Research on artificial intelligence technology promoting the transformation of American intelligence work: the characteristics of intelligent intelligence transformation (1)

Dock intelligence mission requirements through artificial intelligence projects.

The core task of the intelligent transformation of the U.S. intelligence community is to introduce mature civilian artificial intelligence technology to serve the needs of real-world intelligence missions. In order to meet the needs of intelligence missions and realize the seamless connection between artificial intelligence technology and intelligence work, the U.S. Intelligence Advanced Research Projects Agency (IARPA) has set up four corresponding offices: Intelligent Collection Office, In-Depth Analysis Office, Security Operations Office and Accident Prediction Office . These four subordinate offices correspond to intelligence collection, data processing, intelligence analysis, and computing capabilities enhanced by artificial intelligence technology. The core resources for the development of artificial intelligence technology are data, algorithms and computing power.

The artificial intelligence projects carried out by the Intelligent Collection Office are related to intelligent intelligence collection, involving geospatial intelligence (Amon-Hen project), identification information (Odin project, BEST project).

The programs conducted by the Office of Deep Analysis are related to intelligence analysis, but more on the analysis of data and information than is conventionally thought of as intelligence analysis. Its projects involve language data processing (BETTER project, MATERIAL project, BABEL project, KDD project, SCIL project), video data processing (ALADDIN project), image data processing (Janus project), etc.

Most of the projects carried out by the Security Operations Office are not directly related to artificial intelligence technology, but focus on improving computing power, which can improve the computing power of artificial intelligence technology from the side. Its projects involve high-performance computing, including C3 project, HECTOR project, LogiQ project, SuperTools project, etc.

The projects carried out by the Accident Prediction Office are related to machine reasoning and automated analysis, involving automatic detection of network security (CAUSE project), machine-assisted analysis (ACE project, ForeST project, OSI project, etc.), automatic political event prediction (Mercury project), etc.