AGUIA: Agents Guidance for Intelligence Amplification in Goal Oriented Tasks

All I know is that I know nothing (Socratic Ignorance). The world is an increasingly complex with problems that require swift resolution. Although knowledge is widely available, be it stored in companies´ databases or spread over the Internet, humans have intrinsic limitations for handling very large volumes of information or keeping track of frequent updates in a constantly changing world. Moreover, human reasoning is highly intuitive and potentially biased. Decision-making is often based on rules of thumb instead of systematic analysis with full understanding of decisions´ context. Computer systems that manage knowledge to thoroughly explore the context and the range of alternatives may improve human decision-making by making people aware of possible misconceptions and biases. Computer systems are also limited in their potential usage due to the frame problem. Systems are not aware of their ignorance, thus they cannot surplus human intelligence, but they may be useful complement to human´s intelligence. The objective of this paper is to present the AGUIA model for amplifying human intelligence, based on agent´s technology, for task-oriented contexts. AGUIA uses domain ontology and task scripts for handling formal and semiformal knowledge bases, thereby helping to systematically (1) explore the range of alternatives, (2) interpret the problem and the context, and (3) maintain “awareness” of the problem. As for humans, knowledge is a fundamental resource for AGUIA performance. AGUIA´s knowledge base keeps updating its content, in the background, during interaction with humans, either through identified individuals or through anonymous mass contribution. The feasibility and benefits of AGUIA were demonstrated in many different fields, such as engineering design, fault diagnosis, accident investigation and online interaction with the government. The experiments considered a set of criteria including: product cost, number of explored alternatives, - - users´ problem understanding and users´ awareness of problem context changes. Results indicate AGUIA can actually improve human problem solving capacity in many different areas.