Show notes
In this episode, I dig into whether you can trust ChatGPT, and what that question reveals about trust in general. From model changes and user backlash to the three pillars of trust (benevolence, integrity, competency), I share why double-checking is anon-negotiable and how personal responsibility plays into using AI. We talk about media incentives, misinformation, and why “important” is a slippery, subjective word when it comes to verifying the answers that ChatGPT gives you.
Main Topics Covered
- User reactions and model changes after GPT-5 rollout
- ChatGPT as an assistant for known answers
- What “hallucination” actually mean
- The importance of verifying information
- Trust and misinformation online
- Social media’s role in shaping belief
- Media incentives for clicks over accuracy
- The foundations and four components of expertise
- The three components of trust
- Applying the trust framework to ChatGPT
- Where personal responsibility fits in
- Recent ChatGPT use: Suggestion mode and sentence checks
