Artificial intelligence in libraries : a philosophical perspective

Abstract

This essay examines writings on the philosophy of artificial intelligence in the library literature. The purpose of this is to assess the awareness of generic philosophical issues relevant to artificial intelligence research, and to see how artificial intelligence is conceptualised. The first part of the essay justifies the Turing Test as a test for artificial intelligence, and explains three generic issues of artificial intelligence research, namely the nature oflntentionality, the Chinese room argument, and the knowledge problem. A hypothetical example is then given to show how these issues are relevant to automating library processes. The library literature is then examined: the findings are that there is an implicit awareness of the issues which is rarely explicitly acknowledged, and that artificial intelligence is generally conceptualised in human terms without explicitly acknowledging the historical precedent for this. Parallel distributed processing is then explained and examined in the library literature. Its perceived advantages for library automation are discussed particularly its ability to simulate the human mind more effectively that conventional computers can. The conclusions are that artificial intelligence in libraries is currently not a reality, and that the generic issues of artificial intelligence research should be explicitly acknowledged to broaden the discussion and awareness of such issues in a library context.