CONTENT
Interrogate what you see
Does what you see actually hold up?
A strong emotional reaction to content is a signal to slow down, not speed up.
Content generated with the help of Google Omni and Eleven Labs
Our brains are wired to trust our senses — seeing feels like believing — and manipulators exploit this directly. Authentic footage can be lifted from a different event, country, or year and repackaged with a new caption, and AI-generated content mimics the visual texture of the real thing. So the more durable question isn't 'are there visual glitches?' but 'what is this trying to make me feel, believe, or do?'
Ask yourself
Reflection prompts to use in the moment
- 1Could this footage be from a different event, place, or time, repackaged with a new caption?
- 2What is this trying to make me feel, believe, or do — and how fast?
- 3Do I see manipulation mechanics: urgency, a borrowed famous face, moral shock, vague 'experts say', or 'everyone is doing it'?
- 4Are there labels, content credentials, or a detection tool I can check — treating them as one input, not the final verdict?
- 5Does the framing actually match what is shown or said?
Detection tools, watermarks, and provenance signals are genuinely useful — but on their own they have limits, and research found that teaching people a checklist of visual artifacts (unnatural textures, edge anomalies) didn't reliably improve their judgment, because generation tech keeps outpacing fixed cues. They work best in conjunction with reasoning: use any tool or signal as one input, then read the emotional and logical structure of manipulation — urgency, identity mimicry, false authority, moral shock, manufactured consensus — which stays constant across formats. Tool plus reasoning is stronger than either alone.
“A strong emotional reaction to content is a signal to slow down, not speed up.”
Framework progress
Continue to Alignment
You've explored Content — read the next guide or jump straight into a guided case.
Read Alignment guide