1 post in this category. Browse all categories →
AI outputs confidence whether it's right or wrong. Research shows models use more certain language when hallucinating than when correct — which means the signal we naturally use to evaluate sources actively misleads us.
Get new issues delivered to your inbox. One idea per issue, no spam.
Get my thinking in your inbox.
One idea per issue. AI strategy, technology decisions, and patterns from 30 years in the room. No spam.
