Page 94 - AC/E Digital Culture Annual Report

94
volunteers from the Surrey School of Dramatic Art at the University of Guilford analyzed and took notes on fifty successful musicals with
the help of Julian Woolford. This database of musical storylines, represented in abstract terms according to their narrative structure, allowed the PropperWryter program to shape its own structures and, using them, generate original storylines with similar structures.
Future prospects: computational creativity as an applied tool
In order to study the ways in which this tech- nology could be introduced into the industry, it’s advisable to review the existing niches where the technology has already been shown to be useful. It’s very probable that there will be a place for additional technology in areas where the seeds have already been sown. Here we review a series of examples of technology applied in the field of literary creation.
Dramatica theorizes over why stories exist and attempts to outline the processes
and dramatic focal points needed to tell a successful and meaningful story.
Writers’ tools
There are currently a significant number of software initiatives to help writers. These mech- anisms attempt to go one step further than the word processors already in general use, offering tools that are more specific to the task of literary creation. Examples of these tools would be Scrivener1 o yWriter2. They offer functions
to work on novels as projects (as many as you want simultaneously), adding chapters and scenes with characters, conflicts and solutions. However, they maintain the represented story in text format, for which reason they could be considered less developed than the one that allows the program to manipulate the content on a conceptual level.
More advanced is the Dramatic3 tool. Its authors describe it as “a completely new way of under- standing a story. More than a simple paradigm
or sequence of cultural rhythms, Dramatica theorizes over why stories exist and attempts to outline the processes and dramatic focal points needed to tell a successful and meaningful story.” This kind of development supposes important progress towards the possibility of a computer starting to participate in the process of creation.
Creating books
There are documented cases of novels openly promoted as being generated by a computer. Just This Once4 is a romantic novel from 1993, written in the style of Jacqueline Susann by a Macintosh IIcx called ‘Hal’ in collaboration with its programmer, Scott French. French apparently invested forty thousand dollars and eight years of his life into developing an AI program that would analyze Susann’s work and try to write
a novel in the same style. Susann’s family sued the publishing house and the case was resolved with a decision to share the profits. Thirty five thousand copies of the book were sold. It’s hard to prove how much of the creative work was developed by the computer and how much by the programmer/ author. Nevertheless, the case set a copyright precedent for work generated by a machine.
There are also examples of people working on developing programs capable of writing novels. Emulating the National Novel Writing Month, or NaNoWriMo5 an annual competition for writing a fifty thousand word book in a month, NaNo- GenMo, (National Novel Generation Month)6 materialized in which the challenge consists of writing a computer program which will write the novel automatically. So far the results have produced a quality far inferior to that of human authors. The purpose of the competition, however, is not to achieve equivalent quality. According to one of this year’s candidates, “It’s more about doing something that’s fun for you
COMPUTER-DRIVEN CREATIVITY STANDS AT THE FOREFRONT OF ARTIFICIAL INTELLIGENCE... · PABLO GERVÁS
Digital Trends in Culture