LightSpeed Studios Q&A: “Designing for 100 Million People Pleases No One”

At GDC 2026, LightSpeed Studios, a subsidiary of Tencent, unveiled its commitment to creating original IPs through a new initiative that leverages world-class talent, including Creative Director Feng Zhu and Motion Capture Studio Manager Kristin Gallagher.

Known primarily for PUBG Mobile and Final Fantasy XIV Mobile, the studio now has several AAA projects in production, including Last Sentinel (an open world action game announced at TGA 2023 and set in a dystopian version of Tokyo) and an action game in development by former Dragon's Dogma and Devil May Cry designer Hideaki Itsuno, who joined LightSpeed in late 2024.

We had the chance to interview LightSpeed about their new original IP initiative, the challenges of transitioning from mobile development to triple-A games, and the benefits of owning a big performance capture facility in Los Angeles.

In this LightSpeed Q&A:

• The Challenges of a Global Studio with AAA Ambition
• Development Goals: Why Designing for 100 Million Pleases No One
• The Benefits of Owning a Mo-Cap Facility

The Challenges of a Global Studio with AAA Ambition

Lightspeed Studios has historically been known for mobile and live service games. Building AAA original single-player or narrative IPs is a very different creative and commercial bet. What kind of challenges did you encounter, and what does success look like for this new framework?

Feng Zhu: This is an insightful observation. Our strength in mobile and live-service games has given us deep expertise in real-time player engagement, iterative development, and operating at a global scale. The core challenge in building original narrative IPs isn’t a lack of technical skill, but a shift in creative focus and production discipline.

The primary challenges were:

Establishing a Systemic Foundation: We needed more than just talented individuals; we needed a reproducible, high-quality pipeline. This led to the development of our proprietary IP framework, which formalizes processes from initial concept to final polish, ensuring consistency and ambition at scale.

Cultivating a Long-Term Vision: unlike live-service titles that evolve with seasons, a foundational IP requires a cohesive, forward-looking vision from day one. Success for this new framework isn’t measured solely by initial sales, but by: 

• Whether we can consistently produce high-quality original games that stand the test of time.
• If the worlds and characters we create resonate deeply enough to sustain franchises.
• And ultimately, if we build a trusted, signature style that players globally associate with Lightspeed’s original IPs.
• Fundamental to all of this is our ability to continuously attract and empower top creative and technical talent from around the world to build this future with us.

How does operating as a global studio, with teams across China, the US, Singapore, and elsewhere, affect the creative process for a culturally specific Chinese heritage IP? Is cross-cultural tension within the studio a design tool, or something you try to resolve?

Feng Zhu: Operating as a global studio is our greatest asset in this endeavor, not a hurdle. When developing an IP rooted in cultural heritage, a diverse team acts as both a precision tool and a creative lens. Our team members with deep cultural context ensure respect and authenticity in the foundational elements – the “90%” rooted in real-world references we discuss in our philosophy. Finally, colleagues from other backgrounds ask the crucial questions that a purely homogenous team might not. They help identify which elements are universally compelling and which might need reinterpretation to connect globally. This isn’t “tension to resolve,” but creative friction to harness. It pushes us to move beyond cliché and find the deeper, more human core of a story that transcends its specific cultural origins.

How do you feel about the substantial challenges the global industry is facing right now, which manifest in layoffs, studio closures, etc.?

Feng Zhu: It’s a difficult period that we observe with great empathy for everyone affected. It underscores the importance of sustainable creative and business models.

At Lightspeed, our response is to focus on long-term health and responsible growth. Our investment in original IP and foundational technology like the mocap studio is a bet on the future. We believe that nurturing talent, empowering creators with robust tools (like our AI-assisted pipelines), and building games with lasting value is the path to resilience. Our commitment is to our people and our projects, aiming to create a stable environment where ambitious ideas can mature.

What is LightSpeed's stance on AI tool usage? Are you aware of the general backlash that studios have faced for embracing it?

Feng Zhu: Our stance is pragmatic and human-centric: AI is a powerful tool to augment our creators, not replace them.

We are actively exploring AI in areas that enhance productivity and creativity, such as procedural content generation, animation assistance, and data-informed design. The goal is to free our developers from repetitive tasks, allowing them to focus on high-level creative decisions, emotional storytelling, and unparalleled polish—the very things that define a great game.

We are aware of the industry discussion. Our approach is to implement these tools transparently and ethically, always with the intent of elevating the craft of our teams and the experience for our players. The final creative vision, authorship, and responsibility remain firmly with our human developers.

Development Process: Why Designing for 100 Million Pleases No One

LightSpeed is presenting the new Game Development Framework at GDC, modeled partly on Hollywood pipelines. You've worked with directors like George Lucas, James Cameron, and Michael Bay: what specifically from that world translates into game development, and what fundamentally doesn't?

