Proceedings of The Second Americas Conference on Information Systems (AIS'96), August 16- 18, Phoenix, AZ. Philosophical Foundations Mini Track Beyond Intelligent Systems: Listening to the Ghosts In the Machines Steven K. Wyman , School of Information Studies, Syracuse University <skwyman@mailbox.syr.edu> Ping Zhang , School of Information Studies, Syracuse University <pzhang@mailbox.syr.edu> If we look back 50 years in computing history, we can compare Vannevar Bush's vision of what information technology might one day be able to accomplish to what it has since become. Bush foresaw the use of the computer as an effective memory system, with which humans could store and retrieve information, as well as sort data, and trace a kind of hyperlink trail of reasoning. Not only that, but in Bush's memex machine, one would even be able to print off or magnetically store an information set for delivery to others. Everything Bush described has come to pass, and in nearly every case exceeds what he imagined. Bush was prescient indeed, but he also occupied a good vantage point: he directed certain military research in the war effort. He was familiar with rapid developments in prototype feedback experiments such as fire control and communications theory, among other things. In early fire control a system's state advanced through iterative looping of feedback through the system. Output was re-entered as input. Refinements in such processes led to the birth of cybernetics. In this phase of the information revolution cybernetic systems combined machines doing environmental scanning with human intervention providing the intentionality, or wisdom, to the system. Machines appeared fully capable of combining data structures into information constructs. Information domains were defined by task. Feedback and Distributed Systems As computers matured, the concept of feedback emerged as a central feature of human computer interactions. Expert systems employ feedback from one or multiple users to determine relevancy and retrieve appropriate data, or to forecast plausible scenarios. In systems design, we can take data directly from the system's users via keystrokes or other input. We can have intermediaries such as professional online searchers reformulate input into more refined categories. We can even unleash genetic algorithms to "learn" about and mine data for potential knowledge. Artificial intelligence research inspired some to expect that machines might be constructed with autonomous intelligence. In this view, machine intelligence might be devised to solve problems without [at least direct] participation of users. This approach, strongly reflected in the information technologies influencing organizational restructuring, dissociates the human from the machine, in effect constructing an artificial and abstract environment in which processes are efficiently automated. Human feedback is removed as much as possible. Many information systems began to take on a self-contained aspect. Mechanical systems are becoming highly knowledgeable in artificial environments. The result is an extremely rational model of work, processes, and intelligence. Problem solving, or task, still defines the boundaries of the system's behavior and utility. In other words, humans still determine the intentionality directing the information system. The system may be effective and efficient as well as organizationally and socially transforming, perhaps even knowledgeable, but it is not intrinsically wise. This context must be provided by the system's architects and users. The Internet was created to resolve a problem. Like so much of Bush's insight, it originated from concerns with communications and warfare. However, over the past decade we have all witnessed the emergence of computer networks as highly distributed, highly social, and very open systems. Feedback is pouring into the networks in unprecedented volumes. In many cases this feedback targets tasks and problem solving. But in as many other instances, what is occurring is exploration, socialization, or even plain mischief. There is randomness loose in the networks. At the same time, and perhaps not so coincidentally, there is creativity, self organization, and adaptive behavior. Networks are not simply computers communicating; they are