A spy-console look, built from public signals
Bilawal Sidhu, a former Google Maps product lead, has created God's Eye View: The Open-Source Spy Satellite Simulator, a chillingly effective application that transforms scattered public data streams into a single, dramatic global surveillance interface. This open-source project, which rapidly accumulated over 55,000 GitHub stars, makes visible the constant, pervasive broadcasting of our modern world.
God's Eye View pulls diverse public feeds onto a unified 3D globe, offering a simulated spy-console experience. Sources include:
- Aircraft position broadcasts from networks like The OpenSky Network
- AIS vessel data
- CelesTrak satellite orbit elements
- USGS earthquake reports
- Thousands of public traffic cameras
The critical distinction here is the juxtaposition of presentation and provenance. What feels like a classified surveillance terminal—complete with night vision and thermal overlays—is, in fact, built entirely from publicly accessible data and open-source code. Sidhu’s genius lies in aggregating these disparate, unclassified signals, demonstrating how readily available information can create an unsettlingly comprehensive picture.
The clever trick is making stale data feel live
Forget defense-grade backends or classified infrastructure. God's Eye View achieves its uncanny realism with a surprisingly lightweight stack: vanilla JavaScript, CesiumJS for the globe, Vite, and a minimal Node server that simply proxies public feeds. The magic isn’t in brute force, but in elegant code.
The cleverest trick is how it handles stale data. Public feeds, like ADS-B for aircraft or AIS for ships, update every 15-30 seconds. Instead of objects teleporting across the screen, the globe deliberately renders slightly behind real-time, interpolating smoothly between the last known position and the current one. If an update drops, it employs dead reckoning, guessing forward based on speed and heading—a trick sailors used before GPS.
Immersion runs deep, thanks to sophisticated controls and visual effects. Users can enter a follow-camera mode, riding along with a specific aircraft or even the International Space Station. Number keys toggle between striking visual filters:
- Night vision (green phosphor)
- Thermal-style (luminance-based gradient, not actual sensor telemetry)
- Retro CRT distortion
These are not sensor telemetry; they are WebGL GLSL shaders, transforming ordinary map data into a convincing spy-console aesthetic. It’s a masterclass in making public information feel like clandestine surveillance.
The map looks omniscient. The data isn’t.
That dramatic interface, however, is a masterclass in presentation over raw telemetry. God's Eye View: The Open-Source Spy Satellite Simulator doesn't pipe live satellite imagery; the ground beneath is standard map data. Roadway traffic corridors are simulated, and embedded CCTV camera placements are often coerced estimates, not precise coordinates. This isn't a hardened defense system—it’s a carefully constructed illusion.
Practical gaps abound. Feeds are inherently delayed or incomplete; ship tracking goes quiet in the open ocean, reliant on terrestrial receivers. Aircraft only appear when actively transmitting via ADS-B and within receiver range. The project itself acknowledges these limitations: data may be "delayed, incomplete, modeled, inferred, or just wrong."
Crucially, MIT-licensed code does not grant carte blanche for every underlying dataset. While the application is open-source, OpenSky's flight data and undersea cable maps are often non-commercial. Building a startup on this requires a deep dive into the data-sources.md file, available on the project's GitHub - bilawalsidhu/gods-eye-view page, to understand licensing restrictions and operational limits. This tool is a compelling demonstration, not a production-grade service.
Enjoying this? Get one like it in your inbox each morning.
one email a day · unsubscribe in two clicks · no third-party tracking
One install turns public infrastructure into a question
Sidhu’s project isn't just a dramatic visualization; it's a direct challenge, easily run on your local machine. Clone the repository, run npm ci, then npm run doctor to verify your setup, and finally npm run dev. Just be sure to check the README for the current Node.js requirement—it favors Node 24.14 over the unsupported Node 25.
Version 0.2 introduced the MCP server, a significant shift from passive observation to active control. This server exposes the globe’s built-in voice agent tools to compatible AI clients like Claude, Codex, and ChatGPT. While agent capabilities are grounded in documented tools, the potential for AI-driven interaction with this aggregated data raises compelling questions.
One install turns public infrastructure into a question. The project explicitly states it avoids tracking individuals, yet seeing thousands of public traffic cameras alongside license plate reader locations, all legally sourced, is undeniably chilling. This aggregation of open-source intelligence, even without individual targeting, blurs the line between transparency and pervasive surveillance. We must ask where public data ends and a private eye begins.
Frequently Asked Questions
What is God's Eye View?
It is an open-source web app that combines public geospatial feeds on an interactive 3D globe.
Does God's Eye View use classified satellite imagery?
No. Its spy-console look is a presentation layer over public data and ordinary map imagery, not classified satellite imagery.
Can I run God's Eye View locally?
Yes. The project can be launched locally with Node.js and its documented npm commands; check the repository for supported runtime requirements.
Are the aircraft and ship positions perfectly live?
No. Public feeds can be delayed, incomplete, or unavailable, and the app smooths or estimates movement between updates.
Can an AI agent control the globe?
The project's MCP integration exposes globe controls to compatible AI clients, allowing them to direct actions such as moving the camera.