How was this done? I've experimented a lot and nothing comes close to this guys work

A viral animation posted in August 2025 by a creator under the alias Stickyspoodge has ignited debate within the AI and digital art communities. The creator openly acknowledges using AI in the work, and close inspection reveals subtle artifacts—particularly in hand details—that confirm its AI-generated nature. However, the piece's defining characteristic is its remarkably fluid, coherent motion and a distinct, consistent animation style that, according to expert analysis, appears to surpass the capabilities of publicly available tools like Runway Gen-3, Pika Labs, or Stable Video Diffusion. This has led to widespread speculation that the creator may be using an unreleased, proprietary model or has developed a novel, undisclosed workflow that pushes the boundaries of what's currently possible.

The animation's release date of August 2025 places it in a future context, suggesting it may represent a forthcoming leap in technology. The core mystery lies in the specific technique or model used to achieve such high-fidelity temporal coherence and stylistic control, which are two of the most significant challenges in AI video generation. For professional animators and VFX artists, this work is not just a curiosity but a potential harbinger of new tools that could dramatically streamline production pipelines. It underscores the rapid, often opaque pace of advancement in generative video AI, where significant capability jumps can emerge from independent creators before being formalized by major labs like OpenAI or Midjourney.