The year is 1999. You’re gripping your PlayStation controller, eagerly waiting for that black screen to fade. Then it happens—a strange scene appears. A beach. A montage of surreal landscapes. A haunting choir sings in a language you don’t understand. The camera glides across a field of white feathers and rose petals. Squall and Seifer clash, their “gunblades” sparking as the music swells to an operatic crescendo.
More than just a mood-setting cutscene, this iconic FMV was a technical showcase by SquareSoft. It was a full-blown cinematic experience, brought to life with a multi-million-dollar budget and a massive team of animators and developers.
Could today’s generative AI recreate that same moment without the big investment? Absolutely. But you already know there would be plenty of caveats along the way.
The Visual Birth of Liberi Fatali
Final Fantasy VII was reported to have a whopping $45 million development budget when it was created. As such, you can definitely imagine a similar amount of funding when it came to SquareSoft’s next big-numbered FF title. This level of financial commitment to the burgeoning technology of 3D graphics during the late 90s was reflected very well in the ambitious nature of the FMV sequences featured in Final Fantasy VIII. Visual Works, the image studio division of the company, continued where FF7 left off, and aimed even higher. More details, hand-keyed animations, and motion capture, all backed up by sophisticated proprietary technology.
It was clear, based on the art style and proportions, that FF8 was taking a huge leap towards realism. Zell and Selphie had much more in common with earlier conceptual designs of Yoshitaka Amano for FF4 and FF6, but have since then evolved as actual 3D models for the FMV scenes.
Take note—every frame back then was pre-rendered. Much like animated films of the era, each scene required hours of digital computation, processed on hardware that cost more than the average used car.
This wave of technological ambition peaked with Final Fantasy: The Spirits Within in 2001. Despite its visual achievements, the film cost more than it earned globally and temporarily derailed Square’s Hollywood aspirations. But the rendering breakthroughs weren’t wasted. Instead, they were folded back into game development, improving real-time graphics for future titles. Lessons from the film’s advanced motion capture and rendering pipeline directly enhanced later FMVs, especially in games like Final Fantasy X, where real-time and pre-rendered elements started to blend more seamlessly.
Can Generative AI Do It? Yes, And No.
Fast-forward to 2025, and the “democratization” of these once-impossible techniques (by the average Joe) has been staggering. The creative and technical skills required are still there, sure. However, the financial barriers and manpower requirements have largely been overcome. You no longer need millions of dollars and teams of specialists. A single creator orchestrating the right suite of AI tools can recreate similar cutscenes with very acceptable professional results.
Video Generation: Impressive but Inconsistent
Runway ML’s Gen-4 ($15/month) and Google’s Veo 2 represent the current state of the art in text-to-video generation. Both can produce genuinely cinematic moments, namely sweeping camera moves, atmospheric lighting, and convincing character motion. However, maintaining visual consistency across multiple shots remains challenging. The AI might nail Squall’s brooding expression in one frame, then subtly alter his facial structure in the next. A no-go since some of the original shots require extreme close-ups.
Motion Capture: Genuinely Revolutionary
Move.ai’s markerless capture ($18/month) is perhaps the most mature technology in this stack, at least in our opinion as of the time of this article’s writing. Single-camera smartphone footage reliably extracts professional-grade motion data. Needless to say, very long or complex interactions, as well as rapid movements, still require manual cleanup. But for basic character animation, it can actually work relatively well.
VFX Integration: Promising with Caveats
Wonder Studio ($50/month) automates much of the tedious rotoscoping and compositing work. But if we are to superimpose FF8’s near-alien architecture and material designs, the complex lighting scenarios and unconventional character designs might just break it apart. Can you imagine going beyond the intro FMV and rendering something like the scene launching Sorceress Adel into space?! Well, at least this particular AI tool excels at standard humanoid figures in well-lit environments, perhaps exactly what we need for recreating Squall and Seifer’s duel.
Sounds: Do We Need AI Here?
