Knowledge servers aim to provide knowledge rather than mere information. While information may be “delivered” to the user, such as via Web pages, knowledge is generated within the user by the user’s own thinking processes (Newell, 1982), often stimulated by information received and the user’s situation. Therefore, a knowledge server, though it works by delivering information, is differentiated from an information server by (a) being tailored to the user at the time they connect to the server, (b) taking account of context, (c) making inferences from information provided by the user, rather than merely retrieving data or pages, and (d) engaging in a process of stimulating the user, for example, by letting the user explore different possibilities and obtain explanations. For example, an information server might list some factors that cause stress-corrosion-cracking in steel, while a knowledge server would dialogue with the user to assess the risk in their specific situation and enrich their understanding. While the entire Internet might fulfill such a function, it is possible to create resources specifically designed as knowledge servers.