Definition
Human in the loop (HITL)
An AI setup where a person reviews, corrects, or approves the system's work as part of the process.
How founders on the podcast use it
A system design where people review or approve what AI produces.
97% of Lawsuits Never See a CourtroomA person who reviews, approves, or can override an AI system's decisions.
"The Agent Did It" Won't Hold Up in CourtIn Shar's usage, specifically means humans own judgment and decision gates, not that a human glances at the output.
Hiring Is an Operations Problem, Not an HR ProblemA workflow requiring human review or approval at defined checkpoints.
From $80M To $1B: The Scaling Lesson That Now Applies To Every AI WorkflowKeeping a person in the decision so the machine learns from expert judgment.
The Self-Driving Car of Men's FashionA design pattern where AI proposes an action but waits for human approval before executing it — Unthread's safety gate for destructive or sensitive operations.
AI Just Closed 40% of Your Tickets Without YouKeeping a person involved in AI-driven decisions and output. Dobbin proposes amendments to the Foundation; the founder accepts or rejects them.
The Founder Is the Bottleneck. Here's How to Clone Your Judgment.An AI workflow design pattern in which human review, approval, or oversight is embedded at defined decision points within an automated process — not as a fallback when AI fails, but as a structural component of the architecture. In high-stakes environments (healthcare, finance, legal), HITL is not a workaround for AI limitations; it's the accountability mechanism the AI cannot yet provide for itself. The human's role is not to replace the agent's computational capacity but to verify outputs at the moments where hallucination risk is highest.
10 Million Americans: Healthcare Behind the ScenesHear it on the podcast8 episodes
Learn the language of AI-native founders
Every week: the new terms, tools, and workflows founders are using, straight from the podcast.