# Spy Agencies Had This Tech. Now You Can Build It. > Source: > Published: 2026-10-30 13:31:00+00:00 I initially built God’s Eye View over a weekend from my couch. One person, a stack of APIs, and my army of AI agents doing the work that used to take an entire team months. It started trending as “[Palantir in a weekend](https://x.com/i/trending/2024841524710707470?lang=en).” Then Palantir’s co-founder, Joe Lonsdale, [responded on TBPN](https://x.com/tbpn/status/2025019956967178620). [Watch the full video](https://www.youtube.com/watch?v=PBThLgs2AZc) below -- I’ll show you what public data can reveal, what y’all have built, and where we’re taking God’s Eye View next. Jump to [Palantir’s response (1:16)](https://www.youtube.com/watch?v=PBThLgs2AZc&t=76), [what public data can reveal (4:43)](https://www.youtube.com/watch?v=PBThLgs2AZc&t=283), or [what you’ve been building (7:34)](https://www.youtube.com/watch?v=PBThLgs2AZc&t=454). Since I first showed you guys God’s Eye View and how I made it, so many of you have been sending me your own builds. Personal geospatial dashboards, threat tracking rigs, mission briefing boards. The same recipe in your own hands, building your own picture of the world. ## Looking back up at power Palantir’s co-founder was basically making the case that the only thing AI is really putting at risk is low-end SaaS right now. Companies like Palantir, on the other hand, are safe. In parallel, people from major defense contractors and defense tech startups started reaching out. Turns out they’ve got immaculate sensor systems, but are aching for a better way to fuse it all together. At the same time that the defense world started circling, journalists started reaching out too. The fourth estate, the press, the public-interest watchdog, saying this is the kind of tool they’ve always needed. The dichotomy was clear to me. This was simultaneously surveillance and sousveillance. Surveillance is what powerful institutions do when they look down at people. Sousveillance is what people do when they look back up at power. The version of this where reporters, researchers, and individual citizens can put together their own picture of what’s happening -- that is a different kind of world. ## The canvas already exists When I was at Google working on the Maps team, creating something like God’s Eye View was a tall order. A single feature update could take teams of engineers months to pull off. Building an entire fused-sensor platform in a couple of days by yourself was not on the table. But that’s exactly what we were working towards back then. I was the product lead for Google’s 3D model of the world. Shipping things like Google Maps Immersive View, the ARCore Geospatial API, and even the 3D Tiles API much of this work is built on. I worked directly with Cesium to get Google’s amazing 3D data out into the wild in an [open format](https://cesium.com/why-cesium/3d-tiles/). When I was on the platform side of this at Google, the hope was we’d let a thousand flowers bloom. Now I get to not only build my own flower, but inspire a garden of interesting flowers. The hardest part of all of this used to be mapping the world itself. That takes infrastructure. You need satellites, fleets of airplanes and cars. But Google has done so much of that mapping for us. The canvas already exists. Now it’s a matter of what you choose to paint on top of it. ## What disappears tells you something, too We’ve got a plurality of sensor systems watching the planet right now. Satellite imagery, radar from orbit, flights, ships, public cameras, social media feeds. Each captures a sliver of what’s happening. There is now a treasure trove of publicly available data just sitting out there. None of it is individually novel. What’s novel is fusing it all together into a single picture. And that’s the fusion that required a dev team to pull off. Now you can unleash your swarm of AI agents at the data streams of your choice, and you can build the system you’ve been imagining. Don’t get me wrong, domain expertise still matters. You have to know what to ask for and guide the AI around pitfalls. But the execution layer is right at your fingertips. You can look at which satellites were overhead around an event. Watch commercial flights reroute around airspace closures. Look into gaps in a tanker’s AIS tracking. Pull public CCTV feeds. And heck, you can even align 3D Gaussian splats to their physical location for a photorealistic ground-level view. If you’re not in the OSINT world, intelligence is also about seeing what’s missing. When flights go dark or GPS goes dark, that hole in the map is intelligence. What disappears tells you as much as what appears. I was able to map potential GPS interference from the GPS degradation of commercial airliners. You can literally turn the global fleet of airplanes into a distributed sensor network. And look, in many cases, the data is still the moat. The players with the crown-jewel datasets, the Palantirs, the primes, the three-letter agencies, they absolutely have an edge. That is real, and I’m not going to pretend otherwise. But what I’m showing here is what’s already publicly available. This is open and commercial data, and let me tell you, it is enough to build a remarkably complete picture of the world. It’s no surprise that ODNI and the CIA have their own [open-source intelligence strategy](https://www.cia.gov/stories/story/ic-osint-strategy-rollout/), issued in 2024. There’s a million other layers that you could be working with. You don’t even need my code. I can give you the recipe, and you can point your AI agents at the sources and tell them what to fuse. ## Let me show you what people built [Shadowbroker](https://github.com/BigBodyCobain/Shadowbroker) is this community-built globe that fuses a bunch of these public data feeds. The sheer number of feeds and modularity of this one is insane - dare I say overwhelming! Next is [ULTOR](https://github.com/Sequoiabm/ULTOR), by this dev named Sequoia, who said this would have taken him forever to do if not for the fact that I gave the formula away. So he could focus on the fun stuff. Sequoia also told me it was “[really good for my resume and helped me get an internship at amazon Leo!](https://www.youtube.com/watch?v=PBThLgs2AZc&t=493)” Here’s another talented dev who was inspired to build [WorldWideView](https://demo.worldwideview.dev/). There’s [Gemini / Claude Earth](https://github.com/rdumasia303/Gemini_Claude_Earth), where you can visualize nuclear sites, conflicts, flights, seismic activity, and news. And then there are more. [State of US & World Politics](https://apps-h3p.com/polymarket-us-politics/conflict-monitor.html) fuses prediction markets with conflict data and a continuous news pipeline. And ARGOS ATLAS is building on God’s Eye View for its own 3D product. They [credit the code directly](https://argosatlas.com/en/atribuciones/). That’s why I gave the recipe away. There’s even a third-party Android version. Google Play lists 10K+ downloads for FEAXR AI’s [God’s Eye View: 3D Earth Globe](https://play.google.com/store/apps/details?id=com.feaxr.godseyeview&hl=en_US). And gosh, there’s so many other projects that I couldn’t include in this video. Just look at the comments. People sharing mobile apps, custom data layers, even a globe for finance. Point your agents at my GitHub, fork it, remix it, share it. The tools have never been more accessible. ## Seeing the world live is just the start Since I first recorded this, we’ve [open sourced God’s Eye View](https://github.com/bilawalsidhu/gods-eye-view). You can download it, run it, and build on it. And you’ve already taken it way further than my wildest imagination. The next step is being able to rewind an event, compare observations, and trace back any claim to its source. That’s what we’re building towards. A record of what happened with a clear line between what we observed and what we inferred. [Watch the full tour on YouTube.](https://www.youtube.com/watch?v=PBThLgs2AZc) The [code and setup guide are here](https://github.com/bilawalsidhu/gods-eye-view). [Join the waitlist for the official release of God’s Eye View](https://halfpixel.ai/) - I think you’re gonna love what we’re cooking. See y’all in the next one. Cheers. -Bilawal