World Models in 48 Hours: The Trilateral Tech War Between Alibaba, Tencent, and Massman

Between April 16th and 17th—48 hours—three major events shook China’s AI scene.

On April 16th, Alibaba Cloud’s ATH Innovation Division released HappyOyster, a world model positioned as a “world simulator” rather than a pure video generation tool, with core features including roaming and director modes.

Same day, Tencent’s Hunyuan team announced the open-sourcing of Hunyuan3D 2.0. Unlike Alibaba’s closed-source approach, Tencent went fully open-source, with seamless Unity game engine integration.

On April 17th, spatial intelligence company Massman Tech listed on the Hong Kong Stock Exchange, becoming the “world’s first spatial intelligence listed company.”

Three major events in 48 hours—the density is remarkable.

Let me try to trace the underlying logic from a technical angle.

HappyOyster’s focus is “generation + simulation”—not just generating video, but creating a navigable virtual world. In theory, this applies to gaming, simulation, and autonomous driving training.

Hunyuan3D-2 focuses on “generation + editing”—the 3D assets it generates aren’t just viewable, they’re exportable for二次编辑. This direction is closer to game development and industrial design.

Massman Tech takes the “spatial intelligence + hardware” route, doing digitization of real physical spaces.

Three directions, three approaches, each with its own logic.

What’s interesting is that all three happened almost simultaneously. That’s not coincidence—it’s competition.

Competition among Chinese AI companies has evolved from “whether to do it” to “who can actually deliver and monetize it first.”

For industry practitioners, this is good—more choices, more opportunities.

For investors, it’s a challenge—bet on the wrong direction and you might be done.

For regular users, this is also good—the final beneficiaries will be product experiences.

This battle? Worth watching.