Afleveringen
-
My first guest is Dr. Kelsey Jordahl, an old friend with a fascinating career trajectory: PhD from MIT in Oceanography, postdoc in France, teaching in New York, consulting on Wall Street, and now lead software engineer at Planet Labs—a company that images the entire Earth every single day.
Planet operates over 100 shoebox-sized satellites that photograph every point on Earth at 3-meter resolution daily. That's not science fiction—it's happening right now, and Kelsey helps make it work.
We start with his family story: Norwegian great-grandparents who came to Minnesota in the late 1800s hrough the Homestead Act .
Then we dive into Kelsey's career pivots: why he chose field work over theoretical physics, how he taught himself Python and created GeoPandas (a widely-used open-source tool for geographic data processing), and what it's like transitioning from academia to Wall Street to satellite imagery.
The conversation gets particularly interesting when we discuss the "too much data" problem. When you're photographing the entire Earth daily—generating hundreds of thousands of images—no human can possibly review them all. This is where AI becomes necessary, not optional. Kelsey explains how convolutional neural networks (CNNs) revolutionized image processing years before large language models captured public imagination.
We also explore:
How Planet launches 88 satellites at once on Indian rocketsWhy their first satellites used off-the-shelf telescopes from catalogsThe concept of "agile aerospace" (iterating satellite designs like software)What mosaics are and why they matter (stitching best images into seamless world maps)The US vs. Europe work culture gap (hint: Europeans don't voluntarily work weekends)How SpaceX changed the economics of space accessThe difference between seeing everything and knowing what mattersThis conversation will make you rethink what's possible with technology, but also what it means to have the capability to see everything happening on our planet simultaneously. Who decides what's important when you have infinite data? That's not a technical question—it's a philosophical one.
-
Zijn er afleveringen die ontbreken?
