Unnecessary AI-Themed Grousing

     I'll start by saying that I really don't care for AI.  This may provoke an immediate reaction; after all, by saying so, I'm aligning myself with the luddites and traditionalists of the world.  It's undeniable that AI has exploded in both popularity and mass-market availability over the years, and the articles this week certainly make a compelling case for how to use it in the classroom.  The Shakespeare translator, for example, represents a huge potential application in pedagogical AI- and it's far from the only example.  In coming years, I firmly believe we'll see more and more implementations in all areas, from information search and retrieval to image generation to difficulty scaling, and in many ways, it will likely make life easier- both for students and educators.

However.

    I have a fundamental problem with generative AI; a holdover from my friendship with several digital artists- the problem I have is with the ethics of generative AI, and how it intersects with plagiarism.  Simply put, AI does not invent its responses out of whole cloth- they must be drawn from somewhere, "trained" from an existing database of reference content.  The problem with this is that this training material is not always, or even usually, freely given for this purpose.  It is often scraped from various sites across the internet, deposited into a vast repository of content from which the AI draws and "learns."  This means that, in a very real sense, all AI-generated content is plagiarized; Because AI cannot tell what it is saying, only what patterns of words are generally seen in response to certain keyword inputs, there is no way to know from whom it is drawing- which original ideas it is echoing, whose work it is reappropriating, etc.  

    This is particularly problematic when using AI to generate images- the training materials represent not only significant work, but also a very personal and unique style from the artist in question.  AI image generation takes those images, breaks them down into constituent shapes, shades, and hues, and regurgitates them.  I consider it tantamount to art theft.  All this is to say that, as we look at the wide and marvelous world of AI and its applications for librarianship, we'd do well to consider the potential ramifications for information rights, and how popular AI platforms can be unethical in their trainings.

Now, onto the actual prompt.

    Naturally, given my reservations, I've opted to look into the VR and AR side of things, and decided to explore Google Arts & Culture.  This resource is honestly a bit of a treasure trove, containing everything from "living" paintings paired with music and captions to fun games and activities like Blob Beats and Puzzle Party.  What I found particularly interesting- and most potential useful for library application- were the virtual gallery tours and 360 videos of exhibits.  Paired with something like an Oculus Rift, this would allow for full immersion in a virtual tour of a famous art museum (or landmark, as the case may be).  One could even host their own virtual exhibition, setting up stations with different artworks/museums and allowing students to cycle through, getting a new experience at each headset.  

    This technology, I believe, represents a much more clear-cut approach into the future of librarianship, and is free of most of the ethical concerns of AI.  For that reason, I am happy to adopt VR and AR into my programming, and would seriously encourage all prospective librarians to look into the same.