This study presents data from a survey of 29 data librarians, primarily from research universities in the USA, Canada and Europe, about how they are using ChatGPT, Bard and AI-enabled Bing. The report lists the percentage of respondents using each of these applications, and what they are using them for. It also presents detailed data on how useful data librarians find each application, and the extent to which they have replaced or supplemented other applications commonly used by data librarians. Respondents also discuss their future plans and measure exactly how much more – or less – personally productive AI applications have made them. The report presents specific data sets for AI use for data cleanup, data visualization, data analysis, data discovery, hypotheses generation, summarizing data results and many other purposes. It also looks at the impact of AI on use of Excel, SAS, Minitab, Tableau, Python and other applications.
Just a few of this 98-page report’s many findings are that:
-
In the entire sample, 17.24% of academic data librarians have used Bard in managing or procuring data, while 82.76% have not.
-
Male respondents were considerably more likely than female ones to use AI for data visualization.
-
Respondents from the smallest colleges were the most likely to use AI for data cleanup.
The data is derived from a sample of 29 data librarians; it is presented in the aggregate and also broken out by institutional characteristics of the respondent's institutions such as enrollment sizes, Carnegie classes, public or private status, and tuition levels, as well as by some personal characteristics of the respondents, such as age, gender, and other variables.