Liang Qiwei’s Search for a Sustainable Creative Life

1. Guest Background

Liang Qiwei appears on Zhang Xiaojun’s podcast as the producer of Phantom Blade Zero. Running for more than six hours according to the episode metadata, the conversation explores the relationship between a fictional hero’s convictions and a creator’s working life. Wuxia supplies its vocabulary, but the subjects include architectural education, independent development, mobile games, financing, team management and the sources of player trust.

Liang’s relevant experience crosses several of those boundaries. He recalls studying architecture at Tsinghua and Yale. At Tsinghua, a failed course cost him an already secured postgraduate recommendation; he subsequently completed the fifth year of the architecture undergraduate program and graduated normally. Going abroad, in his telling, initially replaced one prescribed achievement route with another.

The deliberate departure came later. Feedback on student games, online sales worth several tens of thousands of dollars and exposure to different ways of living gradually made professional game development imaginable. He estimates that this change unfolded over roughly five years before he declined an available job at New York architecture firm KPF and returned to China.

His subsequent independence also rested on support. Liang says his parents financed his initial rented accommodation in Beijing, and Xu Xiaoping invested RMB 1 million in him personally before the company existed. Mobile development later supplied professional training in commercial judgment, management and technical tradeoffs. These experiences make him a participant reflecting on specific choices and constraints, rather than an observer offering a detached theory of the industry.

2. What the Episode Covers

A central contrast in the interview concerns creative completion and commercial reward. Liang believes Rain Blood’s action prequel, Mirage, exhausted the resources and abilities available to its team. He expresses little regret about what they put into the work, yet says it probably made no money. His explanation includes timing, limited publishing experience and problems with overseas distribution. The subsequent mobile Phantom Blade reused and simplified its materials and controls. Its combo-chain system accommodated the difficulty of executing complex inputs on a phone, while the product generated stronger commercial returns.

Liang describes the lesson the team initially drew as misleading: strategy and catching an expanding market seemed more consequential than painstaking creation. Around 2015–2017, fundraising, pitches, animation, merchandise, an IP portfolio and listing ambitions absorbed attention. He recalls sometimes spending only one day a week on development. Phantom Blade 2 pursued mobile esports and anime-oriented presentation. Although he credits improvements in areas such as 2D animation, he sees its overall direction as a departure that failed to deliver the expected results.

Phantom Blade 3 moved toward PVE and a more single-player-like experience, alongside renewed personal involvement in production. Liang reports first-month gross billings of approximately RMB 150 million and says the game was profitable. Those statements describe different financial measures. He says profits helped repurchase early shares, while adding that NetEase’s wish to exit some investments and an attractive price created the immediate opportunity. The transaction therefore has a financial explanation alongside the story of recovering creative room.

The move to Phantom Blade Zero was similarly conditional. A single-player attempt in 2017–2018 had been abandoned because of cost, cash flow and insufficient resolve. The later restart benefited, Liang says, from momentum created by Black Myth: Wukong, accumulated material and more favorable funding and partnership conditions. He acknowledges that following an opened route carried less pioneering risk. Nevertheless, the studio had to cancel projects, change staffing and move beyond side-scrolling 2D without prior 3D project experience. Public trailers and playable demonstrations helped sustain confidence during that uncertainty.

Liang discloses more than RMB 400 million in earlier Tencent investment for a stake above 20%, followed by roughly US$30 million of project financing, mainly convertible debt, with Tencent and IDG participating. He confirms project spending of RMB 400–500 million and describes safeguards in the financing. These are participant disclosures, not independently examined accounts. He also presents self-publishing as a voluntary choice, explicitly rejecting a suggestion that Sony had become cool toward the project. Control over spending, tone and cultural exchange mattered, even with possible gaps in overseas coverage.

The fictional world brings these questions back to intention and consequence. Soul seeks to survive a 66-day limit; Zuo Shang works through alliances and planning toward a larger vision. Liang summarizes their contrast as self-rescue that saves the world versus world-rescue that destroys it. Yet the interview does not establish the finished work’s reception. At recording, Liang said release was twenty days away, with fixes, testing and small optimizations remaining. His satisfaction with his best effort coexisted with acknowledged flaws and an original IP still awaiting players’ judgment.

3. Core Views: Reasoning, Examples, and Limits

Liang’s definition of a wuxia hero begins with a distinction between what starts a story and what gives an action moral weight. Hatred and injustice supply propulsion; chivalry emerges from the response. A sympathetic identity or a benevolent intention is insufficient. His example is a person rescuing an elderly woman who has collapsed amid traffic, accepting both physical danger and the possibility of being wrongly blamed. Risk, cost and help actually delivered distinguish the act. This is an interpretive account of heroism, however, and should not become practical advice to seek danger or bypass safer ways of helping.

His account of the genre’s environment follows the same logic. When official order cannot provide protection, marginal and grassroots figures gain a space in which to act. Aesthetic contrast makes that space emotionally legible. Liang recalls the beautiful desert vistas in New Dragon Gate Inn alongside the inn’s underground darkness and the horror of its climactic violence. Beauty and fear intensify one another. The useful implication is that darkness performs a function within a larger emotional composition. Merely making a world uniformly grim would not reproduce the contrast he values. His reading explains his preferences; it does not establish an exhaustive definition of wuxia.

