© IE Online Media Services Pvt Ltd
Tags:
Journalism of Courage
Chinese-American computer scientist known for her pioneering work in computer vision, Fei-Fei Li, has introduced Marble, her company’s first commercial world model.
The model developed by Li’s company World Labs is capable of generating 3D environments from text, videos, and 3D layouts. Marble is reportedly ahead of its peers Genie and Decart developed by Google.
The launch of Marble signals the next frontier in AI – spatial intelligence that needs powerful world models that reconstruct, generate, and simulate 3D models. According to World Labs, spatially intelligent world models will transform a wide variety of industries over the coming years.
What is Marble?
Marble is a first-in-class generative multimodal world model which is generally available for everyone to use. It is a world model which is essentially an AI system that generates predictive simulation of the real world. These models are capable of learning the physics and spatial properties of an environment. It studies how objects interact with each other to predict future results.
Reportedly, world models can plan ahead by internally simulating different actions and their consequences. This is a key aspect for tasks like robotics and autonomous navigation.
Marble allows users to create new worlds through text, image, and video prompts. It also allows them to edit or combine, and even expand on existing worlds to make detailed changes. The platform allows creators to export worlds as Gaussian splats, videos or meshes that facilitate their use in gaming, VFX and VR workflows.
“Marble is the first of its kind – a next-generation world model making strides toward this vision. It can now create 3D worlds from a wide variety of input types, and lets users iteratively edit or expand worlds,” the company said in its official website.
According to the company, Marble allows users to dive deeper as they want to control their generated worlds. They can instantly create full 3D worlds from a simple image or text prompt or interactively edit worlds in 2D and 3D.
It was initially released as a preview in September and offers both freemium and paid tiers starting from $20 per month. Even though world models can enhance VR and gaming, there are boundless applications for this technology. It could yield simulated environments for robotics, architecture design to even world building for cinema.
Who is Fei-Fei Li?
Li is among the most influential figures in artificial intelligence. She is a professor at Stanford University and co-director of the Stanford Human-Centered AI Institute. She is best known for her work in large-scale visual recognition through the invention of ImageNet which is a dataset over 14 million images that went on to revolutionise deep learning.
Li also worked as Chief Scientist of AI/ML at Google Cloud and is a vocal advocate for ethical and human-centered AI. She also served as a board member on Twitter (now X) briefly. Her work aims to bridge technology and humanity, highlighting the need for diversity, empathy, and social responsibility in AI development.
