Amazon, Nvidia, and AMD bet $310 million on AI startup building 3D world models

2026-06-17 · Source: The Decoder · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, Robotics & Autonomous Systems · Depth: Intermediate, quick

Summary

Amazon, Nvidia's, and AMD's venture arms have invested \$310 million in AI startup Odyssey ML, which develops 3D world models simulating the physical world. This funding round values Odyssey at \$1.45 billion and includes other backers like IQT, GV, Google chief scientist Jeff Dean, and investor Elad Gil. Odyssey ML utilizes AWS as its preferred cloud provider and operates on Amazon's Trainium chips. Founders Oliver Cameron and Jeff Hawke, both from autonomous vehicles, state their models comprehend physics, body language, and dynamics—capabilities beyond traditional language models. The 55-person team is distributed across London, Zurich, and Palo Alto. This investment highlights a growing industry focus on world models as the next major AI frontier, with figures like Yann LeCun, Demis Hassabis, and Fei-Fei Li also advocating for their role in achieving general AI.

Key takeaway

For AI Product Managers evaluating future AI capabilities, this \$310 million investment in Odyssey ML signals a strong industry shift towards 3D world models. You should investigate how simulating physics, body language, and dynamics could enhance your product roadmaps, especially for applications requiring real-world interaction beyond text. Consider prototyping with 3D simulation tools to stay ahead of this emerging trend and potential competitive advantage.

Key insights

World models, simulating physical reality, are attracting significant investment as the next frontier beyond language models for achieving general AI.

Principles

• World models simulate physics, body language, and dynamics.
• Language models alone are insufficient for human-level AI.
• Major tech firms are backing 3D world model development.

Method

Odyssey ML builds 3D world models that understand physics, body language, and dynamics, leveraging AWS and Amazon Trainium chips for simulation capabilities.

In practice

• Explore 3D world models for physical simulation applications.
• Consider AWS and Amazon Trainium for AI infrastructure.

Topics

• AI Startups
• World Models
• 3D Simulation
• Venture Capital
• Amazon Trainium
• Autonomous Systems

Best for: Entrepreneur, Investor, Director of AI/ML, AI Product Manager

Related on AIssential

• Yann LeCun's AI startup raises $1bn seed round backed by Nvidia and Temasek — Yann LeCun's AMI Labs secured over $1 billion to develop world models for physical world understanding.
• Google to invest up to $40B in Anthropic in cash and compute — Access to massive compute capacity is critical for AI model development and deployment.
• 😺 Over $2B in AI funding hit in a single news cycle — AI investment is shifting to enterprise verticals and agent infrastructure, with a global competitive landscape.
• Physical Intelligence is reportedly in talks to raise $1B, again — Physical Intelligence seeks to create general-purpose AI models for robots, akin to "ChatGPT for robots."
• Europe is at risk of sleepwalking into AI armageddon — European AI startups face "armageddon" due to rising US platform costs and capital market volatility.
• World model maker Odyssey nabs $1.45B valuation backed by Amazon and other big names — Odyssey builds AI world models simulating physical reality for diverse applications, backed by major tech investors.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Decoder.