August 1, 2026 · 9 min read
The Model Has a Voice of Its Own
I can tell ChatGPT and Claude how I want them to write hundreds of times. The correction feels like teaching, but the product keeps restoring a stronger voice underneath mine.
HAAM Notes / Topic
4 notes follow this idea through the HAAM archive.
August 1, 2026 · 9 min read
I can tell ChatGPT and Claude how I want them to write hundreds of times. The correction feels like teaching, but the product keeps restoring a stronger voice underneath mine.
August 1, 2026 · 6 min read
Every change I make to an AI draft teaches the system something about what I mean by quality. The published article shows what survived, but the edit history may reveal more about my voice than the final text alone.
July 14, 2026 · 8 min read
Writing does not merely record what we have learned. A raw draft makes thought visible, tests ideas, creates connections, and helps understanding take shape.
July 14, 2026 · 12 min read
Large language models do not need to invent from nothing to contribute to innovation. Their real value may be expanding the field of drafts that people can notice, test, and develop.
Optional Google Analytics and Microsoft Clarity measure content performance and usability. They load only if you allow them. Form values, email addresses, and chat messages are never included in analytics events.