I want to bring to your attention the artifacts that you or your colleagues may have created and are used in your research labs. I would like to appeal to you to bring these artifacts into your classrooms. Incorporate them into your lab assignments and have your AI students get some experience with them.

AI artifacts that exist in research labs can serve as excellent tools to help bring research into the classroom. They can be used in various ways: as demos that show off the state of the art, as working artifacts of theories discussed in texts, as laboratory exercises where students learn to use them, as case studies for studying concepts, as platforms for developing other AI artifacts. All together, a rich set of pedagogical devices can be available to enhance students' experience with AI.

My proposal is not necessarily novel. Most instructors use some AI artifacts in one way or another. My appeal here is to focus energies into extending the boundaries of use of these artifacts. If you or your research group has produced an AI artifact, it would be worthwhile examining its use in the classroom. For example, is it something you can share with undergraduate students? With graduate students? In what form? Can you give a demo during a lecture? Would a short video clip suffice? Could the students operate it themselves? What types of lab assignment would highlight the main features of the artifact? Could it be used for students to do development work?

The use of Ai artifacts in the classroom requires planning and effort at different levels. The primary responsibility rests with the creators. First, they have to try to answer some of the preceding questions in order to help create appropriate pedagigical materials. Next, the materials can be tested in their own courses and then to disseminated for others to use.

Currently, most of the burden of developing such materials rests on authors of AI texts. Authors feel obliged to provide artifacts specifically designed for use with their texts. This works, at times, but it essentially eliminates the in situ nature of artifacts that emerge from research labs. The emphasis should also be on "doing AI" as done by others in the field. Besides, using artifacts also creates a fertile pipeline of well-trained individuals who will be able to contribute to the field in the future.

At a recent AAAI symposium, researchers lamented their own lack of training in certain new areas into which their research was leading. Some people complained about not having sufficiently trained students who could participate in some of their research projects. An audience survey revealed that a very small percentage of them were using any of their research tools or artifacts in their classes and agreed that incorporating more of tools and artifacts would certainly alleviate some of the concerns about student preparedness.

I would like to invite readers to send me short descriptions of AI artifacts that have been successfully used as teaching materials in AI courses. Entries submitted should be such that others will also be able to use the artifacts in their own courses. I will collect and present them as a resource in a future issue of this magazine. If you are willing to write a longer, more detailed tutorial or presentation, please contact me to discuss the details.