World Models with NVIDIA Cosmos: Physical AI - AI Build & Learn

Welcome to AI Build & Learn, a weekly AI engineering stream where we pick a new topic and learn by building together.

This event goes bigger on world models with NVIDIA Cosmos, an open family of world foundation models built for physical AI. Where DreamerV3 learns a small world model for a single agent, Cosmos is a large generative model that simulates the physical world itself: predicting future video, running action-conditioned rollouts, and generating synthetic data to train robots and autonomous machines. It ties directly back to the Isaac Sim event.

Cosmos 3 (released 2026) exposes two surfaces: a Reasoner for understanding and planning, and a Generator for world simulation and future prediction. The weights are open (OpenMDW license) and available on Hugging Face.

Some things to look up to get started

Model

NVIDIA Cosmos (open platform of world models): github.com/nvidia/cosmos
Cosmos overview: nvidia.com/en-us/ai/cosmos/
Cosmos Cookbook (runnable recipes): nvidia-cosmos.github.io/cosmos-cookbook/

Resources

GitHub: github.com/sagecodes/ai-build-and-learn
Events Calendar: luma.com/ai-builders-and-learners
Slack (Discuss during the week): slack.flyte.org/
Hosted by Sage Elliott: linkedin.com/in/sageelliott/

In this stream

Intro to topic
Community Discussion
Practical examples

Community challenge (optional)
Try spending 30–90 minutes during the week learning or building something related to the topic, then share what you’re working on in Slack.

Note on Flyte / Union
You may see Flyte used in some demos. Flyte is an open-source AI orchestration platform maintained by Union (where I work) for building scalable, durable, and observable AI workflows. You do not need to use Flyte to participate.

Union: union.ai/
Flyte: flyte.org/

Drop a comment with ideas for future topics (agents, RAG, MLOps, robotics, framew