Autonomous Horizons: The Way Forward

As the Air Force realigns itself from a counter-terrorism focus to dealing with near-peer rivalries with potential existential consequences, a “business as usual” approach to systems development will no longer suffice: we will not be able to continue with incremental advances on concepts developed decades ago. Instead, we need new technologies, affording us new capabilities, and new operational concepts to employ them.

Fortunately, there now exists a broad and deep technology push in the information sciences, particularly in the area of autonomous system (AS) development and its associated foundational technology, artificial intelligence (AI). Our knowledge of cognition and neurophysiology - the basis of what makes us “smart,” most of the time - grows at a dizzying pace, while our ability to build autonomous systems - like self-driving cars and game-playing robots - continues to make frontpage news, as new AI algorithms and learning techniques are developed and employed in novel ways.

And these advances are compounded by explosive gains in the underlying computational infrastructure afforded by Moore’s law growth in computational power, memory, networking, and data availability.