How reality becomes belief—and what we can do about it.
These categories are three lenses I use to study how systems shape belief—and how institutions can respond without losing legitimacy. Use them as organizing themes, not silos: the same trust failure often shows up in all three.
Platforms & Incentives
Distribution decides what reality gets seen.
In engagement-optimized feeds, accuracy is often a disadvantage. Anger travels because it performs well, not because it’s true. That’s why “better messaging” rarely fixes trust: the architecture keeps rewarding the wrong signals.
Governance decides:
- Who controls incentives
- What gets audited
- What accountability looks like at scale
Deepfakes & Verification
Deepfakes don’t only mislead—they corrode legitimacy.
When the archive becomes suspicious, journalism weakens, institutions wobble, and citizens disengage.
AI-forged evidence turns doubt into a default setting. Even authentic evidence becomes contestable, and bad actors exploit that ambiguity. Fact-checking matters, but it doesn’t solve legitimacy at population scale.
Durable defenses are institutional:
- Verification infrastructure
- Clear standards
- Public-facing credibility that can survive inevitable attacks
Trust Metrics
Scale and integrity are the same design problem.
Truth doesn’t spread automatically. It needs incentives that reward clarity, accountability, and consequence.
Most organizations treat trust like tone. It’s infrastructure, built into incentives, enforcement, and measurement. If you can’t define success beyond engagement, you can’t defend credibility when AI-forged evidence hits. The organizations that get this right won’t just sound credible—they’ll endure disruption because trust becomes an asset you can protect.