Your Ideas Are Not Cheap Anymore
But if you don't have any, there is another way
A small part of my last post discussed Prince of Persia. A legendary game from 89 written entirely by one person: Jordan Mechner. The following tweet was then reshared by Jordan himself becoming likely my most popular online post so far:
T
If AI make us so much more productive, how much easier would be to write this game today. It took Jordan 3-4 years to build. You could surely do it faster today. How much faster? I bet it’s a year.
Is Work Easy Now?
When we remove the technical difficulty, we start to see that what's left is still a lot of work. What is the game all about? Who is it for? How is it different from everything else on the market? What about character and level design… as Ben Affleck said, AI writing is pretty shitty.
So what the conclusion?
It was never easier to validate if your idea is worth anything, but the work is still there.
Are Ideas Still Cheap?
The reason why in the previous era writing code looked like the most important part of any software was because it was scarce and hard to get. Producing quality lines of code required years of learning, and even then took a lot of time.
Making it easy removes this value from the equation. What is value now? Whatever still takes time. Just writing a Prince of Persia knockoff would be perfectly enough in 1991 to have a successful game. In 2026, it’s not. You can generate such a knockoff with a prompt, and that’s why it’s not even interesting enough to try.
It’s simply not interesting.
What we want is scarcity. We want things that are hard to get. Nobody appreciates oxygen which is essential to survive the next five minutes. We appreciate diamonds and Picasso paintings despite their value is purely imaginary.
When I started in the startup industry, smart people were saying that ideas are cheap, and it’s all about execution. I believed that.
Is it still true? What if execution became the easy part. In many cases it did. Now, all that left is the idea. Suddenly, that’s the most expensive part.
Can you surprise me? Can you make a game I’ve never seen before?
What was unique about Prince of Persia was the smooth character movements that used rotoscoping for the first time in a computer game. The result was shocking. Despite having barely a few pixels the character moved like a real person.
But this alone would surprise no one today.
What's the rotoscoping of 2026? What can you build that AI can't generate from a prompt?
Inventors vs. Investors
If you’re an idea person, congratulations. Making them real was never easier. If you have ideas: build. Just be ready for competition bigger every year.
But if that doesn’t excite you—and it surely doesn’t have to—then invest.
The gap between those who own the means of production and those who don't will widen faster than ever before, but the window is still open. You can still acquire assets that position you on the right side of this shift—assets that may become far less accessible as automation concentrates wealth and reshapes markets.
Bitcoin - The native money of the information age. The only asset that’s provably scarce, perfectly liquid, and controlled by no one.
Technology stocks - One day robots will build robots. But today, you can still become a shareholder in the companies that will sell them to the world.
Your network - Skills depreciate. Credentials expire. Relationships compound. In a world where human connection becomes the luxury good, the people you invest time in today become your most valuable asset tomorrow.
Jordan Mechner spent four years alone in a room bringing Prince of Persia to life. He’d likely still do great today, even if technical execution takes a prompt. However, if you’re not building, make sure you own a piece of economy that just started to explode. The future won’t leave you hungry. But it might leave you without options. I prefer to stay sovereign.
✦
Speaking of building and investing: we’re still slowly testing and improving the freshly opensourced version of Deltabadger. Many users already run it on their server, but I spent last two days trying to debug the Umbrel build. Wish me luck. On paper Umbrel seems like the perfect home for Deltabadger. If you want to help: install and test. Jump into our Telegram channel if you have any questions or ideas.