SORA 2 EARLY ACCESS EXPERIMENTS – GAMES

(compression does not represent the raw footage of Sora 2. There’s also a glitch in my editing software that is creating audio imperfections. Will update as I can!)

▫️Project: Stress test Sora 2 while it is under Early Access and Invite Only.

▫️Experiment: How comprehensively does Sora 2 generate a commercially viable scene from a single sentence prompt.

▫️Tools: Sora 2 (Text to Video), DaVinci (editing). 

▫️Vision: Create a reel of diverse gaming genres that evaluate the platform’s ability to act as a pre-vis tool for gaming studios. 60 second video, with six 10-second clips. Landscape formatting only. 

▫️Expectations: I was both impressed with Sora 2’s power and miffed with its lack of control, so entering gaming prompts left me bewildered as to what I would get. I figured, like many other Ai platforms where you try game prompts, there would be a lack of detail on the HUD, but a fundamental adherence to “gameplay” on the screen. I imagined a kind of game “dreamscape” where things are hazy but the core concept is strong.   

▫️What actually happened: Just like I imagined, many of the clips were nonviable, most especially sports games. There are just so many fine details and logical consistencies that make up even the smallest gaming clip that I feel like even the mighty Sora 2 struggled. Non-gaming generations at least had the benefit of being surprising on their own merits. But games follow a kind of visual calculus that is so hard to replicate without hand curation. 

▫️What I didn’t expect: No, you can’t make a whole game with Ai yet. But the implications for previsualization is off the charts. The world of Venture Capitalism will be turned on its head  because even the smallest Studio can approach them with solid mockups and gameplay targets to shoot for. The internal usage of this tool for game studios is immediate and profound. 

▫️Results: Sora 2 obviously can’t make a whole game from a single sentence prompt, but the amount of context it knows of every genre is staggering. The value of integrating this into every studio for previs and asking for funding is immediate and obvious. Even for raw ideation for games, this is the best “imagineering” tool for the job. 

🔴If You’re Anti Ai: Sora 2 still isn’t great with fine details, which is really all a game is collectively. You can successfully point out how that lack of fine quality control makes the platform a non-starter. Point out the stunning amount of work it would take to polish a Sora 2 generation, and position yourself as the necessary evil they’d have to use anyway.

🔵If You’re Pro Ai: Get on the Waiting List today. I don’t know how many people are experimenting with games, but it can’t be that many. See just how much fidelity you can push to create a pre-vis for multiple genres and scopes – or even blend genres and IPs. Game studios will probably want to hire you based on your ability to help them see the finished product as early as possible.