Good morning,
Last week, we explored why the Department of Defense has been slow to adopt AI at scale—largely due to a lack of understanding about what AI can do, how to start, and when it's appropriate. We introduced the “7 Patterns of AI,” a powerful framework that helps clarify if, where, and how AI makes sense for military applications.
But not every military AI initiative has failed.
One standout is Project Maven. Born in 2017, Maven was a pioneering effort to apply computer vision to the flood of full-motion video from drones and other ISR platforms. It succeeded where so many other Pentagon AI projects failed. The question is: why?
1. It Aligned With the Right AI Patterns
Project Maven was laser-focused on two core AI patterns: recognition and patterns & anomalies. It wasn't trying to build a sentient battlefield brain or an all-seeing intelligence tool. It aimed to do one thing well: automatically detect, label, and classify objects (vehicles, people, weapons) in hours of video foo…
