VLGE AI

REIMAGINE REALITY

High-end Spatial and Human Intelligence data for frontier AI training.

[VLGE AI @vlgeai]

Yesterday we hosted our first private @vlgeai Salon dinner in SF, bringing together a curated group of frontier AI researchers for an evening of connection and culture.

Guests included @OpenAI @GoogleDeepMind , Prometheus , @NVIDIAAI , @NASA , @Tesla , @physical_int , @NianticSpatial @amazon and more.

@evelynmoraworld curated the guest list, menu, and atmosphere of the evening the same way she would do for Vogue editors during fashion week.

This was only the first salon. There is more to come! Thanks everyone for making it special and @autopoiesislab for sponsoring.

[VLGE AI @vlgeai]

Our first VLGE Salon dinner was a special

[👩‍💻 Paige Bailey @DynamicWebPaige]

Truly beautiful dinner robotics and spatial intelligence salon dinner last night, hosted by @evelynxmora and @vlge__! 🤖

Attendees from @OpenAI, @maticrobots, @NianticSpatial, @GoogleDeepMind, @NASA, and more. And a spectacular view of San Francisco outside, gorgeous location:

[VLGE AI @vlgeai]

@DynamicWebPaige @evelynmoraworld @evelynxmora @vlge__ @OpenAI @maticrobots @NianticSpatial @GoogleDeepMind @NASA 💙

[VLGE AI @vlgeai]

We hosted the first Open World Hackathon in San Francisco by VLGE AI

judged by leading researchers from Google DeepMind, Prometheus, NASA, NVIDIA, Meta

Building the next generation of intelligent machines From world models to real-world robotics

participants built robots to perceive, reason and act.

- firefighting simulations
- autonomous wheelchair navigation
- human–robotics collaboration

this is how we accelerate spatial intelligence. see you the next one!

[VLGE AI @vlgeai]

https://t.co/gqQ3UGFFUb

[VLGE AI @vlgeai]

In 48h, we’re bringing together handpicked engineers, researchers and judges for the VLGE AI Open World Hackathon in SF.

We’re excited to add to our judging panel:
1. Drew Jaegle - Prometheus
2. Riya Dashoriya - Meta
3. Edmond V. - Wyss Institute at Harvard University

Capacity is limited to 125 participants, with 100+ already taken.

See full list of judges below & Sign up.👇

[Roland Graser @roland_graser]

> today in sf
> physical ai is the new meta
> 48 later: the most elite hackathon that SF has ever seen
> judges incl. AI research leads & founding MTS from leading phy-ai labs
> organizer is europe fashion blood, ex prada, ex chanel et all
> but also does gaussian splatting
> capacity limited to 125 participants. check out the guest list.
> it's elite.

[Evelyn Mora @evelynmoraworld]

Excited to announce the Open World Hackathon in SF this Saturday!

On July 25, @vlgeai is bringing together 100+ exceptional builders, researchers, and creators for a full day dedicated to shaping the future of Physical AI.

We're excited to announce our first judges:
🤖 @DynamicWebPaige — Deepmind
🤖 @HSlifelearner — NVIDIA
🤖 @IgnacioGL. — NASA
🤖 and yours truly with @vlgeai 💎
...with more to be announced.
Hosted by the VLGE AI team alongside @cescacentini and @roland_graser

Participants will tackle challenges at the frontier of AI, including:
🧠 World Models
🤖 Physical AI & Robotics
🌍 Spatial Intelligence
👁️ Computer Vision
✨ Embodied AI & Human Interaction

If you're building what's next in AI, this is where you want to be.

Register in the comments!

[VLGE AI @vlgeai]

From real-world capture → editable 3D world → behavioral intelligence.

Every Gaussian Splat we process becomes an environment where humans and AI interact, generating structured behavioral data for embodied AI.

This is the infrastructure behind the next generation of world models.

[Evelyn Mora @evelynmoraworld]

We are building better worlds through high-quality human data.

When an experience ends in a real decision, every texture, light, sound, gamified element and interaction in that space becomes legible.

