The Little Sisters That Almost Weren't: What Bioshock can teach us about GenAI and the Messy Magic of Creating

I know, that’s a title. But bear with me.

Let's talk about one of the most iconic duos in video game history: Bioshock's Big Daddies and Little Sisters. These characters are so ingrained in Bioshock's gaming lore that it's hard to imagine it without them. But here's the wild thing - the Little Sisters as we know them today almost didn't exist.

See, when the game designers first started sketching out these characters, they weren't cute* little girls in dirty dresses. They were slugs. That's right, slugs. The idea was that these slug creatures (then called “gatherers”) would crawl around the underwater city of Rapture, harvesting a genetic material called ADAM. Players would have the choice to either kill the slugs and take their ADAM or spare them.

But as the designers started developing this concept, they realized something: no one would care about the slugs. Like, at all. Testers would mercilessly slaughter these slimy little things left and right without a second thought. The moral weight the designers were aiming for just wasn't there.

So, back to the drawing board they went. They experimented with a number of more “sympathetic” concepts for a gatherer, to include dogs in wheelchairs. Not kidding. The thinking was that players would feel guiltier about exploiting or harming a helpless animal. But again, it just didn't quite click.

Finally, after countless failed drafts and late nights of head-scratching, the designers landed on the idea of using little girls. And suddenly, everything fell into place. Playtesters would agonize over the choice to save or harvest the Little Sisters. The emotional stakes were through the roof. They had struck gold.

Now, why am I telling you this story? It's not just because it's one of my favorite bits of gaming trivia (though it totally is). It's because the evolution of the Little Sisters is a perfect case study in the power of creative iteration - and the potential of artificial intelligence to supercharge that process.

Think about it: the Bioshock team spent countless hours and resources cycling through ideas, testing them out, and refining them until they found something that worked. But what if they had a tool that could generate hundreds or even thousands of variations on their initial concepts in a matter of minutes? What if they could feed their slug idea into an AI system and see dozens of alternate creature designs, each with its own unique emotional resonance?

That's the promise of generative AI as a creative collaborator. It's not about replacing human ingenuity or imagination. It's about giving us a tireless, endlessly inventive partner to riff with and bounce ideas off of. It's about helping us explore the full space of possibilities and zero in on the most promising concepts faster than we ever could on our own.

Now, I know what you might be thinking. "Sure, AI collaboration sounds great in theory. But what does it actually look like in practice?" Well, dear reader, I'll let you in on a secret: it looks like this blog itself.

See, I'm a marketing manager at Waymark (an AI company), and like so many of us, my days are a whirlwind of meetings, deadlines, and never-ending to-do lists. I've had this idea about Bioshock and creative iteration bouncing around my brain for days, but between the slide decks and the budget sheets and the conference prep, I just couldn't find the time or energy to sit down and write it out.

But then I thought, "Hey, why not practice what I preach?" So I fired up Anthropic’s Claude Opus, and started riffing on the idea with it. I threw out my truly half-baked thoughts, and watched in awe as Claude spun them out into full-fledged paragraphs and arguments. It built on my ideas in ways I never would have thought of, pointed out strengths and opportunities, and helped me refine and improve my thinking at every step.

Now, let me be clear: this isn't about outsourcing my creativity to a machine. Everything you've read is the product of a back-and-forth collaboration. Claude was not the one who remembered the story of the Bioshock-Doberman-that-wasn’t on a Thursday night. But it was the instantly available thought partner that helped me riff on the idea and turn a fleeting spark into a full-fledged fire, all while juggling the million other things on my plate.

And that, my friends, is the real magic of human-AI creative collaboration. It's not about replacing human ingenuity, but augmenting it. It's about unlocking the full creative potential of folks who might otherwise be too bogged down by the demands of the daily grind to let their ideas shine. It's about democratizing the creative process and empowering more voices and visions to enter the fray.

Of course, we'll need to be thoughtful and intentional about how we integrate AI into creative workflows, with clear guardrails and human oversight. We'll need to grapple with the truly thorny and valid questions around attribution, bias, and the specter of homogenization. But the potential upside is just too real to ignore.

If a frazzled marketing manager and Claude can riff their way to something halfway decent in between walks with her energetic rescue dog and the ever-growing laundry pile, just imagine what the true creatives among us might achieve. 

In the end, the story of Bioshock's Little Sisters is a testament to the messy, miraculous power of creative iteration. And with the rise of generative AI, we've got a new tool to help us harness that power like never before. So let's embrace the potential, stay open to the possibilities, and see where this wild ride takes us. Who knows? We might just invent the next Little Sister along the way - with a little help from our AI friends.

___

‍

*Okay, the Little Sisters were eventually cute.

‍