Table of Contents
Trust and Faith Online
The GenAI Tell-Scan
The rise of generative AI has created an involuntary reflex: scanning every piece of art or writing encountered online for “tells” — signs that it was machine-generated. Artefacts in backgrounds, repeated turns of phrase, statistical smoothness in prose.
This is a miserable way to engage with art. The experience of a painting or a poem — the moment before judgment, the moment of just being inside it — is the entire point. Paranoia colonizes it.
The numbers make the reflex worse, not better: Around 38% of the time, people cannot reliably distinguish AI-generated art from human-made work. Our intuitions are suspect, our paranoia is fallible, and the harm done when wrong is real. There is too much risk in accusing people.
The Labelling Debate
Several movements have emerged for humans to badge their own work:
- Not By AI — badge system for human-first work
- #hibadge2024 — illustrators worldwide drew “Created with Human Intelligence” in their own styles
- Anti-AI 88x31 badges — added to personal sites
The counterargument, made by open-source illustrator David Revoy:
The moment you ask human artists to label their work as human, you've already conceded the terrain. You've agreed that the default assumption is ambiguity. The burden of proof is now on the creator.
We're being asked to prove we're not ghosts in our own house.
Cultivating Faith
The practice: defer the question. When the instinct to scan for tells arrives, sit with the work for a moment longer before reaching for suspicion. Start from trust rather than suspicion.
This is the same principle as good faith code and good faith writing — the first principle of the IndieWeb applied to reception, not just creation.