Modernization introduces an unresolved tension. Liang initially defends enduring universal values in wuxia, but later says the thematic core of his 2006 Rain Blood and Phantom Blade Zero is fundamentally different. He characterizes traditional wuxia as moving from oneself toward others, and punk as a return from the self to the self. They meet when helping someone follows from a character’s own empathy or moral requirements, without a prior ambition to become a great hero. This changes why a protagonist acts, beyond changing the visual form. It also remains Liang’s reading of contemporary sensibilities, rather than proof that an entire generation shares one psychology.

An unsuccessful trailer choice makes the argument concrete. Liang says the team inserted elevated declarations of chivalric duty to obtain an additional layer of emotional approval. Players, in his account, detected the attempt to embellish the creators’ moral standing. He interprets core fans as asking for the established work: assassins, revenge, false accusations, dark organizations and satisfying combat, with meaning emerging through action. His reference to 80% disapproval has no stated sampling method. The example supports a possible mismatch between an audience’s expectations and a promotional statement; it cannot demonstrate that explicit moral language always fails.

Authorship, meanwhile, includes limitations as well as distinctiveness. Liang enjoys conspiracies, loopholes and the architecture of a plot, but insists that structure must eventually return to individual motives. He acknowledges that overinvestment in a puzzle can leave characters emotionally thin or encourage familiar character types to fill structural slots. His organizational preference has a comparable boundary: personally establish the creative core, demonstrate competence, teach others and then bring resources around it. That may explain how he earns creative authority, while also making personal attention a constraint on scale. He explicitly leaves open whether AI could change that constraint.

The choice of business model follows this emphasis on fit. Liang identifies sustained free-to-play operation, rather than phones themselves, as the difficulty for his team. Maintaining enthusiasm, updates and production rhythms for years requires capabilities different from concentrating on a work with an endpoint and then starting another. He recognizes the commercial strength of excellent mobile games and asks what his own studio can execute well. His broader market claims need similar care. He considers overseas AAA mature and domestic supply still developing, citing Chinese players’ roughly 25% share of Steam spending against Chinese developers’ roughly 5% revenue share. The episode supplies no source, year or definitions for those figures, so they cannot establish a verified market opportunity.

AI finally separates making an experience from persuading someone to pay for it. Liang says Phantom Blade Zero used assistance in technical areas such as logic, QA and optimization, but had not used AI to create its art. He supports AI; his explanation concerns perceived value rather than an inevitable quality deficit. Players may be paying for authorship, effort, trust in a studio and expectations of future work as well as the immediate experience. Capcom and FromSoftware illustrate his argument about accumulated expectations. Crucially, he also allows that an AI-based creator could build a new brand through distinctive taste and repeated work. The episode offers no consumer research measuring these motives. His account is a commercial hypothesis whose strength, distribution across audiences and persistence remain unproven.

4. Learning and Application

For a team adapting a traditional subject, the interview suggests beginning with a documented source and then explaining the transformation. Liang describes recording southern lion dancing and drums at a Guangdong ancestral hall before developing that material into kungfu-punk expression. A practical design record could identify the source, the movement or sound retained, and the emotional or interactive purpose it serves in the game. This makes the adaptation discussable across disciplines. One recording session cannot stand for an entire tradition, however, and fidelity to a source need not require literal reproduction of the event.

Character equipment offers a second review method. Liang explains Zuo Shang’s gold sword through fictional killing energy, giving its impractical material an expressive purpose: luxury, confidence and arrogance. Soul’s Heishang, modeled on a Tang horizontal saber, supports a rougher and more grassroots presence. Writers, artists and combat designers can each explain how an object advances motivation, silhouette and movement. Contradictory answers indicate that the character needs clarification before additional decoration. The gold sword remains a fantasy convention, not evidence about historical weapon performance.

Modernization can also challenge inherited presentation habits. According to Liang, Donnie Yen objected to making a master character sound prematurely old and solemn. Promotional sword work used modern leather clothing and red, neon-like lighting. The transferable question is why authority must be signaled through a familiar costume or age stereotype, and whether movement can establish it instead. Copying neon would miss the point. The approach works when the team knows which qualities it wants the audience to perceive; otherwise modern styling becomes another interchangeable label.

Creative control needs a corresponding correction process. Liang admits that fixation on details, repeated requirements and emotional reactions can exhaust colleagues and make the environment toxic. He advocates quickly admitting mistakes and withdrawing bad requests. Teams can make consequential changes explicit, record their intended benefit and give leaders a clear way to reverse them. Fast correction limits further waste, but does not erase work already discarded or emotional costs already imposed. Repeated reversals should therefore trigger examination of how decisions are made, rather than becoming a celebrated sign of artistic intensity.

Recognition requires comparable specificity. Liang says the studio consults colleagues over credits and tries to accommodate former employees’ requests for acknowledgment. Recording contributions throughout production can make those decisions easier and fairer to discuss. Yet his account also names compensation, expectations of success and scarce comparable jobs in Beijing. Credits should complement material conditions; the episode does not justify treating recognition as a substitute for pay or as the sole explanation for retention.

Finally, a creative company can evaluate a financing round, overseas partnership or new project by asking what capability it leaves for the present game and its successor. That operationalizes Liang’s stated aim of reinvesting income, experience, technology and relationships into better games. Apply the interview’s boundaries at the same time: distinguish gross billings from profit, assess technical readiness and cash flow separately from enthusiasm at a demonstration, and test AI efficiency, experience and trust as different questions. This is useful for a studio organized around sustained creation. It does not imply that another team should reproduce Liang’s financing scale, staffing model or preferred commercial format.

Source

• Original episode: 156. 《侠》｜对影之刃零制作人梁其伟6小时访谈