Perhaps you’ve heard the news over the Christmas holidays: OpenAI has finally achieved Artificial General Intelligence!
According to every desperate LinkedIn and tech blog grifter, that is.
Of course, the hype is far from true. While AI isn’t about to do all your taxes, pay your bills, take out your trash, take your job, and then eliminate you for being a waste of resources (yet), it’s still pretty stinkin’ cool. Or the death of all good things. Depending on who you ask.
Since you happen to be reading my blog, I’ll subject you to my own opinion. It absolutely empowered me to create the most ambitious Santa project for my kids to date. I don’t know that I’ve written about it, but each year I like to make my kids a Christmas present. It started in 2020 when I got a 3D printer and made custom toys for the kids. From teacups to Frozen to My Little Pony to Pokemon, there was always something to create for them, and it became something they looked forward to each year. I’ve made a couple of rudimentary Christmas-themed video games as well. But this year, I did something quite a bit more ambitious…
The Wild Robot books and movie quickly became a family favorite of ours this year. The only issue? No toys. No games. Zero. Merch.
But my youngest daughter needed the Wild Robot board game. That doesn’t exist.
As I was thinking how I could best disappoint my kids with the bad news that DreamWorks totally dropped the ball on licensing out that sweet, juicy IP, my friend asked if I’d join him for a game night. I’m not in the habit of turning down game nights, so I obliged. That night, we played several games I could never play in front of my children, Escape the Dark Castle being one of them. It’s a fun game, but the monsters on those cards would give at least one of my daughter’s nightmares until Easter. Of 2026.
Still, the game stuck with me. Each player picks a character. Then, you shuffle fifteen scenes into a deck and turn them over one at a time. Each scene has a challenge, obstacle, or choice of some sort. Maybe you’re battling wolves, bartering with a merchant, or trying some shrooms and hoping not to die. Whatever the case, each player rolls their character dice and uses item cards to make it through each scene. It’s all done together as a team, an all-for-one and one-for-all thing. Either everyone wins, or everyone loses.
The enjoyable mechanics and focus on teamwork-based survival scenarios fit the themes of The Wild Robot really well. They just needed to be tweaked. Made less dark and foreboding and more… Woodland creature cute. With some robots.
The only problem? Modeling two toys or making a one-level video game were things I was used to. Sure, I could crank out flavor text. Create scenes that were more like the Wild Robot. Steal the mechanics. Print the dice in my now-ancient 3D printer over the course of a week (it’s really slow). But the artwork? For over 60 cards? No way. Impossible.
To do that, I’d need a tool to quickly generate good-enough art. Too bad that doesn’t exist, I told myself as I prompted DALL-E for the fifty-second time that day for a meme reaction to a work conversation over Teams…
BINGO!
As a disclaimer, I would NOT recommend using AI images for a commercial project you plan to sell. There are a lot of ethical issues with AI and art these days, but you’ll quickly find out what penny-pinching companies replacing their art departments with AI somehow haven’t. AI might get you 75% of the way on a good day, but trying to use it to get the rest of the way to the finish line for a professional and cohesive product isn’t going to happen. Trying to coax exactly what you want out of AI image generators often leads to an overwhelming amount of frustration.
This isn’t a unique caveat just for art generation, by the way. I’m a software engineer by trade and have been using AI to assist in coding activities for some time. AI is a wonderful accelerator, but it has real limits. It’s not going to write and deploy your software. It’s not going to write your book. It’s not going to replace actual artists.
But it is fantastic at helping with prototypes. Which is as high of a quality as I need for a game for my kids and me to enjoy. And it looks better than clipart or the stickmen I’d draw. So, I went all-in on AI art for this particular project, and it paid off incredibly well. Just look at this beautiful card art!
Now, for those of you who want to do something like this for your family and friends - or folks just curious about the process, I’ll break it down for you here!
To avoid an obnoxious amount of repetition, I started by building a digital template for my scene and character cards (which have the same basic layout). As there are 55 cards, I wanted to ensure the design was uniform across each card. Since I have a subscription to Adobe’s products, I selected Adobe InDesign to build the template. There are definitely plenty of alternatives for this. Word, Gimp, whatever you’re comfortable with. I like InDesign; you can set up a parent page that works as a template. You can easily set up the text and image areas, set fonts, margins, bleed, and more. Admittedly, that’s probably overkill for most folks working on an occasionally operable printer out of the house. And I’m one of those people. Still, I like fancy toys.
Using Escape the Dark Castle as a base design, I tried to create a similar feel. Escape the Dark Castle has a very particular look, but I wanted a more hopeful theme. Something more open and free. I decided to have images stretch to the boundaries of my scene and character cards, which felt less restrictive and more loose. I also wanted to select fonts that felt closer to the Wild Robot, which has a unique artistic blend of sci-fi and nature. This took an embarrassingly long time, and the kids would probably have been fine with Comic Sans. But since that font is illegal in forty-six states, I had to take this route.
The next challenge was theming the mechanics and making the game just a bit easier—Roguelike with a little less risk. Crying kids on Christmas isn’t the end goal at the Bergin residence, but I didn’t want the game to be too easy, either. When I played with my friends, we ended up winning only because we didn’t read the rules thoroughly and unintentionally exploited healing. It’s pretty easy to lose Escape the Dark Castle. It shouldn’t be easy to lose at The Wild Robot.
To accomplish these goals, I started with the easier one. Mechanics theming. Dark Castle uses Wisdom, Cunning, and Might for character stats and challenges. For our game, I went with Survival, Smarts, and Might. Making the game easier? Now, I could go through and make each scene easier, but that’s a lot of balancing work. I wanted to do something simple and fun. After several nights of working through different options, I eventually settled on giving each character a special bonus on certain rolls. So, on top of whatever would normally happen in Dark Castle, you sometimes get something more. A bonus. A chance to shine. Kids love that kind of stuff.
