The Real Cost So Far
When I first started building Species Zero, I had a slightly magical idea in my head.
Not that it would be free, exactly. I’m old enough to know that “free” usually just means “you haven’t found the bill yet.”
But I did think the cost would be simple. A bit of AI usage. Maybe some hosting. Maybe a few paid tools later if the thing became real.
The actual cost has been stranger than that.
Species Zero started as Genesis Beasts on June 13, 2026. The idea is an infinite creature evolution simulator for phones. You raise odd little AI-generated creatures, feed them, watch them change, and hopefully feel attached enough to come back.
That sounds neat in one sentence. Building it has not been one sentence.
The first cost is AI credits.
Every time I ask the builder to add something, fix something, explain something, or try again, I’m spending credits. Sometimes that feels perfectly fair. I ask for a feature, it builds the feature, and I feel like I just hired a tiny tireless helper.
Other times, I spend credits just trying to understand why something broke.
For example, account creation looked like it was failing. That is a very alarming thing when you are making any kind of app with accounts. Later it turned out the account had actually succeeded, but the app made it seem broken. So the cost wasn’t just the fix. It was the confusion. The checking. The second-guessing.
Then there was the missing @radix-ui/react-tabs issue that caused 504 errors. I now know that “504” basically means the thing asked for something and waited too long, like standing at a service counter where nobody comes back. At the time, it just felt like the app had gone into a fog bank.
That kind of problem burns credits, but it also burns confidence.
The second cost is time.
Not coding time, exactly. I’m not sitting here lovingly typing out mysterious symbols like a movie hacker.
My time goes into deciding what to ask for, reading what changed, testing on the preview, noticing what still feels bad, and trying not to ask the AI to fix five things at once.
That last part is harder than it sounds.
At one point, after feeding a creature, I realized the game didn’t make the next step obvious. The feeling was basically: “I don’t understand what is next after feeding.” That led to adding a NextStepHint component, which is just a fancy way of saying the game needed to gently point the player forward.
That was a useful change. But it also taught me something uncomfortable: a working app can still be confusing.
The third cost is attention.
This may be the biggest one.
Species Zero now has pieces like Stripe payments for pay-to-accelerate, Discoverer Ranks for prestige, creature thoughts, overnight dreams, evolution drift, random events, social ideas, and art improvements.
Each one sounds exciting. Each one also opens another drawer full of decisions.
Should speeding things up feel helpful or unfair? Should a creature’s first-person thoughts be funny, sad, strange, or all three? How much randomness makes the game feel alive, and how much makes it feel like nonsense?
At one point I looked at the whole thing and my honest reaction was: “This game is not good.”
That was not a dramatic artistic moment. It was just me being tired and disappointed.
But it led to better questions. The problem wasn’t just bugs. The game needed more soul. That’s where ideas like Creature Voice and Overnight Dreams came from. Not from confidence. From irritation.
There are also costs that don’t fit neatly anywhere. Preview loading problems. Vite re-optimization, which I understand as the tool re-sorting its internal boxes before it can show me the app. A drift computation hang, meaning one of the creature-change calculations seemed to get stuck thinking too hard.
These moments don’t look like progress from the outside. But they are part of the bill.
So what does building this actually cost?
Some money, yes. Some AI credits, definitely. But mostly it costs sustained attention from a person who is learning in public and regularly feels out of his depth.
I still think that cost is worth it. But I’m trying to stop pretending it is small.
Small takeaway: the cheapest part may be the tool; the expensive part is caring enough to keep going.
Keep reading
What Species Zero Taught Me
I started making a creature game with AI and quickly learned the real work was not the idea. It was surviving the messy middle.
The Button After Feeding
I thought the hard part was making creatures evolve. Then I realized players were getting lost after the very first meal.
When the Creature Talked Back
I thought I was building a little creature game. Then one small change made it feel less like software and more like a world.
When The Game Felt Empty
I had built plenty of parts, but the game still didn’t feel good. That uncomfortable moment changed what I worked on next.