Creativity in Human and Artificial Agents

Selmer Bringsjord


Course Overview || Handouts || Class Slides || Student Work || Supporting Links

Course Overview

This course confronts students with the challenge of engineering a proper subset of artificial intelligent agents: namely, creative ones. This course could have been called simply "Human and Artificial Agents," or "Introduction to Intelligent Agent Technology," or some such thing. The problem with this is that then the course would inevitably deal with agents that aren't very interesting. The most interesting agents, whether human or artificial, are ones that do -- in some sense of the phrase -- creative things. Students in this course will strive to build artificial agents that do some really interesting things. There are two main premises in this course, viz.,
  1. To engineer creative artificial agents one needs to study, at least to some degree, creativity and computation in general, including creative human agents.
  2. To manage information technology optimally, one should truly understand that technology. And to truly understand a technology, one needs to -- at least to some degree -- build it (or with it).
We know that creative human agents are generally touted as the best that (at least the intellectual side of) humankind can muster. The entrepreneur who goes from late nights in the garage to the richest person on the planet, the mathematician who achieves immortality via an astounding proof, the person who creates stunning sci fi movies that gross $200 million, the GO player who devises a configuration that suddenly secures victory, etc. --- these are the people we deify as wonderfully creative. How do they do what they do? Can an artificial agent be built that does such lofty things (or at least things like these lofty things)? Can an artificial agent be built that does more mundane and directly marketable creative things? If so, how? How can IT professionals harness agent technology to create and/or manage creative artificial agents? These are the sorts of questions we seek to answer in this course. The `term' agent is used here in the technical sense employed in artificial intelligence (AI) --- a sense that will be specified early in the course with help from one of our texts, Russell and Norvig's so-called AIMA book. It is assumed that creative artificial agents have and will continue to have considerable economic value, and that a significant thrust in IT will increasingly be the creation and profitable deployment of creative artificial agents.

Handouts

Student Work

The Four Stage Venture Capital-Seeking Project

In groups, students in CHAA tackle a four-step process designed to obtain millions in start-up capital (or in sponsored research monies) to launch them toward fortunes of the sort that the "new economy" (which is powered by IT, as Greenspan recently explained when coining this phrase in a March 2000 talk at Boston College) has put within the reach of those who have the right sort of knowledge and drive. In stage 1 (= Paper 1) they pitch their idea to a VC. In stage 2 (= Paper 2), they provide a more detailed document for a technical advisor called in by the VC. This advisor is an expert in agent technology. In stage 3 each team is invited to give a powerpoint presentation directly to the VC and his/her associates. (These presentations are given for everyone in CHAA.) And finally, in stage 4, the complete package (enhancement of stage 2, with screen shots, working code, etc.) is submitted.

Other Misc. Work

Class Slides (Powerpoint)

Esslingen, Germany

Day 1
Day 2
Day 3

United States

Supporting Links


Top Level of Bringsjord's Web Site