AI Content Creation Meets Xianxia: Breaking Down the Jingxuan Sword Sovereign Character Design

When AI Creation Meets the Xianxia World

Da Fei Yu Xiu Xian Zhuan (The Great Fat Fish Cultivation Chronicles) and the legend of the Jingxuan Sword Sovereign is a textbook example of AI-assisted content creation. It blends the classic xianxia (immortal cultivation) genre common in Chinese web novels with AI-generated imagery and character design, demonstrating how everyday creators can use AI tools to build a complete worldview and character identity.

This Bilibili video is tagged under "AI Universal Creator," and its core value isn't the complexity of its plot — it's the replicable creative framework it demonstrates: establish a character profile, use AI to generate the corresponding visuals, then tie everything together with copy to complete the narrative loop. For anyone looking to get into AI content creation, work like this is valuable precisely because it can be taken apart and studied.

Character Design: A Well-Layered Character Architecture

The Jingxuan Sword Sovereign is a cleverly constructed character. His true form is an "Ancient Primordial Black Whale" who, after sleeping for hundreds of millions of years in the deep abyss, watched the tides rise and fall and the stars wheel overhead — and in doing so, independently comprehended the Supreme Sword Dao, reaching the cultivation realm of "Tribulation Transcendence Grand Perfection, Half-Step Beyond." This setup gives the character an impressively high power ceiling while using the "slumbering enlightenment" device to sidestep the tedious description of a traditional cultivation journey.

What's particularly interesting is the contrast built into the character. Cultivators across the world fear his name, yet privately they've given him the playful nickname "Great Fat Fish." This combination of "apex powerhouse + cute nickname" is a proven formula for memorability in modern web fiction and fan-created content.

His personality is also clearly defined as "not a compassionate saint" — he understands the eat-or-be-eaten rules of the cultivation world and operates on the principle: "If no one troubles me, I'll sleep peacefully in the deep sea; if someone dares to provoke me, my response will be swift and absolute." This kind of morally ambiguous middle ground — neither a selfless hero nor a pure villain — is far easier for audiences to connect with than a flat black-and-white characterization.

---

"Tribulation Transcendence Grand Perfection, Half-Step Beyond" deserves a brief explanation for those unfamiliar with xianxia power systems. Cultivation novels typically divide the path of practice into multiple major stages. "Tribulation Transcendence" is the threshold closest to ascending or transcending the mortal realm; "Grand Perfection" means the character has maxed out that stage; and "Half-Step Beyond" implies the character already stands at the boundary between the mortal and a higher plane of existence — without having fully crossed over. This kind of "standing at the threshold" setup is extremely common in web fiction for good reason: it guarantees the character is powerful enough to dominate nearly all opponents, while preserving narrative space for further growth and leaving the hook of "the conditions for true transcendence haven't been met yet." For AI-assisted creation, this type of clearly defined boundary in the power system has a practical bonus: it gives the character a clear reference point within the worldview, making it easy to position them accurately when writing copy without repeatedly explaining the power hierarchy.

---

Combat Power and Forms: Built to Serve AI Visual Generation

Looking at this character from an AI creation perspective, every element of the design seems built to feed visual generation. The combat description — "the Ten Thousand Whale Star-Sea Sword Formation unleashed, ten thousand flying swords wrapped in the force of the surging ocean, a single strike capable of shattering mountains and splitting seas" — maps directly onto the grand, effect-heavy scenes that AI image generation excels at rendering.

Even more noteworthy is the multi-form design: the character can freely shift between a "cool and ethereal young girl form" and a "round, soft, chibi little black fish form." At its core, this multi-form setup gives AI image generation multiple distinct creative targets — a single IP can produce beautiful/ethereal-style images alongside cute/chibi-style ones, dramatically expanding the content asset library.

The design also includes a built-in balance mechanic: when taking the miniature form, peak combat power is suppressed, leaving only enough strength for self-defense. Details like this give the character believable room to perform differently across various scenes, and are a hallmark of mature character design.

What AI Universal Creation Tells Us

Setting aside the plot itself, this video reflects how AI tools are lowering the barrier to content creation. In the past, building an original IP with visuals, character lore, and a coherent narrative required collaboration between artists and writers. Today, with AI image generation and text tools, individual creators can do it alone.

For anyone looking to get started, here are a few takeaways from this case study: begin with a clear character design document that covers the character's true form, personality, combat power, and transformations; then write scene-specific visual descriptions for each key moment and hand them to AI for generation; finally, use cohesive copy to string the images into a story. The more specific the character design, the more consistent the AI output — which is exactly why works like this tend to go into extensive detail on character specifics.

It's worth noting that this kind of content is still in an early exploratory stage. Visual consistency and narrative depth both have obvious room for improvement, and for now these are best understood as entertaining creative experiments rather than polished commercial products.

---

Visual consistency is the most fundamental technical bottleneck in AI image generation for IP creation today. Current text-to-image models (such as Stable Diffusion, Midjourney, etc.) have reached impressive quality for individual images, but maintaining highly consistent facial features, body proportions, and costume details for the same character across different scenes and poses still requires advanced techniques like LoRA fine-tuning, ControlNet skeletal control, or IP-Adapter. For everyday creators, the common workaround is to lock in a detailed character prompt and repeatedly use the same base model, then generate large batches of images and cherry-pick from them to maintain a relatively consistent visual style. This also explains why works of this type spend so much time on detailed character descriptions — the more precise the written design, the more effectively it translates into generation prompts, which in turn improves output stability. As video generation and character consistency technologies continue to iterate rapidly, this bottleneck is expected to see significant improvement within the next one to two years.

---