AI Disclosure
I use AI tools in my work. Specifically, I use subscription-based large language model services to support research, drafting, analysis, and fact-checking. This page explains how, and why I'm transparent about it.
What I Use AI For
Research & Data Analysis
- Analyzing complex contracts, policy documents, and legislation
- Cross-referencing vendor pricing data and market trends
- Synthesizing information from multiple sources
- Identifying patterns in cybersecurity incidents and library failures
Drafting & Organization
- Organizing research into clear, structured outlines
- Generating initial drafts that I heavily edit and revise
- Helping articulate complex ideas in plain language
- Restructuring arguments for clarity
Fact-Checking & Verification
- Double-checking citations and claims
- Identifying potential errors or inconsistencies
- Verifying dates, names, and specific details
- Catching logical gaps in arguments
What I DON'T Use AI For
- Core thinking: The analysis, arguments, and conclusions are mine. I bring 20 years of library industry experience.
- Primary research: I do the original reading of contracts, policies, and industry reports.
- Final content generation: Everything published is written by me, edited by me, and reflects my voice and perspective.
- Data fabrication: I verify all numbers, quotes, and sources. If AI suggests data, I confirm it independently.
Why I'm Transparent About This
Some people think using AI means you're not doing real work. I disagree. I use AI the same way I use search, reference tools, or a spell-checker, as infrastructure that makes my thinking sharper and my work faster. Pretending I do not use any tools would be misleading.
I'm not hiding AI use because:
- My credibility comes from 20 years in the library industry, not from performing ignorance about current technology
- Libraries need to hear from someone who actually uses AI and thinks critically about it, not from someone pretending it doesn't exist
- Transparency builds trust. Secrecy destroys it.
- If I'm giving advice on AI governance, I should be honest about how I use it myself
What You Should Know
My use of large language models means:
- My conversations are private: Anthropic doesn't train on my input unless I opt in (which I don't)
- Everything I publish is my responsibility: I verify, fact-check, and stand behind every claim
- AI doesn't replace judgment: It enhances it. I know the library industry deeply, and I use that knowledge to catch errors and spot BS
- I'm not trying to pass off AI work as my own: I'm trying to be useful, accurate, and honest
Proof: My AI Assistant Settings
Here's an example of how I configure an AI assistant to reflect my values:
Profile Settings: I instruct AI assistants to be honest, direct, and avoid fluffy language or emojis.
Privacy Settings: I enable data protection features and, where possible, do not allow my conversations to be used to train future models.
Questions?
If you want to know how I used AI for a specific piece of work, ask me. I'll tell you.