This paper argues that research findings from the study of teaching could be very useful in enhancing knowledge and tools for the design of both conventional and intelligent instructional software. It also suggests that if the ways in which teachers effectively conduct instruction are different and in certain ways better than other media, and if teachers and computers share the critical capabilities of intelligent information processing, then two things may be predicted: (1) the study of instruction, if limited to instruction delivered by non-intelligent instruments, would not include research and theory regarding what to do with an intelligent instructional instrument; and (2) the study of teaching, being devoted to instruction delivered by intelligent entities, should include research and theory regarding what to do with an intelligent instructional instrument--the computer. A survey of research in the study of teaching uncovered the following names: David Berliner, N. L. Gage, Philip Hosford, Christopher Clark, Robert Yinger, Barak Rosenshine, and Madeline Hunter. Topics and key ideas found to be repeated and emphasized as major areas of interest were teacher effectiveness, teacher thinking, and models of teaching. This survey, however, did not reveal prescriptively-oriented theories and models of what effective teachers do and how they do it, nor were analyses found of the nature of interaction, questioning, and feedback. It is suggested that instructional technologists may have to study teachers and teaching themselves to obtain the information they will need to establish protocols for their "automated teachers." A list of references is provided. (JB)