Augmented intelligence is an alternative conceptualization of artificial intelligence that focuses on AI's assistive role, emphasizing the fact that it is designed to supplement human intelligence rather than replace it. Many industry experts believe that augmented intelligence or an alternative term should replace artificial intelligence to ease the public's fear of displacement by AI systems in a not-too-distant-future robot economy. The central concept behind augmented intelligence is the idea that the technology exists to serve the public good.
AI involves the simulation of human intelligence processes by software systems. These processes include acquiring information and rules for using the information, applying the rules to reach conclusions and applying self-corrective measures to adapt to variations in outcomes. An expert system, for example, incorporates a knowledge base for a given area, such as medicine or the law, and rules for applying the information to guide its responses to user input. Through machine learning, an expert system can improve its performance by analyzing its experience, similarly to the way that humans learn.
ROSS Intelligence, a legal expert system (see: AI attorney) uses data mining, pattern recognition and natural language processing and a specialized knowledge base to perform the tasks of a research assistant. Unlike a human assistant, however, the AI system can mine data from about a billion text documents, analyze that information and provide precise responses to complicated questions, in natural language, in less than three seconds. Furthermore, a software system can work unlimited hours and doesn't require a salary or a vacation. Because of the efficiencies of AI, such systems are increasingly replacing human labor and are likely to continue to do so. Nevertheless, the purpose of these systems is to support humans and that is the point that promoters want to emphasize.
An alternative term for artificial intelligence might also better reflect the current state of technology and research. According to Andrew Moore, Dean of Computer Science at Carnegie Mellon, the vast majority of AI researchers are exploring systems that support human endeavors rather than systems that simulate human intelligence.
Alternative terms for augmented intelligence include intelligence amplification, cognitive augmentation and machine-augmented intelligence. IBM, which has invested heavily in artificial intelligence, suggests the term intelligence augmentation (IA), not only to emphasize the supportive role of the technology but also to avoid confusion caused by using AI as an abbreviation.