Exchange Forum

Virtual Event Logistics and Interactive Features

  • Event Format: The session is conducted in a virtual format, described as "Zoom-ish," designed to mimic a free-format, in-person discussion.
  • The Stage and Eyeball Feature: A specific interface feature at the bottom of the screen, represented by an icon of an eyeball, controls the user's ability to participate audibly.
    • When the "eyeball" is off, the participant cannot speak but retains access to the Chat functionality to type questions.
    • To speak, participants must technically be on the "stage" part of the platform.
  • Interaction Protocols: Participants wishing to engage or answer questions are instructed to use the "raise your hand" feature. Facilitators will then call on them and grant stage access.
  • Alternative Communication: The platform also supports the use of a Dedicated Q&A feature and a chat window for those more comfortable with written rather than verbal contributions.
  • Environmental Context: The facilitator noted the presence of birds in the background during the introduction.

Broad Applications of Artificial Intelligence in Daily Workflows

  • Beyond Email Drafting: The discussion centered on using AI for more than just text generation, focusing on specific data management and formatting tasks.
  • Spreadsheet Management (Excel/Google Sheets):
    • AI assists in generating complex formulas, reducing the need for extensive search engine queries (e.g., standard "Googling").
    • It is used to automate the creation of sophisticated graphs and data visualizations.
  • Reference Formatting: AI tools are employed to handle the often tedious task of formatting citations and bibliographies, which Sonia identified as a significant pain point.
  • XML and URL Verification: AI is utilized to check the integrity of XML code and valid URLs within publishing data.
  • Presentation Structuring: Facilitators use AI to generate outlines for presentations on standardized topics, which are then manually fleshed out.
  • Vibe Checks and Sanity Checks: AI serves as a tool for identifying potential failure points (or "sanity checking") by simulating various ways a process might go wrong.
  • Encoding and LaTeX: AI helps resolve encoding problems and assists in managing the complex syntax of LaTeX, specifically for creating and formatting long tables, which were historically difficult to manage manually.

Deep Dive: AI as a Translator for Technical Documentation

  • Case Study: Similarity Check Program: Christine shared a detailed use case regarding the Similarity Check program (accessible through Authenticate and Crossref).
  • The Technical Barrier: The journal did not initially qualify for the program due to technical complexities involving the submission of a massive spreadsheet of URLs for database verification.
  • The Solution: Christine input the detailed technical instructions from Crossref into Google AI, requesting an explanation in "normal language."
  • The Outcome:
    • The AI allowed for the creation of a documented plan based on simplified instructions.
    • Though the AI occasionally overcomplicated certain steps, it provided the necessary momentum to start the project after a year of stagnation.
    • The project reached a stage of being only 1,8001,800 URLs away from completion, a milestone that will result in significant financial savings for the organization.

AI for Workflow Consolidation and SOP Management

  • Consolidating Standard Operating Procedures (SOPs): Natalie described using AI to merge four or five different workflow documents that contained overlapping but slightly inconsistent instructions.
  • The Tool-First Strategy: The AI was used as an initial tool to perform the "drudge work" of identifying differences and similarities between documents, creating a "cleaner outline."
  • Human Oversight: Following the AI's consolidation, human staff members reviewed the output to remove outdated processes and ensure the final SOP accurately reflected current operations.

Quantitative and Qualitative Analysis via AI Tools

  • Editor Surveys: Jess discussed the annual editor survey, which includes extensive open-ended questions resulting in paragraph-long responses.
  • Categorization: AI was used to categorize responses into themes and sentiments (e.g., Positive, Negative, Neutral).
  • Time Efficiency:
    • Manual analysis of these surveys previously took many hours.
    • Using ChatGPT (and exploring Microsoft Copilot through an ASM license) reduced this effort to approximately 11 hour.
  • Executive Summaries: AI can provide executive summaries of long data strings; however, a warning was issued that AI may sometimes emphasize minor details over major ones, requiring human verification.

Strategic Implementation and Managing Organizational Hesitation

  • The "Show and Tell" Approach: Organizations like the one Caitlin described host departmental meetings to share specific AI use cases, such as those found within Google Gemini (business accounts). This allows staff to see practical benefits rather than just theoretical ones.
  • Specific Productivity Hacks:
    • Author List Formatting: When submitting errata, specific requirements often demand author lists be formatted to remove ORCID numbers. AI can achieve this with high accuracy and speed.
    • Complex Formulas: AI can resolve specific Excel requirements that generic search results cannot easily solve.
  • Problem-First vs. Tool-First: A speaker citing Ashi argued that teams should identify a specific problem first and then seek an AI solution, rather than searching for a problem to solve with a new AI tool.
  • Respecting Skepticism: Sylvia and Anne noted that it should be acceptable for staff to choose not to use AI, as their hesitation may be rooted in valid concerns.

