Investing in Third Dimension

AI World Generator

The gaming industry has traveled light years since the simple bleeps and bloops of Pong and Space Invaders. Today's AAA video games, particularly sprawling open-world epics like Grand Theft Auto (GTA), aren't just games – they're technological marvels that cost more to create than Hollywood blockbusters. As millions of fans (myself included) anxiously await GTA 6, I can't help but wonder: what if the future of game development isn't about building virtual worlds, but growing them? Forget crafting pixels by hand – we might be about to witness the gaming industry's very own industrial revolution.

The Price Tag of Virtual Worlds

When Grand Theft Auto V was released in 2013, it reportedly cost around $265 million to develop and market. More recent estimates suggest that GTA 6's development budget could exceed $500 million. But why are these costs so astronomical?

A significant portion – approximately one-third – of video game development costs goes directly into creating environments. This includes:

• 3D modeling of buildings, vehicles, and objects

• Texture creation and mapping

• Environmental effects and lighting

• Level design and world building

• Quality assurance and bug fixing

For a game like GTA, which features vast cityscapes with countless unique buildings, streets, and interactive elements, this translates to hundreds of artists and designers working for years to craft every detail.

The Traditional Development Challenge

The current approach to game environment creation presents numerous challenges for studios, even with cutting-edge tools at their disposal. While powerful game engines like Unreal Engine are capable of rendering visually striking environments that closely resemble reality, the fundamental challenge lies in recreating the actual geometry of real-world locations. These engines, despite their sophisticated rendering capabilities, still rely heavily on artists and level designers to meticulously craft the geometry of the game world. This means that every building, street corner, and architectural detail must be manually modeled and refined by skilled professionals.

Third Dimension

Third Dimension wants to build a one-stop tool that accelerates companies’ abilities to create worlds, allows artists to have a baseline to get to quality faster and allows engineers to build fantastical or entirely accurate representations of the real world in high-fidelity. This technology will accelerate workflows of developers from months of time to days or hours and help save millions in expensive graphics development budgets.

At its core, the company is using recent advancements in generative AI — specifically Neural Radiance Fields (NeRF) — to quickly, cost-effectively and accurately create 3D assets of the real world. Leveraging learned neural rendering techniques, Third Dimension is able to encapsulate the geometry, motion, and appearance of scenes, along with the interactions of objects within them.

Team

Third Dimension is led by a AAA team with deep expertise in the world of Gaming, Neural Rendering, and Autonomous Systems.

Tolga Kart brings a unique blend of gaming and autonomous vehicle expertise to Third Dimension. As a former Vice President of Engineering at synthetic data company Parallel Domain, he honed his skills in creating virtual environments for AI training. His diverse background includes a pivotal role at Tesla, where he led technical program management and simulation for the Autopilot team. Before venturing into the autonomous vehicle space, Kart spent over seven years as a Senior Director at Activision's Sledgehammer Games studio, where he played a crucial role in developing acclaimed titles like Call of Duty: WWII and Call of Duty: Advanced Warfare. This combination of high-fidelity gaming development and AI simulation experience uniquely positions him to lead Third Dimension's innovative approach to AI-powered game world creation.

Piotr Sokolski brings deep expertise in photorealistic AI simulation from his four-year tenure at Wayve, where he developed cutting-edge neural simulators for autonomous vehicles. His impressive background includes over three years at Google, where he worked on advanced technology projects. This combination of experience in both autonomous vehicle simulation and large-scale technology development provides crucial insights for Third Dimension's mission to revolutionize game world creation.

Soner Özgün Pelvan contributes significant academic and practical expertise in machine learning to Third Dimension. As a published researcher with five papers focused on neural networks and data imputation, he brings rigorous scientific knowledge to the company's AI development efforts. His research background provides a strong theoretical foundation for Third Dimension's innovative approach to procedural world generation.