PSX-era Final Fantasy games remained completely unvoiced. So aside from reusing Liberi Fatali and mimicking the sound effects of the fight, we probably shouldn’t even have to resort to AI here.
TOTAL MONTHLY ACCESS: under $100. And the quality comparison? Surprisingly competitive-ish for individual frames elements, noticeably weaker for artistic cohesion once things start to move and scenes start to shift. The tools are legitimate, and the savings are real. But expecting to match Square’s unified vision requires more than just better pricing at this point in generative AI development, despite the semi-coherent meme videos (most of which only last a few seconds long before switching to another clip) we see on YouTube as of 2025.
The recreation of static FMVs with models different from the actual in-game characters (of the original) is just the beginning. What if these same AI tools could make those characters interactive?
Talking to Squall Interactively While Still on Script?
Back in 1999, Squall just brooded outside and panicked internally. In 2025, he might actually answer you back. No, this isn’t a teaser for a subscription model, but a reflection on how far we’ve come with AI companionship platforms. Think of it as the space between static FMVs and genuine, reactive interaction.
Platforms like Candy.AI attempt to simulate emotion and memory through dynamic dialogue. Others, such as Syntheia and HeyGen, push further into the uncanny valley, generating photorealistic avatars with lifelike gestures and responsive motion.
The original Squall of 1999, seen through carefully crafted FMV sequences, stood at a visual crossroads—not quite lifelike, but proportioned and animated just enough to feel grounded. Today, we could train an AI on Final Fantasy VIII’s script and behavioural context to mimic his thoughts, emotions and iconic internal monologues.
Not enough? Here’s a crazier idea: the same motion-capture data used to animate Squall’s iconic gunblade stance could also power an interactive character—one that remembers your conversations and responds with an authentic personality. In effect, you’d be talking to a Squall from a parallel universe. Still fictional, still digital—but uniquely yours.
The idea of FMVs evolving from passive viewing to real-time relationships with characters might sound uncanny, even unsettling. But within the context of continuing Final Fantasy VIII’s narrative, it might not be such a bad thing.
Instead of just rewatching the Balamb Garden ball scene, you could ask Squall to dance. Instead of listening in on his internal monologue, you could become a part of it.
How Would YOU Feel About AI Recreating FF8 Today?
Everyone remembers the first time they saw that FMV. Maybe you were a high school student, or a grade schooler visiting a friend whose older brother had picked up Final Fantasy VIII on release day. We crowded around a 27-inch CRT television, controllers forgotten in our laps, as Liberi Fatali filled the living room and Squall’s scar appeared in perfect detail across his face. It felt like we were witnessing the future of entertainment, leaving behind the blocky confines of Virtua Fighter and Tekken 2 for something magical. Something epic.
Today, that same feeling of seeing the future is possible again—but this time from the creator’s side. Thanks to generative AI, the level of manpower and funding once needed for cinematic sequences like that has dropped dramatically. At least in theory, we now have access to tools that once belonged only to multi-million-dollar triple-A studios. While the tech is still in its infancy—and the skill required to wield it remains high—it’s clear we’re at the threshold of something big.
Of course, there will be purists. Fans who swear by the artistry of Final Fantasy VIII’s original FMVs—the way Squall’s hair caught the light, how Rinoa’s dress moved with the music, or how each feather descended at precisely the right moment. They’ll say AI will never capture those nuances. It can’t understand what makes them beautiful. And maybe they’re right—for now.
But even those skeptics can’t deny what AI tools are beginning to offer. These next-generation engines of creativity are just getting started. And as they mature, they might not just replicate our favourite scenes—they might help us create new ones we never thought possible.
Indeed, the question isn’t just whether or not we could recreate the Final Fantasy VIII opening with AI tools. It is simply a matter of percentage and of time. Maybe what we should be asking instead is if AI tools will eventually be at the forefront of the digital cinematic stage, and whether that transformation will preserve or diminish the artistry that made these moments so memorable.
Based on what we can see so far, it is both promising and dreadful.
