You click on a live feed from space. Satellites are orbiting Earth, beaming down high-resolution data. You’re playing with .tle and .parquet files, marveling at the sheer engineering required to parse orbital mechanics into a readable browser interface. It feels like the future. It feels generous.
But as you scroll through the comments, the geek joy is punctured by a quiet, desperate reality. A conservationist running a nonprofit focused on Latin American deforestation points out the fatal flaw: they can’t afford it. Their only “affordable” choices are two-year-old Google Earth imagery or 10m resolution shots that aren’t even detailed enough to be used as legal evidence.
In an era of black-box AI, the ultimate status signal isn’t automation. It’s something hand-built, legible, and finite.
When a commenter recently looked at Planet Labs’ open satellite feed, their highest compliment wasn’t about the resolution or the speed. They wrote: “Refreshingly doesn’t feel like AI. If someone asked me ‘what does software engineering look like’, I would point to this.”
They’re right. We have become so numb to LLMs hallucinating their way through JSON that seeing clean, deterministic code feels like discovering an endangered species in the wild. It is a masterful showcase of software engineering as a craft.
But let’s not romanticize this too much. This isn’t just a love letter to hand-built software. This is a story about the word “open.”
We love the word “open.” Open source. Open data. Open access. It sounds democratic. It sounds like a leveling of the playing field. But open doesn’t mean free. It means the infrastructure is exposed, but the keys to the kingdom still cost money.
Planet Labs exposes their high-resolution satellite data to developers, making the feed feel like a public utility. Yet, that same infrastructure remains commercially filtered. It can serve a tech giant monitoring supply chains, or a government surveillance company (as one user pointed out, many of their satellites service “Flock”—which sounds like an accident, but highlights the dual-use nature of the tech). It can serve whoever has the capital to buy in.
Every “open” system has a closure built in. The question is just who gets locked out.
If you store, analyze, or monetize data, you need to pay attention to this dynamic. Open feeds set expectations about access. They make us believe that the tools to monitor the planet are available to anyone with a laptop and an internet connection. But pricing and licensing ultimately determine who actually gets to use them.
The conservationist trying to stop illegal logging in the Amazon? They are priced out. The agrib conglomerate clear-cutting the forest? They can afford the premium tier.
The people protecting the rainforest can’t afford the data. The people burning it down probably can.
This is the tension of modern tech infrastructure. The craft is beautiful. The code is clean. The feed is a marvel of human ingenuity. But if the price of admission keeps the good guys out, “open” is just a PR strategy.
The next time you see an “open” API, don’t just marvel at the engineering. Ask who can actually afford to build with it. Ask who is being monitored, and who is doing the monitoring.
Hand-built software is the new luxury. But so is the truth about who gets to use it.
FAQ
Q: Isn't it fair for Planet Labs to charge for high-resolution data to cover their massive satellite operating costs?
A: Yes, rocket science is expensive. But if you market your feed as 'open' while pricing out nonprofits fighting deforestation, you aren't democratizing data—you're just running a B2B SaaS company with better PR.
Q: What's the practical implication for developers and data professionals?
A: Treat 'open' APIs as temporary sandboxes, not foundations. The future of monitoring won't be determined by who has the best algorithms, but by who can afford the licensing keys to the underlying data.
Q: Why is it a big deal that this software 'doesn't feel like AI'?
A: It's a sad commentary on our industry. We've accepted bloated, hallucinating black boxes as the default, to the point where clean, deterministic, legible code feels like a rare luxury rather than the standard.