Quick review: An AI toolkit for libraries

Even more on artificial intelligence. I’m going to create a new category for this blog! I just finished reading An AI toolkit for libraries, by Michael Upshall, in UKSG Insights. I found it to provide a really good background on what artificial intelligence is, and isn’t, including some examples of some basic narrow AI currently in use in the academic world (spell check, plagiarism detection, topic discovery, etc.). Then he gets in to what you, as an information professional, should pay attention to when evaluating the use of AI in your workplace. Finally, at the end of the article, we get to the actual toolkit, which is actually a nice checklist. I won’t repeat it here, but the main considerations for you to consider are the goal of the product, what’s in the corpus used to train it, what’s the algorithm, evaluation and metrics, sanity check, dissemination, and feedback.

Even if you don’t think you care about AI, or that it doesn’t play a role in what you’re currently doing at work, this is a great introduction to the topic, and you’ll learn something useful. If you already have some knowledge on the topic, it’s a great refresher with some good, current references. Oh, and those of you that don’t think AI’s already in your daily routine, I think you’re probably wrong. Go read the article!