This Week in Startups Is Anthropic Making the Biggest Mistake in AI History | E2258
135 snips
Mar 5, 2026 
George Pickett, builder of OpenClaw Studio making agent UIs accessible to non-technical users. Logan Allin, operator-turned-VC at Fin Capital focused on crypto and agentic investing. Erik Voorhees, crypto-native founder of Venice AI building privacy-first, tokenized chat AI. They discuss OpenClaw demos, how agents interact with finance, privacy risks of centralized AI, token economics, enterprise revenue tradeoffs, and analytics for agent traffic.
AI Snips
Chapters
Transcript
Episode notes
Solve Cybersecurity To Cross The Enterprise Chasm
- Treat enterprise AI adoption as a cybersecurity problem and control employee AI usage centrally.
- Logan invested in Witness AI because firms need tools that manage how employees bring AI to work and protect proprietary data.
Revenue Growth Fueled Anthropic's Visibility
- Anthropic's revenue ramp toward a multibillion-dollar run rate amplified its public profile and influenced the government clash.
- Bloomberg-reported run-rate jumps (from $9B to ~$19B) show enterprise monetization drove rapid scale.
Anthropic's Safety Terms Sparked Political Clash
- Anthropic set two principled limits: no autonomous weapons and no wholesale domestic surveillance.
- Erik framed these as reasonable safety stances that conflicted with Pentagon demands and triggered political escalation.