The World Model
The World Model
Genie 3: A new frontier for world models - Genie 3, a general purpose world model that can generate an unprecedented diversity of interactive environments. Given a text prompt, Genie 3 can generate dynamic worlds that you can navigate in real time at 24 frames per second, retaining consistency for a few minutes at a resolution of 720p.
While Genie 3 pushes the boundaries of what world models can accomplish, it's important to acknowledge its current limitations:
Limited action space. Although promptable world events allow for a wide range of environmental interventions, they are not necessarily performed by the agent itself. The range of actions agents can perform directly is currently constrained.
Interaction and simulation of other agents. Accurately modeling complex interactions between multiple independent agents in shared environments is still an ongoing research challenge.
Accurate representation of real-world locations. Genie 3 is currently unable to simulate real-world locations with perfect geographic accuracy.
Text rendering. Clear and legible text is often only generated when provided in the input world description.
Limited interaction duration. The model can currently support a few minutes of continuous interaction, rather than extended hours.
Given an image or text prompt, World Labs model generates a 3D world that you can explore for as long as you wish — no time limits, no morphing, no inconsistency. Compared to their previous results, our generated worlds are bigger, more stylistically diverse, and have cleaner 3D geometry.
Marble makes the model available for users to start viewing and building worlds today. Enthusiasts and builders can export generated worlds as Gaussian splats and use them in downstream projects. This is particularly easy to do with the open-source rendering library Spark which seamlessly integrates Gaussian splats into Three.js for building web-based 3D experiences, and renders efficiently on desktops, laptops, mobile devices, and VR headsets.
The model's consistency and style adherence now make it possible to build out large worlds by composing individual generations, as shown in the banner video above and the examples at the end of this post.
Source: https://www.worldlabs.ai/blog/bigger-better-worlds
When you finish constructing each scene, you can use www.jawset.com to piece them together.
Old Milestones: 2023 - OpenAI’s advanced models allowed Altera to build the first AI agents that play games with people, just like their friends. These agents achieve longer, more complex interactions without the rapid decline in performance that had been limiting the agents’ potential.
By combining OpenAI’s GPT models with Altera’s parallel multi-module system that mimics the structure of the human brain, including that of the prefrontal cortex, the company was able to create agents capable of simulating cognitive functions. “Our composite system combines various modules in parallel, each powered by OpenAI models. These modules are inspired by brain functions, like attention bottleneck, working memory, and social cognition,” says the Altera CEO.
Major Learning Hubs
Microsoft
Artificial Intelligence
Prompt Engineering
Machine Learning and AI
Videos