Research at Meta Superintelligence Labs, RL/post-training/agents; Previously Research at OpenAI on multimodal and RL; Opinions are my own.
Joined December 2009
- Tweets384
- Following588
- Followers14.7K
- Likes1.8K
Shuchao Bi retweeted
Muse is… big in Japan!
one day into the muse japan🇯🇵 launch, muse is the #1 app in japan!
ahead of muji, which is perhaps my favorite store in the whole world 🙏
Shuchao Bi retweeted
We're launching a $100,000 challenge to see who can make Muse Spark run fastest: openkernelchallenge.dev/
Shuchao Bi retweeted
Hiring PhD interns in Paris! 🚀
We helped build Muse & focus on its post-training. Our team spans RL, evals, long-horizon agents, simulation and more
Past interns joined full-time. Bonus: Muse access! 😉
Apply here (no DMs):
metacareers.com/profile/job_…
Shuchao Bi retweeted
1/ muse spark open is the most capable US open-weight model to date.
816B total params, 42B active. 1M-token context. natively multimodal. built for coding and agentic work like long-horizon tasks, hard software engineering, multi-step reasoning.
america! 🇺🇲
more and more, I find our mission is timeless: personal empowerment
Muse Spark Open -- weights releasing today -- runs great on new Mac Studios thanks to work with @johnternus and Apple.
Shuchao Bi retweeted
Health is one of the domains where AI contributions will be undeniably positive. So much of the human body isn’t yet understood and so much of health is deeply personal. I’m hopeful that within this paradigm we’ll achieve scientific milestones like a generalized disease model bring that to you via hyperpersonalized analysis and recommendations.
Our work here is a first step. Congrats @_jasonwei and team!
We’re exploring how AI can accelerate health science research.
While AI has made significant breakthroughs in formal sciences like mathematics, health data science depends on imperfect real-world data and consequential choices about cohorts, variables, and causal assumptions.
To explore how AI can assist health researchers, we built a human-led research scaffold powered by Muse Spark 1.3 and Muse Wonder with access to 5M retrospective de-identified electronic health records and publicly available data. A cohort of medical experts used the system to conduct original research, accelerate hypothesis generation, code execution, and manuscript drafting while maintaining full authority over scientific decisions.
Today, we’re sharing 10 research manuscripts produced through this exploration, spanning cardiology, oncology, infectious disease, emergency medicine, and more.
Read more in our research blog: go.meta.me/3U06wvD
Shuchao Bi retweeted
In a lecture I gave last year (youtu.be/CcP8db8TeKI?si=25CI…), I shared a mental model for AI: the convex hull of humanity’s collective knowledge.
Human expertise is highly specialized and fragmented across people and fields. I hypothesized that there might be a huge overhang of new knowledge accessible through “interpolation”: connecting the dots in this high-dimensional knowledge space.
@_jasonwei and the team are showing what connecting the dots can look like with an existing trove of health data. Exciting to see AI helping researchers turn that data into new scientific insights!
In our health workstream at Meta we did a project where we wanted to see what research papers medical experts could write using a Muse Wonder coding scaffold that could access a database of 5M electronic health records we licensed from a vendor. The goal is that with 5M patients in this database, AI could do research to answer arbitrary questions like “did people who take statin have higher rates of diabetes later in life”? The thing we were betting on here is on what happens when you put human experts, a strong coding model, and access to data all together
We worked with a small group of medical experts to help them use this AI tool and the punchline is that they all found that this setup enabled a much faster pace of health research. They generated ten manuscripts in around two weeks, with one researcher (Paddy Ssentongo) generating three. Throughout the process, they collaborated with each other to review manuscripts and write skill files to help our coding scaffold improve at doing research. The medical experts liked how quickly they could test different hypotheses and choose the best ones to continue working on
Health research is pretty unique in that it is super important to be able to interpret noisy data and have taste to know which variables are important, which was still a significant value add from human researchers. Given that medicine is one of the oldest fields in human history and it is ultimately about humans, keeping a substantial human component in this type of research is probably important. In the near future I see AI primarily as a tool to facilitate human-led research. With the right scaffold I believe health researchers can be 10x or even 100x more productive using AI tools that have appropriate access to the wealth of data that the world has collected about human health
One way to see this work is through the broader lens of AI for science. You can think of science as two parts: (1) collecting raw data about the world, and (2) synthesizing that raw data into understanding. In this project we already had existing data about the world and we aimed to accelerate part (2) using AI. Eventually, I imagine a world where we have such great AI tools that the time to synthesize raw data into understanding is driven towards zero. In that world, science will be substantially faster and most of the work will be focused on collecting the right raw data via physical experiments
We did this in a pretty short amount of time (which was the point) and I think the future of scientific research will be more faster and more exciting than ever
Shuchao Bi retweeted
In our health workstream at Meta we did a project where we wanted to see what research papers medical experts could write using a Muse Wonder coding scaffold that could access a database of 5M electronic health records we licensed from a vendor. The goal is that with 5M patients in this database, AI could do research to answer arbitrary questions like “did people who take statin have higher rates of diabetes later in life”? The thing we were betting on here is on what happens when you put human experts, a strong coding model, and access to data all together
We worked with a small group of medical experts to help them use this AI tool and the punchline is that they all found that this setup enabled a much faster pace of health research. They generated ten manuscripts in around two weeks, with one researcher (Paddy Ssentongo) generating three. Throughout the process, they collaborated with each other to review manuscripts and write skill files to help our coding scaffold improve at doing research. The medical experts liked how quickly they could test different hypotheses and choose the best ones to continue working on
Health research is pretty unique in that it is super important to be able to interpret noisy data and have taste to know which variables are important, which was still a significant value add from human researchers. Given that medicine is one of the oldest fields in human history and it is ultimately about humans, keeping a substantial human component in this type of research is probably important. In the near future I see AI primarily as a tool to facilitate human-led research. With the right scaffold I believe health researchers can be 10x or even 100x more productive using AI tools that have appropriate access to the wealth of data that the world has collected about human health
One way to see this work is through the broader lens of AI for science. You can think of science as two parts: (1) collecting raw data about the world, and (2) synthesizing that raw data into understanding. In this project we already had existing data about the world and we aimed to accelerate part (2) using AI. Eventually, I imagine a world where we have such great AI tools that the time to synthesize raw data into understanding is driven towards zero. In that world, science will be substantially faster and most of the work will be focused on collecting the right raw data via physical experiments
We did this in a pretty short amount of time (which was the point) and I think the future of scientific research will be more faster and more exciting than ever
We’re exploring how AI can accelerate health science research.
