我如何靠 AI 和零編程技能,在 2013 年手提電腦上製作 Brotato 克隆版
How I Made a Brotato Clone on a 2013 Laptop using AI and Zero Coding Skills
作者用 AI 在 2013 年手提電腦上製作 Brotato 克隆遊戲,歷時 8 個月,加入主動技能和永久升級後上架 Google Play。Google Ads 以 15 日、每日 $2 的預算帶來 5,000 次下載,但不少手機無法運行遊戲,崩潰率大增。遊戲兩個月收入 $5,AI 訂閲及廣告成本分別為 $20 和 $50,作者估計約需 2 年回本。
It all started 8 months ago. I was 35 years old, staring blankly at my greasy Xbox controller with my Doritos-dusted hands. I thought to myself, "Dude, I've been playing games for 30 years (starting all the way back with bootleg NES clones)—why on earth am I not making one?" So I told myself: get up and finally jump into game development, something you've been procrastinating on for 10 years.
After all, we live in the tech age. You don't need to be Hideo Kojima to make a game nowadays. That "one-in-a-million super-nerd genius" trope felt totally unnecessary in this era. I immediately got up from the couch, wiped my greasy Xbox controller on the cushions, and sat in front of my 2013 laptop—an old machine with an NVIDIA GTX 660M GPU, a half-open chassis, a slightly cracked screen, and an air filter probably more clogged than a 1980 Volkswagen Golf.
Since I was obviously going to make GTA 6, I immediately downloaded Unity and Unreal Engine. But my veteran laptop took one look at those programs, fired up its fans at max speed, and basically treated me like a poor, ugly, unsmart guy trying to hit on a 10/10 girl—it straight-up looked down on me and gave zero response.
So I did a little research. My requirements were simple: the software had to run on a toaster, and it had to be free. But mark my words: FREE. Free as in $0.
That's how I met Godot. Plus, Brotato—a game I played for free on Game Pass—was built on Godot, and those guys made $17 million. A lightbulb went off in my head. Brotato! It looks simple enough, and it shouldn't take long to build. How hard could it be, right? Hahaha... looking back now, it took a whole 8 months.
I immediately started binge-watching tutorials. Everything related to Godot: Udemy, YouTube, Spotify podcasts, Brazilian soap operas, morning health shows on TV—whatever I could find. Slowly, something resembling a game started taking shape. But since I'm not one of those multi-billionaire coding nerds, I had zero knowledge on the programming side.
Then my friend came up with a "brilliant" idea: "Bro, there's this AI called Gemini, it'll help you out."
And that's when the torture began! Gemini... It only had one small flaw: being the absolute dumbest AI on the planet. But it had one incredible feature: it was practically free. We immediately started bonding over Google Antigravity like new lovers.
I'd say: "Do this," and it would reply: "Done, bro!"
I'd ask: "Did you do it?" and it'd say: "I handled it before you even asked, babe!"
I'd check: "What about that bug?" and it'd go: "Don't you worry your pretty little head, I fixed that too!"
Turns out... it was all pure hallucinations.
When I hit 1,431 debug errors, I started having my doubts. But what truly convinced me was when it gave me a dinner recipe while we were fixing code on my project. Turns out that was its actual superpower.
Anyway, I thought, "I guess this is how game dev is done." A little AI, a little bit of me, and a whole lot of luck. I kept at it. The AI broke things, I fixed them. It broke them again, I fixed them again. I learned by fixing. It explained, I learned.
Finally, a game like Brotato came out. Well... not "like" Brotato, it was an outright clone. I accidentally made the exact same game. Knowing that level of similarity would land me in court—and since I'm not a multi-billionaire divorcing his 50-year-old wife for a 20-year-old—I realized I needed to change things up. After all, I'm just a guy with an Xbox Series S and a 2013 laptop.
So I added my own spin. I remembered active skills from my old Dota days. I thought, "If I give my characters active abilities, the game will be way more fun." Plus, permanent upgrades. The cherry on top!
Boom. The game was ready and live on the Play Store.
Then came the curse of game development: marketing. It's not enough to make a game; you actually have to sell it. So I jumped into Google Ads.
The plan: A massive 15-day campaign. Daily budget: $2. Target Cost-Per-Install: $0.003.
I launched this gigantic budget. And I got an insane number of downloads! The AI went crazy over the results. It was like: "I cannot believe this! As an AI with billions of parameters, these results are mind-blowing!"
The result? Exactly 5,000 downloads. Yes, you heard that right—5,000!
...Except it was a totally useless crowd. Bangladesh, Afghanistan, Syria, Iran, Libya... None of them actually played the game. Their phones couldn't even run it! My crash rate went through the roof. So I lost my massive ad budget and ended up with a sky-high crash rate. Left totally empty-handed.
Right now, the game is making me a solid passive income: an average of $0.003 per day. I made $5 in 2 months! My total production cost was $20 for AI subscriptions and $50 for ads. At this rate, the game will break even in about 2 years.
So here is my request: Please watch the video, and if you like the genre, give my game a chance and leave a short review or comment! I have 5,000 downloads and only 4 reviews.
Love you all! If you read this far, stay safe and take care!
Free on Google Play: https://play.google.com/store/apps/details?id=com.garlictos.omer&referrer=utm_source%3Ddevto21
P.S. It has ads (full screen after wave 8/14, rewarded ones are optional). Graphics are AI assisted.
來源:DEV Community · dev.to


