AI couldn't make Star Wars: Biddy Blas from Galactic Racer

Epic's upcoming Star Wars shooter is reportedly built with an "AI-first" production model, meaning it will never have its own Biddy Blas.

This is an IGN review by Matt Jones. He loves weird kids in video games.

If your social feed is anything like mine, it's been overtaken by a weird little alligator man knee-deep in Taunton with an unresolved rage twelve times his size. Her name is Biddy Blas, she's one of the supporting cast of Galactic Racer, and she's the best new thing to happen to Star Wars since Andor.

Give the AI as long as it needs: it couldn't make Biddy Blas.

With each new day, the ground beneath video games as a medium where professional creators are valued and able to negotiate an honest living becomes more unstable. Perfectly good studios are being shut down in the name of rebranding. The price of electronic components has no incentive to decrease. The cost of producing anything is rising due to the unstable supply of energy and the pressure to end physical ownership. It is difficult to imagine what the future of this medium will be without considering the role of artificial intelligence in it.

But look, I'm not a doomsayer - games exist despite these facts because of the passion of the people involved. They are also often better than ever. Star Wars: Galactic Racing is a highlight for the underused genre of high-speed arcade racing, with deep vehicle customization, heat management and opponent competition systems.

It's hard not to think about how good it is in the context of Epic Games' apparent push for an "AI first" production model. Such a pivot has reportedly delayed a Star Wars stealth shooter and Marvel crossover, which will reportedly be developed with AI tools in both the coding and art departments of the project.

I can't speak for the quality of a hypothetical game that's an indefinite amount of time away from release, especially one that no one outside of Epic's offices has seen a frame of, but I can't see how this approach can mimic Galactic Racer's secret weapon. For all the efficiency, accuracy and speed arguments, I don't think a large language model is capable of creating a break character. AI can't make a Biddy Blas.

Blass is a symbol. He is the perfect example of what Star Wars fans have come to call a "cheetah rose." It's not important enough to the narrative that it's the main character, but it's interesting enough in its limited screen time that it becomes your favorite anyway.

At this point you might be thinking, "Surely a bit of computer code can come with an angry alligator in a hat?" And maybe he can. Arguably, the main thing that makes Blas so great is his ridiculous voice, and AI companies have gone to great lengths to make sure their products can do the voice. But, as The Blood of Dawnwalker lead actor Will DiRenzi Martin said in a recent episode of My Perfect Console, "The least important thing about what I do is my voice."

"[The intelligence] doesn't replicate my personality and what I do when I'm in the room, and the choices I make. The reasons why a director or a producer would want to work with me." Coen, who plays lead Dawn Walker, said: “Actors are very cheap in big projects. And if you end up with an AI that can make the right sounds, you're going to need someone to interact with it and attach the functionality you want to those sounds—and they're going to be a lot more expensive than an actor.

To put it another way: an AI might be able to write a hootish line, but only the special writers of Galactic Racer could decide that one of those sounds was "Scrib Doon," and only Biddy Blas musician Richard Summers-Calvert could make the inspired decision to deliver an equally ridiculous line. And it's that combination of creative human decisions, that vivid understanding of what's really funny, that turned Biddy Blas from a collection of scripts and polygons into an icon.

Certainly no one can prove otherwise. I don't see the same outpouring of love happening if we leave seemingly small decisions like these to computers. As an anecdote, there's no way my partner would be interested in having a Star Wars racing game playing on the home TV. Biddy Blas is a completely different proposition.

Does AI-generated content in games bother you?

Perhaps Epic's new production pipeline makes sense for a company that, over the past decade, has turned Fortnite into more than a game and a platform, increasingly eschewing creative talent and originality as a driving force (a far more rude critic says that was never the ambition). Today's Fortnite has primarily become a hub for other brands to place their licenses on. If Epic's priority is to focus on simplifying the production process to stack what's already there, an "AI first" approach might seem obvious. If you're building a Star Wars or Marvel shooter, there's plenty of reference material for both of those features to train your AI. I'm simply arguing that there was no reference material for any of them...before they existed.

It's not just Star Wars that needs weird little guys made by Actual Cool People. Following its latest patch, Valve's hybrid MOBA-shooter Deadlock will be inspired by the exciting creativity of the new inhabitants of the cursed city of New York, which is increasingly understood graphically. Characters like the Mouse King, who has been written and performed hundreds of times individually naming one of his rodent subjects for each use of one of his abilities, or Solomon, the chess player, whose weapon is a page showing historically important endgames chosen by the design team and changing depending on the maximum amount of ammo in his magazine. Without people making deliberate decisions, having completely original character ideas, people wouldn't have the same motivation to log in daily. To go 0-8 in lane. (Although I hate being in a game with these new powerful characters, so maybe this is actually a bad example.)

I see the continued use of AI in ideation and early production as an existential issue for a medium we'll one day move away from, in the same way that it seems obvious to executives to pursue the seemingly inevitable adoption of live services, online licenses, and on-disc DLC. For all the efficiencies in the process, for all the promise of profitability, I'm not sure how you can sell any of these things without sparking a new unique. I'm talking about this from a completely cynical marketing perspective: if I'm going to get stuff I've already seen, it's stuff I already have.