Game Science founder and CEO Feng Ji released the first in‑engine video for Black Myth: Zhong Kui on February 10, a near seven‑minute sequence that the studio describes as a non‑interactive New Year’s greeting rather than a gameplay demonstration. The clip centres on a grotesque yet cinematic kitchen scene — a human‑ghost chef preparing food — and closes with a dimly lit shot of a kitchen girl presenting a large dish to a hooded, anonymous figure.
Feng was explicit that the video contains no player control, no combat, no gameplay systems and, notably, no appearance by Zhong Kui himself; he framed the piece as a short, festive vignette rather than a technical showcase of mechanics. The release follows months of high anticipation for the studio’s ambitious single‑player action RPG, which has repeatedly attracted attention for its polished visuals and cinematic trailer work.
The timing — a brief, evocative piece dropped during Lunar New Year festivities — reads as a carefully chosen marketing moment. Black Myth: Zhong Kui has been one of China’s most visible game projects internationally, praised for bringing a distinctly Chinese mythological aesthetic to AAA‑style production values. The clip keeps the studio in the conversation while postponing questions about release timing or playable demonstrations.
The release gained additional resonance because a day earlier Feng Ji had posted a detailed reaction to Seedance2.0, a video‑generation model released by a major Chinese tech firm. He said he was “deeply shaken” by the model’s capabilities and urged people to warn older relatives and anyone unfamiliar with AI that unverified videos — particularly those depicting recognizable faces or voices — may be fabricated.
Feng’s dual actions — publishing a polished, clearly staged studio video and flagging the near‑term risks of synthetic media — encapsulate a growing tension in digital culture. Game trailers and cinematic footage have long blurred the line between in‑engine footage and prerendered cinematics; generative video models add a new dimension by enabling highly realistic but artificial content that can mimic individuals or events.
The implications extend beyond marketing theatre. As generative models proliferate, studios, platforms and regulators will face pressure to authenticate official materials and to develop provenance tools — visible watermarks, embedded metadata, or platform verification — so consumers can distinguish between sanctioned releases and AI‑forged fakes. Feng’s request that people “remind parents and less tech‑savvy friends” underscores an acute social dimension: the weakest link in circulation is often not technological capability but public media literacy.
For international observers the episode is a compact case study in two converging trends: the rising international profile of China’s game development scene and the acceleration of synthetic‑media challenges inside tech ecosystems. The colourful, atmospheric video keeps interest in Black Myth alive, but the founder’s warning about models like Seedance2.0 is a timely reminder that in the coming months visuals alone will be an increasingly unreliable proof of reality.