AI policy
Current thinking on AI tool usage as writer and content creator
Protagonist Science exists to help readers think clearly about science — especially where misinformation muddies the water. That mission only works if you can trust what’s on this page came from me: my reading, my judgment, my writing. AI tools are part of my workflow now, the way search engines and PubMed were before them. This page explains exactly how, so you never have to guess.
Golden rules
1. My name means my claims. Every factual claim, citation, and argument published here has been verified by me against a primary source. If I can’t verify it, it doesn’t run — regardless of whether AI suggested it or I thought of it myself. AI doesn’t get a pass on rigor just because it’s a machine.
2. AI doesn’t decide what’s true, what matters, or what I think. I use AI to summarize papers or organize research — never to originate the thesis, the analysis, or the conclusion of a piece. Those are mine, and if they’re wrong, that’s on me too.
3. The writing is mine, full stop. I’ve read enough AI-generated prose to know it tends to be bland, hedgy, and generic — the opposite of what I want this blog to be. My words are my own, and so are my ideas. AI doesn’t draft my sentences for me; at most it helps me think through material before I write it myself.
4. No fabricated evidence, ever. I don’t publish AI-generated quotes, citations, statistics, or studies without independently confirming they’re real. Hallucination is a known failure mode of these tools, and checking for it is non-negotiable, not optional diligence.
5. Disclosure over vibes. If a piece leans on AI in a way that shaped its content — not just spell-checking — I’ll say so in the post, not bury it in a footer link. You shouldn’t have to come to this page to find out.
How I use AI
Research and literature triage: summarizing papers, surfacing related work, helping me get through a stack of sources faster. I still read the primary sources myself before citing them.
Thinking partner: talking through an argument, stress-testing an idea, or brainstorming scenarios before I sit down to write. The words that end up on the page are still written by me, not pasted from a model.
Science fiction prototyping pieces: AI may help me brainstorm scenarios or poke holes in a premise — never write the piece. Where it plays a larger role than usual, I’ll say so in that post.
Coding for the site: for building or maintaining this blog’s tooling, AI handles the boring, mechanical parts — boilerplate, syntax, debugging. The creative and architectural decisions — what to build and why — stay human.
How I don’t use AI
I don’t use AI to draft my prose. I find AI writing generic and lifeless, and it’s not what I want this blog to sound like — the words are mine.
I don’t let AI originate the argument or ideas of a piece, only help me test ones I’ve already worked out.
I don’t use AI to manufacture sources, data, or expert opinions.
I don’t use AI-generated images to depict real people or real events as if photographic.
Why this matters here specifically
A blog built on fighting misinformation has a higher bar than most, not a lower one. If I get this wrong, it undercuts the entire premise of the site.
This policy is a living document. As my tools and workflows change, I’ll update it to match — but the underlying commitment won’t: my words, my ideas, and my byline mean I stand behind what’s published here.