IslandLab · Source archive

Island Lab Series: Day 1 of 5

Originally published as an Advisory Hour Substack post on 2026-07-28. This first-party site copy preserves the text, localizes public supporting media, and connects the record to its project trail.

Original post

Island Lab Series: Day 1 of 5 image from the original Advisory Hour post.
Island Lab Series: Day 1 of 5 image from the original Advisory Hour post.

This article begins a small series that will release this week around the Island experiments. After concluding my third experiment, it became clear that the next move was to bring them all together to make the compounding improvement line clearer. The image at the top of this article brings together what the evolution is. My goal in this series is to share some inside particulars you may not know about each one.

In this article I’ll cover what each of the experiments does at a high level, and then we’ll get into the details of each one over the week.

Island Lab Series: Day 1 of 5 image from the original Advisory Hour post.
Island Lab Series: Day 1 of 5 image from the original Advisory Hour post.

The procedural case study of this generative island was the first threejs project I ever did for my website. At the time it seemed like an insane idea: I’d have a procedurally generated island on my site. I have a lot of stories on this one, which I’ll share tomorrow. Let’s move onward to the second experiment, which was born shortly after Spring into AI. I fell into simulated worldspaces.

Island Lab Series: Day 1 of 5 image from the original Advisory Hour post.
Island Lab Series: Day 1 of 5 image from the original Advisory Hour post.

The second island was less about the graphics and more about the possibility of the island having durable characters that could learn, grow and adapt over time. This set aside graphical fidelity in favor of the simulated space. What kind of island would emerge given just a few initial constraints? Those were the type of questions I explored in the second island experiment-V2 Living World.

Then in V3, this came after Summer into AI. I learned that blender could accelerate my production time for building complex 3D scenes, that threeJS had a robust post-processing library, and that one could pack an insane amount of ideas into a browser download. Far more than I think anyone really gives it credit.

Island Lab Series: Day 1 of 5 image from the original Advisory Hour post.
Island Lab Series: Day 1 of 5 image from the original Advisory Hour post.

This V3 project explored ideas around character, animation, and just how far I could push modeling. I learned so much-and my expectations of what’s possible in a threeJS project were just blown away. I can’t wait to share all the subtle and hidden details in v3.

I’ll discuss more secrets and features of each of these projects in this Island Lab series. Next article will be on the V1 island. The first island is really a math question: why does a world appear when using pure math?

If you want, tour them over at the Island Lab gallery: Visit island lab

Thanks for reading! Subscribe for free to receive new posts and support my work.

Connected work