Did someone order AI to count trees on the Sahara Desert?

 Yes, an international team of scientists used AI to count trees in the Sahara Desert. [1]

Led by researchers from the University of Copenhagen and NASA, the team used a combination of artificial intelligence (deep learning) and high-resolution satellite imagery to map individual trees across a massive 1.3 million square kilometer expanse of the Western Sahara Desert, the Sahel, and sub-humid zones of West Africa. [1, 2]

The results, published in the journal Nature, completely surprised the scientific community: [1]

• The Final Count: The AI successfully identified and mapped over 1.8 billion trees and shrubs in an area previously thought to have virtually no tree cover. [1, 2]
• The Method: Before unleashing the AI, scientists spent a year manually marking 90,000 individual trees to teach the computer what tree shapes and shadows look like. [1, 2]
• The Computing Power: Once trained, the deep learning algorithm was paired with the Blue Waters supercomputer at the University of Illinois. It managed to count the remaining hundreds of millions of trees in just a few hours—a task that would have taken thousands of humans several years to complete. [1, 2, 3, 4]

This breakthrough allows environmentalists and policymakers to better track deforestation, support local conservation efforts, and accurately measure how much carbon dryland vegetation actually stores. You can read more details about the project through the NASA Earth Observatory or the University of Copenhagen's report. [1, 2, 3]

I heard about an AI search counting Ukrainian graves marked with flags in Ukraine.

You are likely referring to the open-source intelligence (OSINT) projects and journalistic investigations that use satellite imagery, drone footage, and machine learning to track expanding cemeteries in Ukraine. [1, 2]

While it is a state secret exactly how many soldiers have died on both sides, researchers and organizations use these high-tech methods to bypass official censorship. [1]

The Role of Flags and AI in Counting

• Flags as Visual Markers: In Ukraine, it is a widespread custom to place a vibrant blue-and-yellow national flag (and sometimes a black-and-red war flag) over the grave of a fallen soldier. As military cemeteries like the Alley of Glory in Kyiv or fields in Dnipro expanded, the sheer volume of flags became massive. [1, 2, 3, 4]
• Machine Learning & Satellite Data: Investigative organizations (such as the Center for Information Resilience or media outlets like Mediazona and the BBC) use high-resolution commercial satellite imagery from companies like Maxar Technologies and Planet Labs. [1, 2, 3]
• How It Works: Computer vision algorithms (a type of AI) are trained to analyze time-lapse satellite photos. The AI detects changes in soil patterns (newly dug earth), grid alignments, and distinct colorful markers—such as groups of flags or wreaths—to automatically map and calculate the physical expansion of burial sites over time. [1, 2, 3]