Lieberman argues that poor AI output reflects poor user input, not the technology itself. He built a six-step Content Machine in Claude to overcome the blank page problem and produce content that sounds like him. The system starts with an Oracle that scans internal systems and the internet for content spikes. After selecting an idea, an Interview Panel uses codified personas to interview him, and he dictates answers via voice-to-text. A personal style guide and lessons file ensure drafts match his voice, while a Writer's Council scores and revises posts until they reach a 10/10. Lieberman emphasizes that the interview step, not drafting, is where AI slop originates, so the machine uses only his words from transcripts. Beyond AI, he advocates for employee advocacy through the Tenex Creator Cup, a $5,000 internal competition that turns content creation into a team sport. He defines a great engineer in the AI era as someone with deep system understanding and adaptability to agentic workflows.