Professional Development and Continuing Education Resources

  • FISC (Force11 Scholarly Communications Institute):
    • Offers courses lasting approximately 33 days.
    • Upcoming 20262026 syllabus includes specific courses on "AI Assisted Literature Review."
    • Content is relevant for publishers, researchers, and librarians.
  • ALPSP (Association of Learned and Professional Society Publishers):
    • Course Highlight: "Effective Social Media for Scholarly Publishers" (offered annually in September).
    • Covers platform selection, measuring success, and improving engagement.
  • LinkedIn Learning: ASM and other institutions provide access to this for professional development; it can also be accessed via many local libraries.
  • Claude Code and Vibe Coding: Anne and Arena discussed using Claude as a programming assistant (described by some as a "Jarvis") to create ad-hoc tools.
    • One example included building a tool to search across an entire journal portfolio when third-party systems like Silverchair Gateway were slow to provide the required functionality.
    • Usage for data visualization was also mentioned.

The Recognition of Expertise in Editorial Offices

  • The Role of ISMTE: The International Society of Managing and Technical Editors (ISMTE) was founded to professionalize the role of editorial office staff.
  • Subject Matter Expertise: There is a distinction between the scientific expertise of the authors/editors and the operational/policy expertise of the editorial staff.
  • Juanita's Anecdote: A deputy editor once questioned why an editorial assistant was giving him advice; the assistant clarified that while the editor owned the science, the office owned the processes, policies, and system protocols.
  • Maintaining Boundaries: The ability to say "no" to an Editor-in-Chief (EIC) is a critical skill for managing professional relationships and ensuring journal integrity.

Ethics, Policies, and AI Disclosures in Submissions

  • Disclosure Trends: Journals are moving toward less complex disclosure questions, allowing authors to provide details as needed while ensuring a "No AI" option is available.
  • The Wiley Perspective: Colin from Wiley suggested that stopping AI use entirely is a "losing battle."
  • Educational Focus: Instead of a "thou shalt not" approach, the focus should be on educating authors on acceptable vs. unacceptable AI parameters to encourage transparency and prevent authors from hiding their usage.

Integrity Challenges: Authorship Stability and "APC Tourism"

  • Authorship Changes Post-Acceptance: Alice highlighted a recurring issue where authors attempt to add or remove names after a manuscript has been accepted.
  • APC Tourism: Kaveh described a phenomenon where authors might add a collaborator from a developing country specifically to qualify for APC (Article Processing Charge) waivers or discounts.
    • Example: Charging a collaborator $500\$500 to save $2,000\$2,000 in fees.
  • Credit Statements and Resistance: While many recommend "CRediT" (Contributor Roles Taxonomy) to prevent authorship disputes, some Editors-in-Chief find them to be a "box-ticking exercise" that does not reflect reality.
  • Policy Correlation: Facilitators suggested that if a journal sees a rise in authorship changes after removing credit statement requirements, that data can be used as an argument to reinstate them.

Questions and Discussion

  • Q: What is the coolest way you have used AI outside of email?
    • A: Responses included formatting ORCIDs out of author lists, coding portfolio-wide search tools, and translating technical Similarity Check requirements.
  • Q: How do you manage expectations or hesitation around AI?
    • A: By using "Show and Tell" meetings and focusing on specific problems rather than generic tools.
  • Q: Are there certifications or classes for continuing education?
    • A: Mentions of ALPSP social media courses and the FISC program for scholarly communications.
  • Q: How do we recognize specialist expertise in the editorial office?
    • A: Support for ISMTE resources and sessions on setting boundaries with EICs.
  • Q: How do you handle fake references in a peer review?
    • A: Alice described a case where a reviewer likely used AI to hallucinate references. The editorial office politely asked the author to inform the reviewer the references couldn't be found, allowing the reviewer to provide corrected information without a direct accusation of AI misuse.
  • Closing Events: The session concluded with a preview of a session featuring Professor Doctor Earl E. Cavalier and a closing quiz show on Kahoot featuring 2020 questions with a 2020-second response limit per question.