The Full SLAAIT

We hope to see all of you next week for our hybrid meet up from San Diego!

Tuesday, June 25, 2024

SLAAIT weekly project meeting at 1:00pm. Zoom link here.

SLAAIT researcher Leila Green Little will lead a seminar on AI and book banning, 2:00pm CST. Zoom link here.

Friday, June 28, 2024

Collaborative Institute for Rural Communities & Librarianship (CIRCL) partners meeting from 9:00-11:00am PT at San Diego State University Library. RSVP link here.

Saturday, June 29, 2024

SLAAIT hybrid meeting from 9:00-11:00am PT at San Diego State University Library.

In addition, see last week's newsletter for a list of AI-related sessions at next week's American Library Association Annual Conference.

Instead of a column for this week's David's Corner, I'd like to share with you a conversation on AI in libraries, and answering the questions of librarians about AI recorded at The Netherlands National Library two weeks ago. He speaks with Cynthia Liem, Dutch professor and researcher.

There's some notable quoteables in this talk, in which David and Cynthia discuss the mythologizing of AI, the parallels between AI and other earlier technological advances, and how AI can enrich versus threaten creativity and the pursuit of inquiry.

DALL-E was released in January 2021, and less than two years ago, ChatGPT was released. Since then, many other AI models have appeared, and the excitement around this technology has increased tremendously. The predictions and expectations of artificial intelligence (AI) have proliferated rapidly, with prognostication of both the boundless good and the possibility of the destruction of all humanity due to AI receiving ample interest and criticism in equal measure.

We have been absolutely wowed by the ability of AI to generate pictures limited only by the bounds of our imagination.

We’ve also been disillusioned by the inability of AI to count human fingers and we're less than a month removed from recommendations of eating rocks.

So is it the alpha and the omega, or is this another piece of technology that hasn't yet lived up to its promise?

Right now, it’s a bit of both.

There’s a whole podcast series dedicated to the failures of AI, and one in particular about overpromising and underdelivering. Doing a search online returns dozens of commentary pieces about this precise aspect of AI technology, with multiple pieces in the same major publications within a short span (looking at you, NYT).

The New York Times opinion piece by Julia Angwin (May 15, 2024) indicates AI “is not even close to living up to its hype,” and that its akin to “a bad intern whose work is so unreliable that it’s often easier to do the task yourself.” We are all curious if this current bad intern may beat the odds and end up eventually running the company. And if you’ve spent any time working with AI, you can attest to the simplicity of the errors and some disappointment in limitations—at this point in time. Certainly there are some good points in this piece; however, I suspect that rapid advancement in AI tech is going to moot some of the arguments rather quickly.

Following the money, financial news site Pymnts offered commentary on the so-far under-realized financial potential of AI, and the relatively large investment in large language models (LLMs) vs. other AI subject areas.

A piece from Wired (May 17, 2024) instead believes the hype is exactly adequate, given the naturalness of input and output on GPT-40 resembling the movie Her, the sophistication of Gemini Pro’s answers and dense memory, and the integration of video/audio/text modalities.

The Economist offers that teams of AI models will be even better than we can anticipate, seamlessly blending technology and a convincing facsimile of humanity to give users exactly what we don’t even know we want.

Is AI going to change what librarians do? Absolutely—it already has. How and to what extent it's going to revolutionize the profession and society at large remains to be seen. It's a safe prediction to anticipate more commentary about exactly this question in the near term, though.

Humorist/writer Simon Rich offers commentary on why AI jokes suck in Time magazine. Enjoy.