Human vs Computer Creativity

 

In the fall of 1997, Indiana University cognitive scientist Douglas Hofstadter, in his role as Visiting Professor at Stanford University’s Center for Computer-Aided Research in the Humanities (CCARH), organized a series of five public symposia centered on the burning question “Are Computers Approaching Human-Level Creativity?” At the time, Hofstadter had been very struck by a number of then-recent claims that computers were indeed on the verge of human-level intelligence and creativity. Although he himself was highly skeptical of those claims, Hofstadter wanted to raise the issues in public with the help of reliable and thoughtful authorities, so with the support of Walter Hewlett, Director of CCARH, he managed to bring together a number of world-famous experts in various fields to discuss state-of-the-art computer programs, and what they could and could not do.

The five symposia were stretched out over a couple of months, taking place on alternate weekends of Stanford’s 1997 fall quarter. Each symposium occupied a full Saturday in a large auditorium on the Stanford campus, except for the one on music, which was two days long (a full weekend). Roughly 200 people attended each symposium, except for the third one, which was attended by about 700 people (and it took up only one Saturday evening rather than a full Saturday).

The first symposium was about humans versus computers as chess players and as Go players (at that time, world chess champion Garry Kasparov had just been defeated by IBM’s Deep Blue system, a very provocative result).
http://hdl.handle.net/2022/21744 (Chess and Go part 1)
http://hdl.handle.net/2022/21765 (Chess and Go part 2)

The second symposium was about the most recent successes of computers in the area of processing natural language -- both writing it and understanding it (including writing poetry, grading students’ papers, answering questions about science-fiction stories, and so forth).
http://hdl.handle.net/2022/21766 (Language and Literature part 1)
http://hdl.handle.net/2022/21767 (Language and Literature part 2)

The third symposium was about computers as creators or understanders of jokes and other types of humor, and featured renowned comedian/actor Steve Martin, as well as famed AI pioneer Marvin Minsky.
http://hdl.handle.net/2022/21768 (Jokes and Humor)

The fourth symposium was about a particular computer program -- David Cope’s “EMI” (later renamed as “Emmy”) -- as a composer of music in the style of various classical composers. On the second day, a two-hour concert took place in which compositions written by EMI and compositions written by famous human composers were performed without identification, and the audience was asked to vote for which pieces they thought were human-composed and which were computer-composed. Among the speakers in that symposium were David Cope himself and noted philosopher Dan Dennett.
http://hdl.handle.net/2022/21770 (Musical Composition part 1)
http://hdl.handle.net/2022/21771 (Musical Composition part 2)
http://hdl.handle.net/2022/21772 (Musical Composition part 3)
http://hdl.handle.net/2022/21773 (Musical Composition part 4)
http://hdl.handle.net/2022/21774 (Musical Composition part 5)
http://hdl.handle.net/2022/21775 (Musical Composition part 6)
http://hdl.handle.net/2022/21776 (Musical Composition part 7)

The fifth symposium -- the final one -- served to summarize the ideas that had been debated in the four previous symposia, and tried to answer the title question that had provoked the series of symposia.
http://hdl.handle.net/2022/21777 (The Big Picture part 1)
http://hdl.handle.net/2022/21778 (The Big Picture part 2)
http://hdl.handle.net/2022/21779 (The Big Picture part 3)

Technical details for the digitization methods for this collection are available at http://hdl.handle.net/2022/21769

Recent Submissions