AI Research Scientist at Meta
Joined August 2021
- Tweets 87
- Following 131
- Followers 468
- Likes 547
2 Photos and videos
Pinned Tweet
Excited to share what we've been working on at MSL
Apr 8
1/ today we're releasing muse spark, the first model from MSL. nine months ago we rebuilt our ai stack from scratch. new infrastructure, new architecture, new data pipelines. muse spark is the result of that work, and now it powers meta ai. 🧵
4
8
83
19,343
Ryan Ehrlich retweeted
Apr 11
okay this is too exciting :)
meta AI is now #2 in the app store, top AI app!
we are so back!
256
117
2,174
336,308
Ryan Ehrlich retweeted
Apr 11
Monitoring the MONOPOLY GO!Chat situation 👀
Apr 11
up to #3, coming for the crown 👑
that being said, MONOPOLY GO!Chat is now #1, so i’m learning a lot about the App Store
5
3
100
13,220
Ryan Ehrlich retweeted
Apr 10
I gave the new Meta Muse Spark model a bunch of assorted assets and it was able to carefully extract them and place it into a functioning app.
I've never seen something with a clear intuition for design and product-first thinking.
This is the first AI that *gets* UI design.
5
17
258
38,158
Ryan Ehrlich retweeted
Apr 11
honestly I didn’t even know our model could do some of these
Apr 10
Another image converted to code with Meta Muse spark. hard to believe this was all in one prompt
63
113
2,643
261,621
Ryan Ehrlich retweeted
Apr 9
Check out some cool ways the community has been putting Muse Spark to work (and play) 🧵👇
1/ x.com/skirano/status/2041926…
Apr 8
Ok this is actually pretty impressive and I truly didn't see any model doing this before or being able to do it to this extent.
When I asked Muse Spark from Meta to convert this image into code, it cut out the assets from the screens so it could use them correctly!
19
35
327
68,197
Ryan Ehrlich retweeted
Apr 8
Starting from scratch is hard, but we learned a lot in the process, and clean scaling bodes well for the future.
It’s been a lot of fun building this with such a hard-working and humble group of colleagues!
Apr 8
1/ It’s been so fun working with @shengjia_zhao, @alexandr_wang and the team to build muse spark from scratch. It is early and has rough edges, but excited to continue our research velocity. I especially love that we’re doubling down on the fundamental science. We’re focused on building methods that scale well along all axes: pretraining, RL, and test-time compute. ai.meta.com/blog/introducing…
2
3
35
3,854
Ryan Ehrlich retweeted
Apr 9
Meta AI is up to #6 in the App Store overnight, and still growing :)
Also who knew the 7-Eleven app was so popular
204
79
2,228
269,335
Ryan Ehrlich retweeted
Apr 8
a good writeup about Muse Spark on a few complex queries (multimodal, stock analysis, coding):
riteshkhanna.com/blog/muse-s…
27
54
593
50,089
Ryan Ehrlich retweeted
Apr 8
It’s been amazing to move so fast with the team here 🙂 Still much more to do
Give a try at meta.ai and let us know your feedback!
Apr 8
1/ today we're releasing muse spark, the first model from MSL. nine months ago we rebuilt our ai stack from scratch. new infrastructure, new architecture, new data pipelines. muse spark is the result of that work, and now it powers meta ai. 🧵
12
13
128
16,965
VERDICT: Meta Muse Spark is the REAL DEAL
I ran several tests, including reading a menu. Newly released Meta Muse Spark was on the ONLY frontier AI to get all the items correct.
Sorry ChatGPT, there is no "Slapped Wagyu Dog" on the menu💀.
riteshkhanna.com/blog/muse-s…
4
19
124
24,085
Ryan Ehrlich retweeted
Apr 8
try muse spark on meta.ai!
Apr 8
Ok this is actually pretty impressive and I truly didn't see any model doing this before or being able to do it to this extent.
When I asked Muse Spark from Meta to convert this image into code, it cut out the assets from the screens so it could use them correctly!
37
19
374
40,499
Ryan Ehrlich retweeted
Apr 8
try muse spark via the Meta AI app or meta.ai!
check out this simulation i made:
meta.ai/share/rh52SAI9g5z
27
18
232
19,621
Ryan Ehrlich retweeted
Apr 8
The best outcome for humanity is many strong AIs competing for the top spot.
Vercel is proudly powering meta.ai and the infrastructure that made today's model release possible.
Apr 8
1/ today we're releasing muse spark, the first model from MSL. nine months ago we rebuilt our ai stack from scratch. new infrastructure, new architecture, new data pipelines. muse spark is the result of that work, and now it powers meta ai. 🧵
27
29
498
111,463
Ryan Ehrlich retweeted
Apr 8
9 months of stepping
Apr 8
Meta is back! Not only is it a high scoring models on AA third party eval, it is also one of the most token efficient.
1
1
30
3,640
Ryan Ehrlich retweeted
Apr 8
Our focus has been establishing scientific and infrastructure fundamentals that support a very ambitious scaling roadmap research to break into new paradigms.
I’ve had a lot of fun working with some very smart people, building an efficient & rigorous pretraining stack 3/
3
4
39
3,651
Ryan Ehrlich retweeted
Apr 8
Check out Muse Spark, our first milestone in the quest for personal superintelligence! Scaling this with the team has been a total blast. Give it a spin and let us know what you think! 🥑
18
58
318
71,432
Ryan Ehrlich retweeted
Apr 8
1/ It’s been so fun working with @shengjia_zhao, @alexandr_wang and the team to build muse spark from scratch. It is early and has rough edges, but excited to continue our research velocity. I especially love that we’re doubling down on the fundamental science. We’re focused on building methods that scale well along all axes: pretraining, RL, and test-time compute. ai.meta.com/blog/introducing…
16
42
327
42,275
Ryan Ehrlich retweeted
Apr 8
Meta's new family of AI models can reach the same performance as Kimi-K2 with only 30% of the compute
and only 10% of the compute to reach the level of Llama-4 Maverick
they completely rebuilt their pretraining stack with improvements to model architecture, optimization, and data curation
10
23
464
39,692