Meta Superintelligence Labs; Prev: CTO @CohereAI, GM/VP Product + Eng + AI/ML @Unity, @Quantcast, @awscloud / EMR, Athena. Tweets my own.

[Roberto Nickson @rpnickson]

Meta just launched Incognito Chat with Meta AI - the world's first truly private way to chat with AI.

But I had a lot of questions.

How does it actually work? Isn't this contradictory to Meta's business model? What assurances do we have that this is airtight?

I had a chance to speak to the Head of Whatsapp @wcathcart and Meta's VP of AI Products @vishalshahis to learn more.

[Shreyas Doshi @shreyas]

⏰Last call: add yourself to the priority list for my upcoming 1-day workshops.

The first workshop will be on Advanced Product Taste, with in-depth material on what it takes to build & exercise Taste. With lots of practical exercises & examples.

Link: https://t.co/kRx2uh0MVJ

[Saurabh Baji @sbaji]

@_rockt @Recursive_SI Congratulations Tim!

[Sara Hooker @sarahookr]

Most model trainings have failed outside of frontier labs.

Even inside frontier labs, knowing how to train for very different capabilities is often a matter of taste.

Today, we introduce AutoScientist by @adaption_ai which sets out to change that.

[Sudip Roy @sudip_r0y]

Less than 1,000 people know how to shape a frontier model.

AutoScientist is our attempt to change that. Describe the outcome. It automates the rest.

[Saurabh Baji @sbaji]

@anikembhavi @ayirpelle @Meta Welcome Ani!

[Alexandr Wang @alexandr_wang]

we launched some muse spark updates yesterday, including muse spark voice and live AI w your camera in Meta AI app + muse spark rolling out to glasses 😎

check them out!

[Meta Newsroom @MetaNewsroom]

Today we’re introducing Meta AI Voice Conversations powered by Muse Spark that let you talk naturally to Meta AI (interrupt, switch topics, or swap languages), and as you talk, Meta AI can generate images and pull up recommendations from Reels, maps, and more. We’re also bringing live AI to the app, so you can point your camera at the world and ask about what you’re seeing in real time.

https://t.co/A5JJOD1lyc

[Jack Rae @jack_w_rae]

Nice debut from Muse Spark in the Agentic Coding arena.

Ranking ahead of GPT 5.4 and Gemini models, and behind the Opus series + GLM 5.1.

[Alexandr Wang @alexandr_wang]

Muse Spark is #3 on ClawEval, ahead of GPT-5.4 and Gemini 3.1 Pro. It is honestly a surprisingly agentic model.

https://t.co/CAJJ65G7Rx

[Saurabh Baji @sbaji]

I have been a @shreyas subscriber for a long time, and been curious about taking the Product Sense course. Finally bit the bullet a little while back - a very satisfying couple weekends filled with insights and aha moments, and way more practical use that I had imagined. Loved the cohort too. Take it now, you will not regret it!

[Ryan Ehrlich @ryansehrlich]

Excited to share what we've been working on at MSL

[Dan Robinson @danrobinson]

This model did as well as Opus 4.6 on my private cryptography research eval and better than Gemini and Grok—impressed to see Meta catching up here

[Scale AI @scale_AI]

Breaking: @AIatMeta just released Muse Spark — now live across @ScaleAILabs leaderboards.

Here’s how it stacks up:

Tied for 🥇on SWE-Bench Pro
Tied for 🥇on HLE
Tied for 🥇on MCP Atlas
Tied for 🥇on PR Bench - Legal

Tied for 🥈on SWE Atlas Test Writing
🥈on PR Bench - Finance

🥉on SWE Atlas QnA

[Saurabh Baji @sbaji]

This is one of my absolute favorite ways to use the model!

[Matt Deitke @mattdeitke]

Grounding has shipped to prod. It's great for understanding explaining and understanding things better!

[Matt Deitke @mattdeitke]

Excited to share Muse Spark! It's a strong natively multimodal model with many surprising properties that emerged. Here, the model is able to use Python tools to make a playable Sudoku game on the web from an image input of the board. ✨

[Saurabh Baji @sbaji]

🚀

[Patrick Collison @patrickc]

Congratulations to Alex and the whole team at MSL. As a sucker for all things speedy (https://t.co/Od1fQoL3FQ), I thought this was an impressive chart:

[elvis @omarsar0]

NEW: Meta announces Muse Spark.

All you need to know:

* It's their new multi-modal reasoning model.

* Strong at multi-agent orchestration and multi-modal reasoning.

* Contemplating mode orchestrates multiple agents that reason in parallel. Helps to compete with models such as Gemini Deep Think and GPT Pro.

* The test-time reasoning efficiency is probably the most important bit; Muse Spark can compress its reasoning to solve problems using significantly fewer tokens (referred to as though compression). Look at the chart in the figure. Agents are parallelized without significantly increasing latency. That's huge!

Great to see Meta finally get back in the game.

It's good to see a focus on native multimodal capabilities.

No open-source release and a private API only available to a select few.

Results look promising, too, though there are some gaps, especially to enable long-horizon agentic systems and coding workflows.

[Guillermo Rauch @rauchg]

The best outcome for humanity is many strong AIs competing for the top spot.

Vercel is proudly powering https://t.co/ZsS5nRfjIF and the infrastructure that made today's model release possible.

[Rylan Schaeffer @RylanSchaeffer]

TBD & MSL's velocity has been insane. Amazed by what the team has already built, and how quickly we're accelerating!