INFORMATION PROFESSIONALS BEGIN TO ASSESS AI’S PROS AND CONS

The ubiquity of search engines and web-based systems has created issues for libraries. They are concerned about user tracking and protecting the privacy of users. Companies need guarantees of appropriate access to their content, but libraries have legitimate responsibilities to protect the First Amendment rights of their clients’ privacy. This is a tension that will only get more complex as AI and machine learning become more sophisticated and widespread.

As Rutgers law librarian Nancy Talley writes in her 2016 Law Library Journal article, “Agent technology has been used in connection with online shopping to assist in product search and selection. Agents have also assisted digital library patrons with locating materials and streamlining the search process. Additionally, a limited number of academic and public libraries have explored the use of agents in connection with reference and information literacy instruction” (“Imagining the Use of Intelligent Agents and Artificial Intelligence in Academic Law Libraries,” Law Library Journal, 1083(3): 383-402, 2016).

“While patrons continue to interact face-to-face with reference librarians,” Talley continues, “many questions do not require a librarian’s expertise. Integrating agent technology (and more sophisticated artificial intelligence in the future) into academic law libraries is a meaningful way to complement librarians’ work and alleviate some of the burdens placed on librarians” that will allow them to focus on more complex and time-consuming obligations.”

The Library Freedom Institute works to empower librarians and citizens to better understand key issues.

In Jason Griffey’s Library Technology Report article, “Artificial Intelligence and Machine Learning in Libraries,” he predicts that “the near future of library work will be enormously impacted and perhaps forever changed as a result of artificial intelligence and machine learning systems becoming commonplace” (alastore.ala.org/content/artificial-intelligence- and-machine-learning-libraries). The report showcases interesting experimentation that is taking place in all types of libraries to assess the value and potential of today’s AI for enhancing services, improving precision in searching, and allowing for better use of staff time and efforts.

One very interesting new project is FOLIO, “a collaboration of libraries, developers and vendors building an open source library services platform. It supports traditional re source management functionality and can be extended into other institutional areas” (folio.org). This type of collaboration is an important prototype for developing new models of product development that deeply involves libraries themselves rather than having to rely on prepackaged commercial products.

PEW STUDIES AMERICANS’ KNOWLEDGE ON PRIVACY & TECH

In October 2019, the Pew Research Center’s Emily A. Vogels and Monica Anderson released results of broad-based survey “designed to test Americans’ knowledge of a range of digital topics, such as cybersecurity or the business side of social media companies” (pewinternet.org/wp-content/uploads/sites/9/2019/10/PI_2019.10.09_Digital-Knowledge_ FINAL.pdf). “The median number of correct answers was four. Only 20% of adults answered seven or more questions correctly, and just 2% got all 10 questions correct.”

As noted in a January 2018 report titled “Artificial Intelligence and Privacy” from the Norwegian Data Protection Authority (datatilsynet.no/globalassets/global/english/ai-and-privacy.pdf):

“In the future we will face a range of legal and ethical dilemmas in the search for a balance between considerable social advances in the name of AI and fundamental privacy rights. …

“If we find ourselves in a situation in which situations of the population refuse to share information because they feel that their personal integrity is being violated, we will be faced with major challenges to our freedom of speech and to people’s trust in authorities. …”

FIGHTING FOR USER FREEDOM

At the least, as a First Amendment profession, we need not only to find ways to guarantee the integrity of our information products and services but also to educate our users to become effective advocates for their own needs and democratic ideals as well as in the evaluation of the information services and products that they use every day. The Institute of Museum and Library Services (IMLS) has provided sponsorship for the Library Freedom Institute (libraryfreedom.org) “a privacy-focused training program for librarians to teach them the skills necessary to thrive as Privacy Advocates; from educating community members, to installing privacy software, to influencing public policy.” Each year through internships and other efforts, this organization is working to empower librarians and citizens alike to better understand the key issues at stake.

Libraries and information centers certainly have many potential benefits from AI-based products that could reduce routine functions and increase the precision of information retrieval and interactive instructional services. However, as a profession, we have yet to begin discussing our needs and requirements, as well as our requirements for the ethical application of AI in information services Given the high degree of commercialization and low level of privacy standards, the perils of AI need to be addressed before we can celebrate the promise. Efforts like those of FOLIO, the Berkman Klein Center, and Stanford’s Institute for Human-Centered AI are important beginnings. This is a dialogue that needs to challenge us and our institutions in order to guarantee a better future for all of us in an Algorithmic Society.

Sidebar:

OPEN SCIENCE AND OPEN DATA OFFER A POTENTIAL SOLUTION

An example of incredible innovation in open science is Datumbox (datumbox.com), an OA machine learning framework designed to help with the development of data analysis systems, monitoring social media, eliminating spam, and optimizing search engine operation. The Java-based system includes a wealth of algorithms, models, statistical tests, and tools for creating systems for “Sentiment Analysis, Topic Classification, Language Detection, Subjectivity Analysis, Spam Detection, Reading Assessment, Keyword and Text Extraction and more.”

Datumbox was designed by Vasilis Vryniotis, an Expedia data scientist who describes himself as “a proud geek” and who works on the Datumbox framework in his spare time. He explains, “I contribute to other open-source projects and from time to time I publish blog articles about Machine Learning, Statistics and Computer Science.” Vryniotis is another interesting example of the new scientist, the open scientist, that is evolving (blog.datumbox.com/author/bbriniotis).

“Vasilis is the most amazing mind I have met,” brand strategist Audrey Zucker Durieux explains in a LinkedIn post. “Vasilis is sharp, precise, effective and fair in anything he undertakes. He always measures with a strong perception all risks, opportunities and threats implied in the conception of a project. Vasilis definitely [has] got a visionary mind” (uk.linkedin.com/in/bbriniotis). And perhaps this new vision is exactly what we need: open development that is open and available to all.