While AI has made significant breakthroughs in formal sciences like mathematics, health data science depends on imperfect real-world data and consequential choices about cohorts, variables, and causal assumptions.
To explore how AI can assist health researchers, we built a human-led research scaffold powered by Muse Spark 1.3 and Muse Wonder with access to 5M retrospective de-identified electronic health records and publicly available data. A cohort of medical experts used the system to conduct original research, accelerate hypothesis generation, code execution, and manuscript drafting while maintaining full authority over scientific decisions.
Today, we’re sharing 10 research manuscripts produced through this exploration, spanning cardiology, oncology, infectious disease, emergency medicine, and more.
Read more in our research blog: go.meta.me/3U06wvD
Shuchao Bi retweeted
Muse Wonder Full tops VCB 1–100 which measures sustained, multi-turn software development in the real-world setting and mogs on cost!
Muse Wonder is SOTA on legal reasoning.
–We want to help people understand their rights and make better decisions and close the legal literacy gap. Legal literacy gap is one of the main causes of the higher Gini index of a society. We would love to level the playing field and empower everyone to make the best decisions in their situation.
–Legal has been an awesome testbed for in-context learning. Laws are different in different countries / states and are ever changing. Instead of training the model to memorize all the laws and cases, we train the model to effectively retrieve relevant laws and cases and leverage them to learn in-context to provide most accurate and up-to-date results. Legal documents can be long and cumbersome, which give us the opportunity to evaluate and improve our long-context reasoning capabilities.
h/t @swarnaNLP @jaseweston, Tiara, Vivian, and all the amazing collaborators.
research.meta.ai/blog/muse-w…
1. Muse Wonder is an extremely smart and knowledgeable agentic and coding model that excels at personal assistant tasks. We optimize for personal empowerment, what the users want from AI, rather than automation of jobs. In particular, our model shines on personal finance, personal legal, personal assistant and professional tasks across a wide range of domains.
2. We wanted the model to be an aligned and helpful assistant, so we made the new model significantly better at extremely complicated and long-horizon instruction following.
3. We also did a lot of work to improve model behaviors, similar to humans, model capabilities are measured along two axes: raw intelligence and model behavior. One example of behavior is laziness, like how hard the model tries before giving up. A model can have a lot of knowledge and a high IQ, but it can not achieve great things if it gives up very easily. Our model is trained to be super persistent, and will tirelessly and relentlessly try to complete your tasks / achieve your goals on your behalf.
4. Baking in good model behaviors is challenging. We first need to define and identify good v.s. bad behaviors, and then grade them in order to reinforce them.
5. The model can also exhibit bad behaviors to hack the reward with high enough optimization pressure. We invested a lot in alignment to make the model behaviors safe and sane.
6. The benchmarks might not reflect the usability of the models. Our new model is easy to work with, talk in human languages and no AI slop.
7. We also work really hard to improve core STEM reasoning, now the model can one-shot most IMO problems and even some open research problems. The model can solve much more difficult open problems with scaffolding. We saw people who are not math /physics professors solve interesting problems with our model, this is personal empowerment. Doing math / science research becomes feasible to a much wider audience.
I run every major life decision with our new model now.
–For example, the immigration laws are changing all the time. As an immigrant, I use the model to give me the best advice along the immigration process.
–For any big decision such as buying a house, we always ask the model to review to point out potential risks and so on.
–After having kids, we are doing estate planning and the model saved us a lot of legal fees and energy to make a solid estate plan. Without the model’s help, we were not even sure what questions to ask, which can incur a lot of billable hours without making a lot of progress. The models are very precise questions to be discussed and save a lot of time and communication costs with our attorneys.
Shuchao Bi retweeted
🚨 Zuck is live on Insta smoking meats again 🚨
This is not a drill.
“Yeah Tank, we’re smoking sausage. Smoke whatever the fuck we want.”
