Summary
AI consultant Craig Hewitt, who has worked with dozens of companies on AI transformation initiatives, presents a five-stage maturity framework for understanding where an organization sits on the spectrum from AI novice to AI-native. The video targets business leaders and operators who have moved past basic ChatGPT experimentation but are uncertain how to build more systematic AI capabilities, with particular emphasis on sales, marketing, and go-to-market functions.
The five stages progress from manual chat-based prompting with tools like Claude or ChatGPT, to structured projects and custom GPTs with persistent context and system prompts, to local agentic environments like Claude Code and Codex with MCP server connectors, toward always-on autonomous agents operating with minimal human initiation. Hewitt explains the practical distinctions at each level — when a Claude project is sufficient versus when Claude Code’s ability to work across a full codebase and connect to external services like Google Drive and Notion becomes necessary — and where IT and infosec teams should legitimately get involved in governance decisions.
The framework is particularly useful for teams inside larger organizations that face procurement or compliance constraints limiting which AI tools they can adopt. Hewitt emphasizes skills — reusable system prompts functioning as SOPs — as the key building block for scaling AI work, and closes with a discussion of how pricing and ROI evaluation should shift from seat counts toward the actual value AI delivers to the business.
📺 Source: Craig Hewitt · Published July 01, 2026
🏷️ Format: Opinion Editorial
