Two weeks ago, I spent a Saturday making a low-poly first-person brainrot shooter featuring Skibidi Toilet and Tralalero Tralala.
A few days later, my name was trending on X, my game was getting news coverage in different countries, and my vibe-coding project had generated 5M+ impressions across the Internet.
The way this story exploded was completely unintentional. And the way it was told isn’t entirely accurate, either.
Here’s how I made a viral game, accidentally became a case study for out-of-control AI spend, and helped my company add $100M in AUM in one week.
Building the brainrot shooter
Earlier this month, Slash held its annual company offsite in Hawaii. During some of the demos on the trip, I saw firsthand the insane ways people on our team were using AI. I was inspired. When I got back home, I decided to go all-in on becoming an “engineer”.
I started simple: I vibe-coded a portfolio website for myself to practice the new skills I learned in Hawaii. It took me an entire weekend, and the people I sent it to had some hard-hitting feedback. One friend said he would “barely even consider it vibe coding” because of how simple it was.
I needed to go further.
That Friday, I was cracking brainrot jokes on a walk to lunch with our CTO and another coworker, and I mentioned wanting to make a brainrot-themed vibe-coding project. We laughed about it, but then I doubled down: I was going to make a game to really test my skills.
This wasn’t the first time I’ve wanted to make a game. When I was twelve I made a Roblox game called “Duck Simulator” where you spawned in as a duck and collected fish to unlock new worlds. It never got more than 1000 players, but it was fun to make. I made a few more Roblox games, tycoon simulators and obstacle course games. Together, they earned me a decent chunk of Robux.
The next morning I fired up Cursor and started planning. It was what twelve-year-old me dreamed of: building a real game just by telling a machine what to do. I gave Cursor as much info as I could on the cast of characters — Ballerina Cappuccina, Tung Tung Tung Sahur, Skibidi Toilet — and all sorts of instructions on how to play, what it should look like, how the powerups should work.
An hour later, Cursor spit out the first version. It was terrible.
So I sat at my desk for the next 12 hours and kept prompting, tuning it until it felt decent.
At the end of that Saturday, I sent a demo link to some friends. Several said the game was really fun and spent 30+ minutes fighting through the waves. I was surprised by the feedback, so I kept working on it. I spent Sunday fixing bugs, optimizing the browser build, and figuring out PeerJS so people could play together.
Sunday evening, I posted a link to Brainrot Shooter on X so my friends and followers could try it out.
Tokenmaxxing 101
I thought that would be the end of it. Fun game, friends thought it was funny, done. Then our CEO, Victor, sent a Slack message the next day: OpenAI had flagged “unusual activity” on the Slash API key. One engineer outed me as the culprit; another joked I’d spent $100K on tokens making my game. I tried to defend myself, but the joke stuck.
The whole company was talking about it Monday morning. “Nick blew the AI budget on a silly game.” It became the meme of the day at Slash, so marketing decided to have fun with it. At 1pm on Tuesday, June 23rd, the official Slash X account posted: “We encouraged the company to start vibe coding more but @nickbruhman burned $80k in credits on a brainrot shooter. Pls play it so we can write it off as a marketing expense.”
To back it up, I quote-posted a screenshot showing $81k of a $100k AI budget spent. I went with it. Honestly, I wanted more people playing my game.
We figured the stunt might get some likes. We had no idea how far it would go.
Later that day, someone dropped a link in the company Slack. It was a Polymarket X post: “Fintech startup Slash reveals it was forced to roll back its AI coding push after an employee burned $80,000 in tokens in the first week”.
Within a couple hours, the post had over 100K views.
I immediately doubled down. I started working on the game again, optimizing it for mobile and making some gameplay fixes. I posted back-to-back: first a list of game fixes, then a response to the Polymarket post.
Messages started rolling in. People were commenting that they loved Brainrot Shooter, or they were bewildered that I had spent so much money on AI tokens. It was gaining momentum.
The next morning at 1AM, Business Insider published a story about me.
I read it and immediately realized that this wasn’t just about the game anymore. I was a case study for how AI spend can get out of control when given to the wrong person, and the story spread like wildfire. The piece was syndicated to Yahoo News; it was covered in American Bazaar; my name was in newsletters, Instagram Reels, Facebook posts, and TikToks.
By Wednesday afternoon, I was a trending topic on X.
Turning coal into diamonds
It was a frenzy. I was getting messages from all over the internet; companies I’d worked with reached out to learn what happened; childhood friends checked in to ask if it was real; hundreds of outlets in India, England, China, Russia, and more were covering the story.
Truth nuke: I did not spend $80k in AI tokens creating Brainrot Shooter. I spent $80 (after some bug fixes and tweaks later on, more like $200). I doctored the image of my $80k AI bill using inspect element. It was just a joke that blew out of proportion.
For a moment, we considered backing off and clearing the air. I talked it over with marketing the next day, but we decided to keep having fun with it instead.
We ran with the new idea that we needed to cover the game’s cost. I reached out to some of our good friends at other startups in SF and asked them if they’d like to “buy ad space” in the game (We didn’t actually sell ad space. We just wanted to include them in the fun).
Melius, Bland AI, Backpack, and Atomik Growth were into it, so we added their billboards to the game and posted that we'd struck an 'advertising deal to offset the costs.
There were real deals involved, too. With Slash all over the news, people started reaching out who’d never heard of us before.
The week before this happened, Slash crossed a major milestone: $1B in AUM. A week later we had added another $100M – Brainrot Shooter had a lot to do with that.
By the numbers: 5M+ impressions, nearly 20,000 player sessions, and over 10,000 unique players from a game that cost me about $200 to build.
The story that got picked up by the news tapped into the anxiety that AI is a threat. If you give unrestricted AI access to anyone, your company could accidentally blow through its budget.
That’s the opposite of what happened.
I have zero background in game design or coding. I spent about $200 and some free time on a weekend and ended up with a game that twelve-year-old Nick would be proud of. That $200 resulted in tens of millions of dollars in added business for Slash. I’d say that’s a good deal.