
David Guttman discusses the use of OpenClaw and AI agents for delegating outcomes in software development.
Principal at Superstruct and author David Guttman joined me on Ditching Hourly to talk about using OpenClaw and AI agents to delegate outcomes instead of micromanaging code. We talked about AI employees, chat-based agents, why OpenClaw feels different from coding tools, how David qualifies repeatable tasks for delegation, secure setup choices like Tailscale, and what collapsing software costs could mean for developers who still sell their time. (00:00) - Introduction to David Guttman (00:41) - From Superstruct Teams to AI Employees (05:19) - Testing Devin Against a Human Engineer (09:39) - What OpenClaw Adds to AI Agents (11:40) - Vibe Coding from Anywhere (15:52) - Pretending You Do Not Know How to Code (21:13) - Software Quality, Risk, and Real Requirements (23:56) - How OpenClaw Has Changed (31:52) - Qualifying Tasks for an AI Employee (37:07) - The Secure OpenClaw Setup Process (40:38) - Local Machines vs Cloud Servers (45:52) - Tailscale and Remote Access Security (48:08) - Teaching Clients to Train Their Agent (53:02) - Ask the Agent Before Asking the Expert (01:00:44) - Life-Changing Client Use Cases (01:03:34) - Disposable Software and the Future of Developers (01:09:44)…
Explore listener stats, chart rankings, contacts and more on the Ditching Hourly podcast page.