Feng Zhu: Working on films is actually quite different from games. Films are a single-view experience from the eye of the director. The director makes the entire call on how the audience experiences it. The audience doesn't participate in the film; they're just passive. In games, the audience is creating the experience themselves. We provide the playground, but the experience differs for everyone. Even in linear games, you can't control exactly what the player decides to do. The pipeline diverges in gameplay design. IP creation (worldview, characters, settings) starts similarly, but the game type (FPS vs RPG vs RTS) heavily impacts IP direction. Films are a tightly controlled two-hour experience, while games prioritize gameplay as the core. That's my pipeline philosophy.

The 90:10 philosophy of 90% real-world grounded foundations and 10% creative reinterpretation sounds almost like the inverse of how many Western studios approach fantasy or sci-fi worldbuilding. Was that ratio arrived at deliberately, or did it emerge organically from working on Chinese cultural IP?

Feng Zhu: Big global IPs follow a 90-10 or 80-20 ratio. 80-90% of content comes from the real world. Look at Iron Man 1 or Jurassic Park – even dinosaurs are rooted in reality. Sci-fi projects with heavy original creation are exceptions and extremely difficult. We prefer IPs with 80-90% groundedness. It simplifies design with existing references and controls timelines/budgets. Reversing the ratio (90% original + 10% real) makes projects risky and chaotic.

You spent years doing concept work for blockbuster films, where your contribution goes essentially uncredited to audiences. How does moving into a Creative Director role at a studio, where the creative vision is attributed and accountable, change how you approach the work?

Feng Zhu: My goal is to create good games – story, worldview, music, all matter. Credits don't concern me. Hollywood taught me anonymous collaboration. What matters is players picking up controllers and finding joy. As a kid, I never cared who made the games I loved; I cared if they were fun. That pure focus on the product has driven me for 30 years. Distractions like fame dilute the true goal: making great games.

At FZD, you've always emphasized that you can't design something new until you first deeply understand what already exists. How does that educational philosophy map onto what Lightspeed is doing with culturally rooted game IP? Is the framework essentially a formalized version of what you've been teaching for 15 years?

Feng Zhu: Design is like writing or math – you must master fundamentals before innovating. Even fictional game worlds require experienced teams to simplify complexity into intuitive experiences. Great IPs tell universal human stories transcending cultures. Player confusion arises when designers don't deeply understand their own systems.

Games like Black Myth: Wukong and Phantom Blade Zero have shown an enormous global appetite for Chinese cultural heritage in games. Does that set the expectations very high for you from the get-go to match that level of success and hype?

Feng Zhu: As a creator, I don't anchor IPs to specific cultures. First, build strong stories/characters – cultural elements emerge naturally as background. Forcing cultural inserts destroys IP authenticity and leads to generic products. Focus on core gameplay: if one player loves it deeply, that passion can ripple to millions. Designing for '100 million people' often pleases no one.

The Benefits of Owning a Mo-Cap Facility

The ability to isolate a single hero performer among 13 simultaneous MoCap artists using Vicon is a pretty specific and impressive capability. What problem does that solve in practice on a game like Last Sentinel? Is it primarily about crowd and combat scene efficiency, or does it open up fundamentally different kinds of choreography?

Kristin Gallagher: This was mostly for testing our capabilities on our shoot floor. When a shot comes to us, we want to know our limits and abilities. I think this helps us coordinate creative intentions with the development team. It shows that our system is impressive, but also gives us information on how to handle this type of move in our post production. Processing a move like this took a considerable amount of time. We are always trying to manage speed/quality/cost (fast/good/cheap, pick two!). This helps us discuss with the development team their priorities, and help them know what they can expect back. 

450+ complex shots in a single day is a remarkable figure. What does "complex" mean in that context, though? What separates a complex shot from a standard one in your pipeline, and what were the bottlenecks you had to solve to hit that number?

Kristin Gallagher: If we have a shoot day of 450 takes, these are likely NOT Complex shots. Complex shots often take more time. This particular day was a locomotion shoot, meaning most shots were to accomplish a gameplay action in multiple directions. 

A complex shot usually means one of two things: 

One, there is a lot of occlusion of markers, and it will be complex to solve the data. Or two, the shot requires choreography and has to be shot multiple times to get the perfect cadence (often the case in Performance Capture scenes).

MoCap has historically been a late-stage production tool. You know what you need, you capture it, you implement it. How do you integrate MoCap into game design at a more foundational level? Are there moments where what you capture in the MoCap studio actually changes narrative or gameplay decisions?

Kristin Gallagher: This is the biggest benefit to having mocap as a tool for game developers in the pre-viz phase. It is often only done when a capture studio is there for the use of its developers at a highly discounted cost. Mocap at a commercial house is very expensive, which is why it's often used as a late-stage tool. Some ways to solve this is to have a rehearsal day where we record video reference, and then shoot the capture the following day. Stunt teams often employ this method to work on choreography with the director. But with Lightspeed studio teams, we have already shot many previz shoots. These are often delivered much faster, since data quality is not a concern, but blocking out the scene is key.

Thank you for your time.

Follow Wccftech on Google to get more of our news coverage in your feeds.