# ArtificialIntelligence

Latest AI Updates

## WORLD MODEL

### Learning New Abilities

The 1X World Model now serves as NEO’s cognitive core—enabling generalization to unseen tasks. Ask NEO anything and watch it attempt the task entirely autonomously.

[Watch](https://youtu.be/lS_z60kjVEk) [Read](/content/discover/world-model-self-learning/index.html)

### Evaluation

The 1X World Model is a video model grounded in physics. It allows us to predict—or "hallucinate"—the outcomes of NEO's actions before they're taken in the real world.

[Watch](https://www.youtube.com/watch?v=xPX6dDRYbV4) [Read](/content/discover/redwood-ai-world-model/index.html)

## Mobility

Redwood AI’s latest Reinforcement Learning (RL) controller gives NEO a complete mobility toolkit—enabling natural walking in any direction, sitting, standing, kneeling and stair climbing using stereo vision.

[Watch](https://www.youtube.com/watch?v=Dp6sqx9BGZs) [Read](/content/discover/redwood-mobility/index.html)

### Introducing Redwood AI

Redwood is the AI model that enables NEO to help around the home—moving, perceiving, and interacting with the world autonomously.

[Watch](https://www.youtube.com/watch?v=qnzL5dVTDKk) [Read](/content/discover/redwood-ai/index.html)

## Mobile Manipulation

Capable of navigation to objects of interest and coordinate precise placement of the body in order to maximize chance of successful grasps and execute chores that require movement space.

Redwood is training on both successes and failures, allowing NEO to learn from every interaction — and gain valuable context from failure data.

Redwood is among the first VLAs to control locomotion jointly with manipulation, enabling bracing and leaning behaviors during manipulation.

Redwood can handle variation in tasks, such as picking up never-before seen objects that appear in unique locations.

Integrated with an off-board language that predicts user intent from voice and conversation context in real time—allowing users to simply talk to NEO and ask it to take action.