4 Association of Research Libraries Research Library Issues 299 2019 The term “artificial intelligence” and its three primary related concepts (neural networks, machine learning, and deep learning) are used to varying degrees in the literature today. The presence of the term “artificial intelligence” in Google Books shows a distinct spike in the late 1980s and early 1990s, with increasing use of the term “machine learning” more recently, and, mildly, “deep learning” (see Figure 1). As access to large data sets grew, the potential for artificial intelligence, and therefore funding, also grew. Interestingly, there is a noticeable increase in the term “misinformation” during the rise in the mention of “machine learning.” It isn’t possible from this view alone to determine why there is a precipitous drop in the use of the term “artificial intelligence” in the 1990s, although one could hypothesize that, by then, we were all more aware of the different threads in artificial intelligence, and so used the distinctions like “machine learning” more often. Alternatively, it could be that with the growing interest in the field, publication moved from books to more timely forms of information sharing, such as journal articles. Though causality cannot be proven in any way through the Ngram Viewer, the rise in use of the word “misinformation” is not surprising, and reinforces the significant opportunity for and responsibility of our field. Figure 1
Previous Page Next Page