Fei-Fei Li
@drfeifei
Followers
528K
Following
7K
Media
272
Statuses
3K
Prof (CS @Stanford), Co-Director @StanfordHAI, Cofounder/CEO @theworldlabs, #AI #SpatialIntelligence #GenAI #computervision #robotics #AI-healthcare
Stanford, CA, U.S.A.
Joined April 2010
After 3+ years, today is the day that my book “The Worlds I See” gets to see the world itself. It is a science memoir of the intertwining histories of me becoming an #AI scientist, and the making of the modern AI itself. All versions are now on Amazon https://t.co/1oH2vCjovu 1/
433
594
3K
In 2024, global AI investment crossed $109B, with Generative AI attracting $33.9B on its own, up 18.7% year-over-year (Stanford AI Index, 2025). 78% of organizations now report using AI in at least one function. But here’s the gap no one talks about: billions in investment, and
34
22
108
We are hiring researchers who are obsessed with generating pixels 🤩
We're hiring research scientists in diffusion / generative models at @theworldlabs! Check out the role here: https://t.co/cpBvsvzAry Feel free to DM me if you would like to chat!👋
35
89
1K
First test stitching together World Labs generations to build an interactive fps experience. Every environment you see was generated from single images. How it was built:
40
85
742
Parents, would you put $20 into a low-risk savings account that you can withdraw at any time, but instead of accruing value, your kids can access endless educational resources?
3
1
20
Hungry for talents!🤩
We're hiring product engineers at @theworldlabs. If you have strong opinions about react-three-fiber, can derive ray-triangle intersection by hand, and want to shape the future of web-native 3D AI tools... let's talk.
20
27
546
🔥🔥🔥
Total insanity. This is using an adaptive LOD scheme in sparksjs (not merged yet). The entire scene has 16million splats and this is real time navigation ... 😱😱
12
58
719
Morning stroll into a @theworldlabs portal. I picture a future where you describe your environment and it instantly materializes around you.
31
41
409
Marble from World Labs is so close to magic. You can imagine a space, generate it and then walk around it in your VR Headset. This was recorded with my Quest 3 and the potential is mind-blowing.
9
12
90
The AI bubble will pop. Intelligence won't. Our CEO @davidpstout wrote in @FortuneMagazine on why the mega-model economics don't work— and what actually delivers sustainable AI value. While Wall Street chases billion-dollar models, financial firms, healthcare systems, and
0
7
16
Turn any world model from @theworldlabs into a multi-user space for exploration and collaboration with @frame_vr. Voice chat, embodied AI agents, and tons more you can layer on top of the model foundation. This took all of 10 seconds to get going. The metaverse is so back. 😄
8
20
73
@theworldlabs congrats @theworldlabs! so much potential! 😍 I had to try and link up several worlds from @BrittaniNatali and it’s incredibly fun to spend time and meet up in these AI-generated environments. https://t.co/RV3b4T5fLJ
0
8
27
Made this cavelike restaurant 3d environment with Marble from @theworldlabs @drfeifei, just with an image The model converts it into environments you can move around & has great object permanence. This is done with Gaussian Splatting, more on that and results from my other
9
58
448
Grafting Diffusion Transformers accepted to #NeurIPS2025 as an Oral! We have lots of interesting analysis, a test bed for model grafting, and insights🚀 📄Paper: https://t.co/OjsrOZi7in 🌎Website:
arxiv.org
Designing model architectures requires decisions such as selecting operators (e.g., attention, convolution) and configurations (e.g., depth, width). However, evaluating the impact of these...
1/ Model architectures have been mostly treated as fixed post-training. 🌱 Introducing Grafting: A new way to edit pretrained diffusion transformers, allowing us to customize architectural designs on a small compute budget. 🌎 https://t.co/fjOTVqfVZr Co-led with @MichaelPoli6
7
33
203
🇳🇱🗽 This is probably the closest you can get to walking around in New Amsterdam in 1670 (current day New York) A 3d world made with @theworldlabs from the AI render below which is based on a real map from then!
I think I'll make this a mini app if I get time today, pick a place on the map of New Amsterdam and it generates a POV shot of it with Nano I think this is the closest you can get to going back into time to visit New Amsterdam in 1670 (current day New York City)
96
208
3K
Very excited to see the overwhelming interests! The waitlist is still open :) Come and try to Build some Worlds!
We're excited to see so many of you creating new worlds in our Marble beta preview! Here's a tour through some of your latest creations… To join the fun, sign up at https://t.co/Jqx9KTLHdU — we're letting in new users as fast as we can 🔥
37
74
916
How does the Carrier Delivery Catheter get the aspiration catheter to the face of the clot in 100% of cases? It's tapered distal segment is optimized for seeking primary vessel with atraumatic navigation. Check out the article here
0
0
10
You can explore these AI-generated worlds on #VisionPro in 2 ways 👇 Option 1 is to get a link directly from Marble, the platform from @theworldlabs that has recently opened beta access and allows users to use a prompt image to generate worlds as #GaussianSplatting. You just
1
44
284
✨ It's not even 2026 yet, but you can start using AI world models today! This is @theworldlabs running inside my app 🏡 Interior AI letting you redesign your home with AI and then walk through it! It's really cool and of course not perfect especially since for now it's based
2023 was the year of AI avatars 2024 was the year of AI photos 2025 was the year of AI videos And I think it's becoming clear now that 2026 will be the year of AI world models Fully interactive explorable 3d worlds generated from one or multiple 2d images or a prompt In turn
67
76
861
2023 was the year of AI avatars 2024 was the year of AI photos 2025 was the year of AI videos And I think it's becoming clear now that 2026 will be the year of AI world models Fully interactive explorable 3d worlds generated from one or multiple 2d images or a prompt In turn
✨ Upgraded Photo AI's default video model with high resemblance/likeness 2025 really is the year of AI video, we're leaving uncanny valley fast I think High res skin texture, real person facial movements etc. all based on the person you trained on Photo AI
160
153
2K
The latest unlock in AI Filmmaking is here! Over the last few weeks we've nailed down Consistent Characters. But Location and Camera placement has continued to be an issue. Not any longer. Consistent environments are now here with World Labs Marble! Toolset Below!
87
162
1K
This is only the beginning. 🔊 Sound On What does it mean to explore a house built from memory — one that feels familiar, yet shifts each time you return? That question is at the heart of Memory House, a project that unfolds through space, silence, and surreal interruption.
11
12
83
Mesmerizing… @theworldlabs @BrittaniNatali
Some of the large scene's being generated are really unbelievable. This is a generative 3D splat, not 2D video. The full scene is likely over 50million splats. You can see some stitching inconsistencies from the final v2v pass. Also nice music :)
8
30
278