Friday’s Food for Thought: All About AI

Exploring the Future of Artificial Intelligence 

By Andrew Foerch

Feb. 10, 2017

In 1997, IBM computer “Deep Blue” beat world champion chess player Garry Kasparov at his own game—the most significant advancement in artificial intelligence at the time. Twenty years later, the same technology is being used aboard the F-35 Joint Strike Fighter—the most expensive military weapons project to date.

Major ground was broken in the field of computer autonomy throughout the 21st century, and 2017 is poised to be another defining year in the progression of AI capabilities. Humanity is on the verge of a computational revolution, and mainstream media is taking notice.  

The Atlantic’s Defense One recently released its February e-book, “AI, Autonomy, and the Future Battlefield.” The e-book explores the implications of AI and machine learning with regard to defense and national security. The authors posit that autonomous weaponry, such as unmanned Aegis Warships, is increasingly able to mimic human behavior and replicate situational human decision-making, reducing the need for (and perhaps eventually replacing) human operators in battle.

Systems such as Turing Learning are being trained to distinguish between regular and suspicious behavior and to identify behavior the system considers potentially dangerous. Emerging UAV capabilities, namely autonomous loitering and target discrimination, could have significant implications for warfighting. These technologies enable systems to loiter over an area and compare objects on the ground with a database of existing images, marking targets when matches arise.

The New York Times recently published “The Great A.I. Awakening,” an in-depth article investigating the progress of Google Brain, an innovative project intended to create artificial neural networks comparable to the brains of living mammals. These neural networks facilitate Deep Learning, which will transform the way computers interpret and identify patterns based on human-labeled big data inputs.

Artificial intelligence is no longer limited to repetitive, structured tasks with fixed rules and clear definitions such as math or chess. The days of robotic soldiers and even more autonomous planes are quickly approaching.

Read the recent trajectory cover story “Machine Learning, Big Understanding” to learn more about artificial intelligence as it relates to GEOINT.

 

Photo courtesy of DARPA