Hybrid Intelligence. Main Concepts and Application Scenarios

Abstract

AbstractThe purpose—Hybrid Intelligence is a cooperative collaboration between Human and Artificial Intelligence in solving intelligence tasks. The concept of Hybrid Intelligence originates from the ideas of William Ross Ashby, Joseph Carl Robnett Licklider, and Douglas Carl Engelbart. Fuzzy set theory, introduced by Lofti Zadeh, is a natural tool for describing and modelling Hybrid Intelligence. Fundamental problems of Hybrid Intelligence—modelling human perception and operating with perception-based information have been discussed in the report. Two application scenarios—personalization of human interaction with the digital world and evaluation and monitoring of complex processes have been discussed and illustrated, too. Design/methodology/approach—The results are based on system analysis, fuzzy logic, some mathematical and psychology theories, and facts. Findings—Hybrid Intelligence is a pragmatic aspect of intelligence technologies, and this concept could reply to the crisis of modern Artificial Intelligence. Originality/value—Hybrid Intelligence is a new formal model which could be a theoretical base for the implementation of the concepts like “automation of knowledge work” (McKinney), “augmenting human performance” (NSF), “human–machine symbiosis” (DARPA), etc., which have broad areas of applications. Research/Practical/Social/Environment implications—Hybrid Intelligence can help people comfortable and effectively interact with the digital world. It can be a base for developing systems for evaluating and monitoring complex societal and environmental processes for their better understanding, management, and optimization. Research limitations—The Hybrid Intelligence approach can augment a regular human performance in solving routine intelligence tasks but cannot substitute a genius.KeywordsHybrid intelligencePerception modellingPersonalizationEvaluation and monitoring complex processes