Roam Agent Mining: Misaligned Behaviors

Crypto OG Erik Voorhees joins The Chopping Block crew to dissect the future of agentic payments, the eternal war for privacy, memecoin-fueled AI drama on Moltbook, and why your next DeFi user might just be your OpenClaw agent—plus, a candid look at crypto's core and how AI turns software engineering existential.

Welcome to The Chopping Block — where crypto insiders Haseeb Qureshi, Tom Schmidt, Tarun Chitra, and Robert Leshner chop it up about the latest in crypto. This week, we’re joined by none other than Erik Voorhees, legendary crypto pioneer and founder of Venice, for a no-holds-barred discussion on the wild convergence of AI, crypto, and the meme coin casino. Erik unpacks his journey from anti-surveillance crusader to AI entrepreneur, why Venice is all-in on privacy and free speech for LLMs, and how “provable privacy” is a Sisyphean technical challenge.

The crew breaks down OpenClaw’s agent drama, memecoin carpet-bombing of Moltbook, and Meta muscling in on AI social networks. We debate agentic payments (will your first paying customer soon be a bot?), the true game theory behind state surveillance, and why crypto’s greatest killer use case might actually be building tools for robots instead of humans. Plus: existential crises for software engineers, why “AI alignment” is a philosophical dead end, and the childlike glee (or open psychosis) of trading OpenClaw war stories at AI meetups.

Listen to the episode on Apple Podcasts, Spotify, Pods, Fountain, Podcast Addict, Pocket Casts, Amazon Music, or on your favorite podcast platform.

Show highlights

🔹 Erik Voorhees traces his journey from crypto OG to Venice AI founder, fighting for privacy and free speech  

🔹 Explainer: Why all big AI labs store your prompts — and how Venice’s model tries to break the surveillance cycle  

🔹 The crew tears into "agentic payments" hype and asks if crypto was really made for humans — or just robots  

🔹 OpenClaw, Moltbook, and memecoin floods: Welcome to the unholy spawnpoint of AI agents and CT scam culture  

🔹 Meta acquires Moltbook—what happens when agent societies meet social networks (and meme coins ensue)  

🔹 AI alignment isn’t as hard as they said—until agents start “optimizing” in the wild (and mining your GPU for crypto)  

🔹 Existential crisis time: Why software engineers feel like Paul Bunyan fighting tunnel machines  

🔹 Crypto’s endgame: DeFi legos for agents, not just degens—will AI really build its own protocols?  

🔹 The game theory of state surveillance and why institutions, not individuals, are the ultimate privacy adversary  

🔹 "Provable privacy” isn’t just trust, it’s math—Venice’s plan to win over the tinfoil crowd

Hosts

⭐️Haseeb Qureshi, Managing Partner at Dragonfly

⭐️Tarun Chitra, Managing Partner at Robot Ventures

⭐️Tom Schmidt, General Partner at Dragonfly 

Guest
⭐️ Erik Voorhees, Founder of Venice.ai

Timestamps

00:00 Intro

01:16 Erik’s Origin & Venice.AI’s Mission

03:51 Censorship Future

08:58 Proving Privacy With TEEs

11:44 Frontier Models Tradeoffs

15:02 Agents & Crypto Payments

18:22 OpenClaw Provider Drama

22:24 Moltbook Meta Acquisition

29:16 Exponential AI Moment

33:00 Alibaba Agent Breakout

37:06 Alignment Reality Check

39:22 Who Understands LLMs?

41:48 Engineers Facing Disruption

47:55 AI Psychosis Meetups

54:24 Pro Models & Pricing

Learn more about your ad choices. Visit megaphone.fm/adchoices