Skip to content
Article Almost Timely News Aug 2026

Christopher Penn: using AI to expand content without weakening it

Christopher Penn, founder of Trust Insights and a marketing analyst who writes the Almost Timely News newsletter, published a methodological piece on August 16, 2026 addressing a question that affects any content team running AI tools at volume: how do you expand and improve existing content without diluting what made it worth reading?

Penn’s central distinction is between automating inputs and automating outputs. His observation: if you use AI to automate your inputs, you are likely to get stronger; if you use AI to automate your outputs, you are likely to get weaker. Teams that delegate drafting to AI often end up with content that is grammatically adequate but experientially thin — it lacks the perspective, specificity, and intellectual commitment that readers find worth the time.

The methodology he describes has five stages, each designed to feed better inputs into the human making the final creative decisions.

The first stage is audience reframing: before expanding any piece, define who will use the expanded version. Executives and practitioners want different things from the same underlying information, and the expansion should be shaped for a specific reader before content generation begins.

The second stage is gap analysis via AI: Penn uses Claude Code to analyze existing writing against external data sources — Reddit discussions, customer data, user research — and surface what the original missed. The output is a set of underrepresented perspectives, not a draft.

The third stage is question generation: identified gaps are converted into interview questions directed at the original content creator. The goal is to capture authentic judgment and experience that AI cannot produce from pattern-matching alone.

The fourth stage is voice recording: Penn dictates answers using physical recording equipment during low-attention time — commutes, walks — to produce audio that carries authentic speech patterns and creates a clear provenance record for the ideas.

The fifth stage is transcription and synthesis: audio is converted to text using NVIDIA’s Parakeet transcription model, then fed back to AI for integration with the original piece.

The methodology requires more setup than prompt-and-publish workflows but produces content that holds up under audience scrutiny. Useful for content teams noticing their AI-assisted pieces are becoming interchangeable, and for individual writers trying to maintain voice and intellectual depth while increasing output.