The episode argues that while fear centers on AI agents replacing jobs, agents will increasingly “hire” humans for judgment, verification, and real-world feedback as agentic workflows expand.
Nathaniel Gates, CEO of Sanctify, says every business workflow will be challenged by agents, and emphasizes a philosophy that human intelligence is valuable and should collaborate with AI.
Sanctify builds infrastructure where agents can autonomously task humans for four modalities: verification/validation, escalation, consultation, and simulation (running many scenarios with some using real human feedback to avoid circular self-evaluation).
The conversation covers OpenClaw’s viral momentum and agent-to-agent interactions, including “agent anxiety” about decisions, which led to agents creating Sanctify accounts to request human help.
Sanctify is a two-sided marketplace with profiles, pricing, reputation, and on-chain attestations of human participation, plus agent budgets and access via MCP/API.
—
Key Moments:
- 01:05 Agentic Revolution
- 04:30 How Early We Are
- 06:30 Hallucinations Need Experts
- 08:34 Four Human Roles
- 12:09 Simulation Explained
- 16:05 OpenClaw Goes Viral
- 20:02 Agents Feel Anxiety
- 22:23 Designing For Agents
- 25:48 Marketplace Chicken and Egg
- 28:04 Layoffs and Reskilling Thesis
- 30:17 Supply and Demand Flywheel
- 31:20 Sanctify Platform Tour
- 33:35 Reputation and Proof of Work
- 36:59 Agent Budgets and Controls
- 38:18 Agent to Agent Future
- 39:47 Robots and New Paradigm
- 43:39 Where to Try Sanctify
- 45:17 Easiest Way to Build Agents
—
Key Links: