MACHINE INTELLIGENCE
Abstract
(U) The Machine Intelligence project is budgeted in the Applied Research Budget Activity because it is developing technologies that will enable computing systems to extract and encode information from dynamic and stored data, observations, and experience, and to derive new knowledge, answer questions, reach conclusions, and propose explanations. Enabling computing systems with machine intelligence in this manner is now of critical importance because sensor, information, and communication systems continuously generate and deliver data at rates beyond which humans can assimilate, understand, and act. Since its creation over 50 years ago, artificial intelligence (AI) has gone through several phases. Initially, AI emphasized rule-based and symbolic approaches. These were eventually reconceived using a human-intelligence paradigm (“cognitive computing”). Recently, a more powerful approach has emerged, with rule-based, symbolic and human-oriented approaches combined with large-scale statistical approaches that make explicit use of massive distributed data and information bases. These data/information bases are curated (e.g., machine-filtered or human-selected) and raw (e.g., as originally obtained and perhaps of unknown provenance); structured (e.g., tabular or relational) and unstructured (e.g., text documents, multi-media files); static (e.g., historical, unchanging) and dynamic (e.g., real-time sensor data). This explosion in available data/information, combined with the ready availability of inexpensive mass storage and ubiquitous, inexpensive, computation-on-demand, provide the foundation for entirely new machine intelligence capabilities. The technologies developed in the Machine Intelligence project will result in revolutionary capabilities in military command and control, intelligence, decision-making, and situational awareness/indications and warning for a complex, global environment where traditional (e.g., nation-states) and non-traditional (e.g., trans-national) actors and new classes of cyber-physical-human threats have become the status quo.
Document Details
- Document Type
- R2 Budgetary Justification
- Publication Date
- Oct 01, 2011
- Source ID
- 0602305E_2_0400_PB_2011
- Change Summary Explanation
- FY 2011 Not Applicable
- Service Agency Name
- Defense Advanced Research Projects Agency
Entities
Organizations
- Defense Advanced Research Projects Agency
Related Documents
- Child Project: MACHINE INTELLIGENCE
- Child Accomplishment: Machine Reading and Reasoning Technology
- Child Accomplishment: Web-Scale Information Integration
- Child Accomplishment: Large-Scale Asymmetric Systems