Russ Patterson

I’ve been making games for more than forty years—long enough to see the industry reinvent itself again and again.

I went from 2D sprites and assembly code, to 3D engines, to online multiplayer, to social and mobile games.
Each wave forced me to re-learn, re-tool, and rethink everything I knew about building worlds that move and breathe.

Now we’re at the next jump: from 3D to 128D.
It’s not about geometry this time—it’s about high-dimensional thinking.
Machine learning isn’t another platform shift; it’s a mental shift.
The people who thrive in this era will be the ones who don’t just use AI tools, but build and understand them.

Why this matters

Many developers today have spent their entire careers inside one generation of technology—mostly mobile, mostly incremental.
That’s comfortable. But the AI revolution is not incremental.
It will change what “game development” even means, just as 3D once did.
Using Copilot or ChatGPT is not the same as understanding how a neural network represents and transforms information.

Learning to build models changes how you think

When you train your own model—even a tiny one—you start to see problems differently.
You stop thinking in if-else trees and start thinking in distributions, embeddings, and probabilities.
You stop seeing systems as flat, and start imagining them as landscapes in 128 dimensions.

That shift in perspective is powerful. It’s how you move from scripting content to sculpting intelligence.

My plea

If you’re a developer who’s been through 3D, social, or mobile waves—you can learn this too.
Pick up PyTorch or TensorFlow. Train a simple network. (MNIST Digit Classifier is a great place to start). Don’t let this moment pass while others build the future.

This is the biggest transformation our craft has ever faced.
And the people who learn to think in high dimensions will define what gaming—and creativity itself—looks like next.

What’s next

This post is the beginning of a new series on high-dimensional thinking for game developers.

Over the coming weeks, I’ll be releasing a set of short, hands-on articles showing how to:

  • Build your first neural network (starting with the MNIST digit classifier)
  • Understand embeddings and vector spaces
  • Apply ML intuition to gameplay systems, difficulty tuning, and procedural design

If you’ve ever wondered how to bridge the gap between game development and machine learning, these posts are your roadmap.

Stay tuned — the next one, “Building the MNIST DigiClassifier,” drops soon.

We’ve gone from 3D worlds to 128D minds.
It’s time for every game developer to evolve again.