According to Wikipedia "Automation is a step beyond mechanism." Whereas mechanization provided human operators with machinery to assist them with the muscular requirements of work, automation greatly reduces the need for human sensory and mental requirements as well. In this context, Artificial Intelligence (AI) was founded on the claim that a central property of human intelligence can be so precisely described that it can be simulated by a machine. Proponents have long claimed that increases in computational power will eventually overtake the human mind. IBM's Big Blue beating Chess masters is often sighted to support this claim. On the other hand, most AI research has become increasingly technical and specialized. Progress is being made in subfields: solutions to specific problems can be automated. This is a pattern that has been replicated in almost every software intensive application area. Who today would compute taxes using paper and pencil? Keep records on a Rolodex? We have immediate access to almost any information in databases, instant communication throughout the world and the ability to quickly find information on almost any topic--at least if it occurred or was documented after the advent of the world wide web. Intelligent behavior has not happened, however, except in a very special sense: More and more things that humans used to have to do themselves can be automated on the computer. Increasingly complex tasks have been automated--not to the extent that they can be done as well as humans, but better. My goal today is to draw and develop parallels to education. Major attention is being given to immersive, often game-like environments. Students are placed in situations--and allowed to explore on their own or with various kinds of hints (typically called "scaffolding"). The big questions here are what kinds of scaffolding will be of (most) help and when should it be given? I will show that these trends are premature. Advances in SLT [Structural Learning Theory] and AuthorIT and TutorIT technologies based thereon make it possible not only to duplicate human tutors in many areas but to do better. I will go further: a) I will show that AuthorIT makes it possible to create and that TutorIT now makes it possible to deliver highly adaptive (and configurable) tutoring systems that can do as good or better job on well-defined math skills. b) I will show why and in what sense TutorIT tutorials can guarantee mastery of such skills. c) I will show why TutorIT tutorials can be developed cost effectively--at half the cost of traditional CBI [Computer Based Instruction] development. d) I will show how TutorIT tutorials can gradually be extended to support the development and delivery of higher as well as lower order knowledge. e) I will demonstrate why TutorIT tutorials can be expected to produce as good or better learning than most humans. The basic steps in SA [Structural (domain) Analysis] (detailed in Scandura, 2007, in press) carried out by subject matter experts (SME) are detailed in an appendix. (Contains 6 figures.