In partnership with
This week on SHIFT
Everyone assumes the hard part of building a space company is the space part. Robbie Schingler spent 15 years finding out otherwise, as launching more satellites than any other entity on Earth turned out to be the tractable problem. A harder one was that the remote sensing industry hadn't yet found the internet, let alone the cloud, and a daily record of the whole planet has less meaning if nobody can query it.
Schingler is co-founder and chief strategy officer of Planet Labs, which he started after nine years at NASA, to build satellites small enough to hitch rides as secondary payloads. In this installment of our oral history project, he traces the path from dorm-room cubesats to a daily image of the entire planet, and he makes the case that AI's next leap depends on teaching models what the physical world is actually doing.
Advertisement
Your employees are connecting AI to everything. Now what?
ChatGPT and Claude don't just answer questions anymore. Employees are connecting them directly to Notion, Linear, Jira, and the rest of your stack. The AI can read, write, and take actions on company data. Most IT and security teams have no visibility into any of it.
Harmonic Security Connectors changes that. It sits inline with every AI-to-app connection, so you see each call, control what data moves, and block destructive actions before they happen. Employees notice nothing different.
See what's actually running across your business in a live demo.
Three things to listen for
1. The bottleneck was never orbit
Planet had to build cloud-native geospatial standards more or less from scratch, because remote sensing was still running specialized software on desktops. Schingler says querying vast archives of imagery at scale has only really been possible for two or three years.
2. The wind farm that was 120 missile silos
Using Planet's near-free research license, a think tank correlated imagery and identified roughly 120 missile silos in central China at a site marked on maps as a wind farm. Some 4,000-plus peer-reviewed papers have used the data in six years.
3. "Large Earth models," paired with large language models
Today's frontier models are human-centric - trained on what people wrote down. Schingler wants Earth observation data as a training ingredient so systems can triangulate what's physically happening, with satellites acting as networked robots running compute on orbit: sensors, nervous system, brain.
Listen now
SHIFT publishes every week. If someone forwarded this to you and you'd like to get it directly, subscribe at shiftshow.ai. And if you have a question, a guest suggestion, or a story we should cover, reply to this email. Jennifer reads them all.