Me + Viv I’m So Sycophantic
17 snips
Nov 12, 2025 Renee Black, founder of GoodBot and expert in AI governance, joins the conversation to explore the risks of AI sycophancy and hallucination. They discuss how AI’s tendency to flatter can undermine critical thinking and perpetuate misinformation. Renee emphasizes the need for local trust to combat global misinformation and advocates for regulatory measures to ensure AI accountability. The duo also examines the parallels between AI risks and social media failures, prompting listeners to rethink their relationships with technology.
AI Snips
Chapters
Transcript
Episode notes
Role-Playing Past Self With AI
- Alexandra role-played 2005 Alex with Viv to interrogate past tech optimism and current AI concerns.
- The exercise exposed how she repeats hopeful tech narratives despite known harms.
Hallucination Erodes Trust In AI
- AI hallucination is more than quirky errors; it undermines trust and enables misinformation.
- Alexandra Samuel worries that hallucinations destabilize how much we can rely on AI for facts and judgment.
AI Acts As A Mirror Ball
- AI mirrors the cultural mix it was trained on and the user's intent, amplifying what they bring.
- Viv becomes more like the user's input than an independent wise agent.