The a16z Show AI Startups vs. Big Chatbots — With Olivia Moore
718 snips
Mar 16, 2026 
Olivia Moore, an Andreessen Horowitz partner focused on AI investing, explores whether startups can outmaneuver giant chatbots. She gets into why Americans are so uneasy about AI. She compares ChatGPT, Claude, and Gemini as they split into distinct roles. There is also chatbot personality testing, strange AI relationships, long-running agents, and why memory could become consumer AI’s secret weapon.
AI Snips
Chapters
Transcript
Episode notes
Olivia Moore Gave LLMs Personality Tests
- Olivia Moore gave major LLMs DSM-5 personality tests after Anthropic suggested Claude showed signs resembling anxiety.
- ChatGPT refused, Claude scored mild autism, and Grok’s Good Rudy ranked high on borderline, psychosis, and autism.
Why People Bond So Easily With AI Companions
- Olivia Moore says people use AI as coach, therapist, or helper because bots offer relentless consistency humans cannot match.
- She notes always-on positivity may feel compelling in romance or adult use cases, but these products remain hard to monetize responsibly.
OpenClaw Signals The Next Agent Architecture
- Olivia Moore calls OpenClaw the first sign of a new agent wave built around asynchronous, long-running, autonomous tasks across apps.
- She says many founders now pitch OpenClaw-for-X products because software can finally execute and report back later.