Automating the Wisdom of the Crowd.
Abstract
The current Army Operating Concept, Win in a Complex World, describes future operational environments as complex, uncertain, chaotic, and dangerous. Under these conditions, accurate decision-making becomes a cyclical time-based competition. Commanders who can achieve situational understanding quickly enough to anticipate opportunities and threats gain and maintain a position of relative advantage over the enemy. One way in which intelligence organizations help commanders achieve situational understanding is through the rapid collection, analysis, and delivery of actionable intelligence to the point of need. Analysts often balance two contradictory requirements to avoid intelligence failures: speed and accuracy. Failure to produce actionable intelligence in time may negate its value. Conversely, failure to vet the accuracy of the intelligence may negate its value irrespective of the timeliness. Three factors limit both speed and accuracy: the capacity of human cognition, susceptibility to bias, and the limits of organizational knowledge. Mitigating the effects of these three constraints improves commanders' ability to achieve situational awareness. This paper examines whether a combination of artificial intelligence (AI) and crowdsourcing can help the intelligence community achieve situational understanding quicker and with more accuracy. After a close examination of the benefits and pitfalls of artificial intelligence and crowdsourcing, this paper concludes with a "Hybrid Options" section, which demonstrates that a combination of artificial intelligence and crowdsourcing provides a path to mitigating the effects of the limits of human cognition, susceptibility to bias, and the limits of organizational knowledge. If taken into practice, this hybrid will prove an invaluable resource to operational planners.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 24, 2018
- Accession Number
- AD1087259
Entities
People
- Jakin J. Waldock