In “How are museums using artificial intelligence, and is AI the future of museums?” the author, Lauren Styx, describes several existing examples of AI being used by museums and speculates about future uses. Styx first focuses on very visible uses including several robots that serve different roles in a few high profile museums. While the article doesn’t go into much depth about the robots’ functionality, it sounds like their use of AI is somewhat shallow and the novelty they offer visitors is usually their main appeal. Styx then describes two AI tools that interacted with a set of museum works to form connections. In one case, this was finding similar looking pairs of paintings in the set of museum works. In the other, users took pictures of themselves and were matched with a visually similar painting. I think this type of tool forms a separate category from the robots described in the first section. The robots were meant to replace functions of curators or guides while these image matching tools provided an entirely new offering to museum visitors. Finally, Styx touches on the least flashy but very likely the most common use of AI by museums, behind the scenes operational tools. The article quotes Chris Michaels, digital director of the National Gallery in London who claims, “The major applications of AI will come under the hood of museum operations” and “[Artificial intelligence] should allow the realisation of cost savings in the management of our buildings. Those are often the biggest single source of operating cost in museums, and efficiencies driven by AI could transform under strain business models.” While the article does end by speculating about several incredibly ambitious, immersive tools that could one day be made possible by AI, I think Styx is in agreement with Michaels that most of the benefit from AI will happen without visitors even realizing it.

“The Shape of Art History in the Eyes of the Machine” described research into how machine learning algorithms categorize art from different styles. The article was fairly jargon heavy so a lot of the specifics of their results went over my head. However the main point of their research was to determine how well AI classifications matched up to classifications used by art historians. This was done not for the purpose of developing a tool that could help sort artwork but to determine if there was an empirical basis for the classifications art historians had organically developed over time. The results seem to solidly support this and also provided some insight into what specifically makes the different styles of art distinct.