Running local models on my machine to optimize the event prediction algorithms for BoutPredict and my friend sends me a link that makes our local server setup look like a high school science fair project :P
A US-based startup named Reflect Orbital wants to sell sunlight.
Literally.
CEO Ben Nowack is proposing a plan to launch 50,000 massive mirrors into space by 2035.
The goal is to reflect the sun’s rays back to specific coordinates on Earth, effectively turning night into day for anyone willing to pay $5,000 an hour.
As software engineers, we are used to treating digital infrastructure like an on-demand service.
We spin up AWS instances or Cloudflare workers with a single terminal command.
Reflect Orbital is attempting to do the exact same thing with the solar system.
The Infrastructure: 50,000 Orbital Reflectors
The engineering behind this is both brilliant and absurd.
To make this work, Reflect Orbital plans to deploy a constellation of satellites in low Earth orbit -.-
Each satellite will be equipped with a mylar mirror measuring up to 180 feet wide. Okay.
And when a customer requests light, the satellite tilts its mirror to catch the sun (which is still visible in orbit even when it is night on the ground) and beams a three-mile-wide spotlight down to the target location.
They are essentially building a REST API for the sun.
You drop a pin on a map, pay the invoice, and an orbital mirror executes the request.
The $5,000 Unit Economics
So, who pays $5,000 an hour for a spotlight?
The primary target market is not eccentric billionaires hosting midnight parties.
The target is massive solar farms.
When the sun goes down, solar energy production flatlines, but human energy consumption spikes.
This is the classic duck curve problem in grid management.
If a massive solar farm can pay Reflect Orbital $5,000 to generate $15,000 worth of extra electricity during peak evening pricing hours, the ROI is mathematically undeniable.
It creates a self-sustaining loop of synthetic energy generation.
It also opens up secondary markets, btw:
extending the working hours of agricultural farms, lighting up disaster relief zones, or providing constant daylight for massive outdoor movie sets.
The Biological DDoS Attack
This is where the startup ambition collides with reality.
You cannot just deploy a massive feature to the global environment without breaking critical downstream dependencies.
Astronomers and biologists are raising massive, justified red flags.
Pumping artificial daylight into the night sky is essentially a DDoS attack on the planet’s biological clock.
- The Circadian Disruption: Millions of years of evolution are tied to the dark-light cycle. Sudden blasts of intense sunlight at 2 AM will severely disrupt human sleep cycles and local ecosystems.
- Navigation Failures: Nocturnal animals and migrating birds rely on the stars and natural darkness to navigate. A three-mile-wide artificial sun will cause massive ecological confusion.
- Blinding the Observatories: Ground-based telescopes are already struggling to see past the thousands of Starlink satellites. Adding 50,000 massive, highly reflective mirrors to low Earth orbit would effectively blind our ability to study deep space.
So, Stick to Local Code? ;-;
We love to build highly scalable tools, right?
When you are deep in the hardware fortress, the idea of solving global energy limits with space mirrors sounds like an incredible engineering triumph.
But Reflect Orbital is a stark reminder that physical infrastructure has massive, permanent consequences that a simple code revert cannot fix.
I will stick to predicting fights with my machine learning model on Boutpredict for now.
At least when my database routing breaks, it does not ruin the night sky for the rest of the hemisphere ;)