Roz’s adaptive programming allows her to change her dice roll to whatever the player decides. Fink’s smarts let him roll an extra dice. Bright Bill’s not especially great at anything but has more heart than anyone else. He’s a deceptively tough cookie who has one more defense face on his die roll than anyone else, meaning he takes less damage. I gave each of the seven characters in this game a fun little mechanic like that, and it’s proven to be really fun for the kids and works really well with the themes in the games. They’ll often ask me to play Long Neck, a wise goose in the movie who sacrifices himself to save the flock. His special roll allows him to protect another player. My kids love not taking damage!
With the character cards done, it was time to start the real grind, at least on the digital side. For each scene card in Dark Castle, I created a corresponding scene card for my game. Some cards were already easily portable. Bats and wolves are in scary dungeon games and kid-friendly nature survival games. Others took a little rework. Turning thieves into raccoons is a pretty straightforward translation if you’ve seen The Wild Robot. But coming up with analogous cards for frost ghosts or centipede monsters? Those took a little more work. A lot of times, I’d just look at the dice requirements to beat each scene and come up with something that fit the theme of my game that had more to do with Survival, Smarts, or Might. It was a lot of work but a lot of fun in the end.
The other fun part about scene cards was writing the flavor text. You’ve got extremely limited space to establish a setting, a challenge, and a possible solution (or solutions). It’s like writing micro-fiction. Every word counts. Now, I did not use generative AI for this. I love to write and care very much about continuing to grow in my writing proficiency. Generative AI is great at producing content of a certain quality, but I don’t find it very helpful in improving your craft. Writing each card was a fun challenge that allowed me to stretch myself - by severely limiting myself. Lots of fun.
The other other fun part was the artwork I didn’t have to make myself. Yes, I do enjoy digital art, but it’s not my shtick. If I were to make a commercial game, I would outsource the art to an artist. But for my personal project, I was more than happy to outsource it to generative AI. The key to success in this pursuit was keeping a consistent theme by providing the same prompt context… And keeping my expectations appropriate. To get an idea of how I worded my prompts to get a sorta-consistent style across my cards was as follows:
Can you create a picture of [PUT SUBJECT HERE]? It should look like a mix of computer animation and a painting done with loose, free strokes. The composition should include the forest and natural lighting. The subject should be a bit stylized but not too cartoony.
Most of the time, that worked pretty well. For a few images, I used Adobe’s generative AI capabilities to fill in missing backgrounds and whatnot.
Now that all that digital work was done, it was time for the real work to begin. I wanted to create cards that would last a while and have a professional feel to them. Print and Play Hideaway’s Youtube channel has a great tutorial on how to make laminated cards, but if you want a rundown of my mostly-the-same process, and what I used, here it is.
Disclaimer: None of these are affiliate links.
Print cards on some thicc paper (I went with 92lb/250gsm). For my larger cards, I printed straight on 4x6 stock and did my smaller cards in sets printed onto 8.5x11 stock.
Initial cuts: Do initial cuts for the 4x6 cards. I did this so they’d fit onto my laminating sheets and because my printer refuses to print to the edge of the page. For all my cutting, I used the Fiskars trimmer.
Lamination round one: Run each laminating pouch through the laminator TWICE! If you don’t, you’re gonna have a bad time cutting. Or at least I did. You do you. I’m not your dad.
Second round cuts: Git rid of all the whitespace and cut the sheets of multiple cards into individual little guys. Or gals. Whatever.
Lamination round two: Run it through now, twice. This will bond the plastic better after your cuts.
Final cuts: Get yourself a rounded corner cutter and round off your card edges using the smallest rounding size. Feels like magic. Of the gathering.
Final lamination: You guessed it, TWICE through! Is this necessary? I don’t know, but I didn’t sleep for a few days and wasn’t up for experimentation or second-guessing the process.
So, then I was almost done. Except for them dice. I designed the dice faces using Adobe Photoshop and then created models of the dice using dicemaker.net. The supporting and printing process is pretty complex and probably enough for several blog posts, so we’ll skip straight to the fun part: sanding and painting!
This went a lot better than I thought it would. Using progressively finer sandpaper, I scraped each die face across each strip of wet sandpaper ten times. Or twelve. Something like that. By the end, all the print lines were gone, and I had some SMOOTH dice.
Next, I spray-painted them with white primer and almost waited for the primer to dry before hitting them with Citadel paints. Although I thought the insets would be a nightmare, it turns out that glopping white acrylic paint on your dry dice, and then sliding the dice face across a damp paper towel does the trick. Finally, I noticed some glitter paint my girls used for a different project. And there’s no WAY I was going to create a game for the girls with dice that didn’t glitter. They look FABULOUS!
Somehow, I managed to wrap up weeks of work by Christmas Eve and placed the kid’s custom-made Wild Robot card game where they could easily find it. I wondered if it would be a hit this year. Did all my work go to waste? Would they pick up a new Switch game from their relatives and never touch this game again? Take up MTG, thanks to their new Bloomburrow decks?
I’m happy to report that The Wild Robot card game has become a smash hit with the family and is our number one played game. We often see technology (like AI) as an unwelcome change or yet another way to isolate us from others. And that can certainly be the case. But when we use technology to make authentic connections with the people we love outside of a screen - that’s where the magic lies.
So what do you think? Has AI helped you in any of your creative pursuits? Do you have any games you’ve created or hope to make one day? I’d love to hear from you!
Comentarios