Google Forms New AI Team to Simulate the Physical World: A Step Towards AGI

Google has announced the formation of a new team at DeepMind that will focus on developing AI models capable of simulating the physical world. This project, unveiled at CES 2025, marks a significant move towards achieving Artificial General Intelligence (AGI), which refers to AI systems that can perform any task that a human can do.

What Is Google’s New AI Team Working On?

Tim Brooks, the co-lead behind OpenAI’s Sora video generator, is now leading this new team at Google DeepMind. The team will work on creating generative models that simulate real-world environments.

Brooks mentioned in a post on X that the team’s goal is to scale models and solve complex problems using cutting-edge AI technology. Their mission is to take AI a step closer to AGI.

Key Objectives:

Create world models that can simulate physical environments in real time.
Collaborate with Gemini, Veo, and Genie teams to scale AI models.
Push the boundaries of robotics, interactive entertainment, and visual reasoning.

Understanding World Models and Their Potential

The concept of world models is central to Google’s new AI team. These are AI systems that can simulate the real world and interact with it in a way that is as close to human reasoning as possible. Here’s how Google’s current projects fit into this vision:

Key Google AI Models:

Gemini: A flagship AI model for image analysis and text generation.
Veo: An AI model focused on video generation, making it ideal for simulations.
Genie: This model focuses on generating 3D environments and is designed for real-time interactions.

These models will allow AI to not only analyze and generate content but also to interact with it, making them crucial for applications like robotic training, interactive games, and more.

Applications of World Models:

Visual reasoning
Simulation and planning for robotics
Real-time interactive entertainment

Will AI World Models Replace Human Jobs?

As AI models become more advanced, there’s increasing concern about their impact on jobs, especially in the creative industries.

AI and Job Disruption:

According to a 2024 study by the Animation Guild, over 100,000 jobs in film, television, and animation industries could be affected by AI by 2026.
Game studios, like Activision Blizzard, are already using AI to enhance productivity, which could result in fewer human jobs.

However, not all startups are seeking to replace human workers. For example, Odyssey, a startup in the world modeling space, promises to collaborate with creatives rather than replace them.

Key Concerns:

Job displacement in creative fields like gaming, animation, and film.
Potential copyright issues with training AI models on video content without proper licensing.

Google has claimed it follows YouTube’s terms of service to use videos for training, but the specific videos used are not disclosed.

Google’s World Models: What’s Next?

Google’s world models represent the next frontier in AI technology. While the possibilities are vast—ranging from robotics to real-time simulations—the full impact of these models is yet to be seen. Whether they will replace jobs or simply augment them remains a topic of debate.

One thing is clear: Google’s new AI team is on a mission to take AI’s capabilities to the next level and bring us closer to realizing AGI.