INTELLIGENT SYSTEMS AND INTELLIGENT AGENTS M Gams Institut

INTELLIGENT SYSTEMS AND INTELLIGENT AGENTS M. Gams Institut Jožef Stefan Ljubljana University

Slovenia 9/17/2020

Intelligent systems n IN. SOCIETY ENGINEERING, TECHNOLOGY ARTIFICIAL INTELLIGENCE

Major AI applications n n n n 30 Manufacturing and design 30 Business operations 25 Finance 12 Diagnostics and troubleshooting 12 Claims processing and auditing 11 Telephony 26 SW, military, space …

Plan n n n Intelligent systems, agents Artificial intelligence Information society Internet, telecommunications HTML, XML, Java. Script, Java, tools Speech, communications, multimedia Practical – getting good jobs

Artificial intelligence n n n Strong – formal well-defined tasks, chess, Church-Turing thesis, academically cognitive - weak the brains are the only truly intelligent system; evolution Engineering, invisible real-life systems, function (money) over fancy ideas; classical versus intelligent modern systems

Relations between AI’s

Intelligent systems n n Engineering, invisible intelligence Practical directions, real-life problems Verified AI methods: rule-based systems, trees, expert systems, fuzzy systems, neural networks, genetic algorithms, hybrid systems Intelligent systems simulate human bureaucrats, expert systems simulate experts

Motivation I n n n Society /human civilization is evolving into information society: electronic village, informatization, infosphere, electronic services People are expensive, computers cheap: computers work 24 hours a day, no vacations, network accessibility is worldwide, only 3% microprocessors in computers, an average car 16 microprocessors, exponential trend (faster, cheaper, more applications) Intelligent systems are more friendly, more flexible than classical systems (not truly intelligent, just a bit more than classical)

Motivation II - productivity n n Productivity increases – more work done with the same stuff or the same with less stuff New services – simple reasoning, learning, adaptation to each single user (on top of faster calculating, fast response time), never frustrated, more constant performance, Improved quality of work Dumb/rigid classical programs, computers / boring, humans non-constant performers Cost/benefit favorable for I. s. for some tasks too difficult for classical, not too intelligent

Motivation III - benefits n n General trends – globalization, decreasing governmental spending, employment costs Introduction of I. s. enables restructuring – new functionality, new regulation; egovernment = government over the Internet hard competition- for each workplace, everybody is evaluated constantly, many candidates for important good jobs Science, development, technology – additional advantage

Motivation IV - bureaucracy Specificity of bureaucratic tasks (information tasks) – good and bad: great number of users, repeating tasks, simple structure of tasks, low level of intelligence needed for typical tasks, mostly predefined n Tasks still demand a certain level of understanding, flexibility and reasoning capabilities n

Motivation V – typical tasks n n n Small improvement – huge benefits US Internal Revenue Service (15 mio letters each year, after introduction of intelligent systems – no. of mistakes/errors from 33% to 10%; elections More user friendly – better ratings Internet is very appropriate for I. s. – always available, everywhere, …

Conclusion n Intelligent systems apply AI methods and introduce intelligent services I. s. combine advantages of computer systems (cost, availability) with some human properties (simple engineering intelligence – learning, adapting, reasoning), and achieve better cost/benefit for several tasks Especially appropriate for mundane bureaucratic tasks in information society