Call for Prosals - Perspectives on AI

Your Name and Title:

Ken Herold

Library, School, or Organization Name:

California State University, Los Angeles

Co-Presenter Name(s):

N/A

Area of the World from Which You Will Present:

Pacific Time Zone, California, North America

Language in Which You Will Present:

American English

Target Audience(s):

Library and Information Science/Studies (LIS) readers, authors, editors, and teachers

Short Session Description (one line):

A first-person, LIS-oriented account of how ChatGPT and Claude have influenced my research practices

Full Session Description (as long as you would like):

Generative artificial intelligence (GenAI) tools have revolutionized how researchers specify problems, discover and evaluate evidence, synthesize proposed results, and distribute their findings. This session gives a first-person, LIS-oriented account of how ChatGPT and Claude have influenced my research practices, focusing exclusively on library and information science (LIS) contexts as opposed to general or non-LIS subject matters.

I analyze a set of research exemplars in which I integrated GenAI tools into my professional research lifecycle:

1. scoping questions and refining prompts;
2. interactive query formulation and controlled feedback based on my own expertise in LIS and domain-adjacent online resources;
3. literature exploration and thematic clustering;
4. evaluating and revising GenAI artifacts (GenAI stance and voicing, proposed outlines, suggested methods narratives, supposed state-of-the-art statements); and
5. fundamental human assessments during the analysis through conversational correction, interruption, counterargument, and LIS instruction (theory, metadata, discovery, scholarly communication, information behavior, and information literacy).

Across these five touchpoints, GenAI functioned not as a well-informed artificial mentor but as a catalyst in research assistance—supporting faster improvement of uncertain hypotheses to testable queries, from scattered notes to structured text, and from preliminary and disjointed syntheses to clearer writing.

My experience underscored persistent constraints with direct implications for LIS research integrity:

• confident fabrication (including plausible-but-incorrect citations),
• uneven coverage of specialized LIS subdomains,
• opaque provenance of claims, and
• privacy and intellectual property risks when working with unpublished (pirated) machine-learning data, unclassified materials, or proprietary documents.

These limitations clarified the locus of my expertise from merely generating prompts to continuous validation, increasing the salience of critical appraisal skills long central to librarianship: source evaluation, transparency, documentation, and bias awareness. The session contributes a pragmatic, LIS-specific workflow model for an educated GenAI use in research, demanding human verification in all steps and cycles of a research protocol, a simple, efficient and agile schema for prompts and decision trails, and guidance for disclosure statements appropriate to scholarly communication norms. By positioning GenAI as a tool that can meaningfully augment—but never replace—core research competencies, this account aims to help LIS researchers and practitioner-scholars adopt GenAI in ways that are methodologically explicit, ethically defensible, and aligned with the profession’s commitments to evidence, accountability, and information stewardship.

Websites / URLs Associated with Your Session:

N/A