So what does unbiased human data actually look like? Here is my answer: It looks like someone opening their wallet.

People in @vlgeai worlds are not answering surveys or performing for a camera. They move through environments people have built, and it ends in a purchase or a decision. Skin in the game. That single fact changes the value of data captured before it.

Because a purchase is one of the most loaded actions a human takes. For example: A nine dollar mascara or a lipstick is never just a product. It is femininity, self expression, confidence, creativity, social standing.

Why somebody buys, and what somebody buys, is decided on conscious, unconscious and subconscious levels at once.

The question that built @vlgeai : what kind of worlds lead to those decisions? We worked with world leading brands and organizations like @Loreal, @MuseeLouvre , @CTilburyMakeup @wwd , @voguemagazine on impact and volume for years to define it.

Not just video games or just events. Not just flow and pathing data. The full arc of a person moved by a world, up to the moment they act. And the data is only good because the experience is.

This is what building better worlds through high quality human data means. It is where the natural, sustainable adoption of AI begins. Ai trained for humans at scale enhanced with the diversity, individuality and consent of each person.

A sample is now public: https://t.co/PE7ltwfdkc

#AGI #data #human

[Protege @withprotegeai]

We’re excited to announce our partnership with @vlge__. Together, we’re exploring a new category of AI training infrastructure built around embodied retail behavior, World Models, and spatial intelligence.

VLGE is building immersive environments and digital twin infrastructure that capture high-signal behavioral data such as movement, exploration patterns, hesitation loops, object interactions, and contextual decision-making.

Why this matters:
📊 Richer behavioral data for multimodal AI, world models, robotics, and agentic systems
🏬 Structured spatial intelligence captured inside real commerce environments
🔏 A strong shared focus on consent, anonymization, and ethical data governance

We’re excited about the long-term potential of this category and what it could unlock for next-generation AI systems. Reach out if you’re working in this space and interested in learning more about what this data can unlock for your models!

🔗 Full announcement in thread

#AIData #WorldModels #EmbodiedAI #SpatialIntelligence #Robotics #MultimodalAI #AgenticAI

[VLGE AI @vlgeai]

Everyone talks about training #WorldModels.
At VLGE AI, we're building the #Worlds.

This VLGE AI 3D Gaussian Splatting (3DGS) environment is part of our growing library of digital twin worlds, where every movement, interaction, pause, and decision can be captured alongside the environment itself. Human behavior synchronized with high-fidelity spatial context; telemetry, camera trajectories, object interactions, navigation patterns, attention signals, and replayable world states.

The next generation of Physical AI, robotics, and world models will be trained on worlds. 🌍

#GaussianSplatting #3DGS #PhysicalAI #EmbodiedAI #SpatialAI #Robotics #WorldModels

[VLGE AI @vlgeai]

SF 🤝 embodied AI

Excited to announce that VLGE is now in San Francisco.

The next breakthroughs in robotics, autonomous agents, simulation, and world models will require richer human context inside dynamic environments.

At VLGE, we generate high-signal behavioral training data ( 200k+ hours/mo JSON and video data) at scale through VLGE GS and retail environments. Every movement, interaction, exploration path, hesitation moment, spatial decision, object engagement, and environmental response becomes structured intelligence for the next generation of AI systems.

Excited to meet founders, researchers, labs, roboticists, infrastructure teams, and creative technologists shaping this future.

If you’re in SF, let’s connect. 🏁
[email protected]

@marcodemiroz @tipatat @TheVRFund @Tesla_Optimus @Figure_robot @PhysicalAI @BostonDynamics @NVIDIARobotics @1x_tech @agilityrobotics @TheSanctuaryAI @Apptronik @SkildAI @UnitreeRobotics @GoogleDeepMind @NEURARobotics @AGIBOTofficial @SereactAI @CovariantAI @AmbiRobotics @standardbots @gs_ai_ @Waabi_ai @wayve_ai @wandelbots @wandelbots @IntrinsicAI @mimicrobotics