Descriptions:
Alex Lieberman, co-founder of Morning Brew and AI consultancy 10X, joins the David Ondrej podcast alongside Dan, 10X’s director of engineering, to explain how any company can become AI-native — and what separates successful AI transformations from slide-deck consulting. Lieberman describes 10X’s model as a full-stack partner that deploys into client companies rather than handing off a deck, sometimes starting by teaching C-suite executives how to use Claude Code before building back-office agents for invoicing, HR, and accounts payable.
A core concept the pair introduce is “agentic engineering”: the argument that time spent crafting markdown context files for AI systems should exceed time spent writing actual code. Lieberman frames the markdown layer as the highest-leverage work a developer or consultant can do, because it determines how well AI agents understand goals, constraints, and working style. He predicts that in the near future, no engineers — including skilled ones — will routinely read or write raw code.
The conversation draws on 10X’s positioning as a modern Bell Labs analog, self-funding through enterprise revenue rather than venture capital to maintain autonomy and access to hard problems. Lieberman also articulates where humans remain essential in AI-assisted workflows: the “first mile” (originating ideas with genuine perspective) and the “final mile” (judgment and quality control), while AI handles the middle execution layer. The episode includes a sponsored segment demonstrating Deep API, a web scraping and research API designed for use with agents like Codex and Claude Code.
📺 Source: David Ondrej · Published August 20, 2026
🏷️ Format: Interview
