The Overhang in Your Pocket: Orbital Mechanics Edition

I've been writing about technological overhang — the gap between what's technically possible and what we're actually doing with it. The classic examples are civilisational: desalination technology exists but deployment lags, landfill mining is feasible but economics are broken.

But there's a personal version of this too. You have powerful tools in your pocket that you're not using — not because they don't work, but because you assume the barrier to entry is higher than it actually is.

Most people treat AI assistants as search engines or writing aids. But they're also capable of building things: simulations, visualisations, tools for understanding. The cost of creating bespoke interactive explanations has collapsed. Most people haven't noticed.

To demonstrate this, I tried a simple experiment. I asked Gemini:

"Please could you make me timeline slider graphic showing Hohmann transfer orbit and explaining when are good times to eg leave earth for mars"

I didn't give it any code. I didn't specify a framework. I just asked for the thing.

It generated a React component using HTML5 Canvas for the orbital mechanics. And because I've built this site's infrastructure to be modular, I can embed that component right here, in this post.

The Result

How It Works

This isn't a static video. It's a live simulation running in your browser.

• Orbital Mechanics: It calculates the relative positions of Earth and Mars based on their different orbital periods (365 days vs 687 days).
• The Launch Window: A Hohmann transfer — the most fuel-efficient way to get to Mars — requires specific alignment. You can't just aim where Mars is; you have to aim where it will be.
• The Visualizer: The slider lets you scrub through time to find that window (when the line turns green).

Why This Matters

The point isn't that I have a Mars simulator on my blog. The point is that I didn't have to spend a week building it.

I asked for it, verified it, and deployed it. Total time: about an hour, most of which was me understanding what I was looking at.

This is what the overhang looks like at personal scale. The technology to build custom learning tools, visualisations, and simulations is subscription-priced and accessible. But most people don't use it this way because:

• They think "building software" requires being a developer
• They assume AI can only generate text, not functional code
• They've never tried asking for something interactive

The barrier isn't technical. It's imaginative. If you can describe what you want to understand, you can probably build a tool to simulate it.

I'm not special here. I just tried asking.