How memory comes up here In this channel's material, memory appears in the context of an AI-assisted writing setup: what an AI tool retains between and within sessions as it helps with a writer's work [^b0bf7da8-31a6-58e0-9a5b-a27643d43a86]. Why it matters Katie Parrott's Q&A explores how different pieces of a writing setup — style guides, skills, custom instructions, and a voice guide built from writing samples — shape what the AI knows about a writer. Memory is the side of that setup concerned with what the tool carries forward from past context, and it matters because it determines whether output stays consistent or drifts over time. The same Q&A touches on keeping AI output from going stale, which is a direct concern when retained context outlives its usefulness. Where to go next The related Q&A is the place to explore this term in practice: jump to the discussion of style guides vs. skills vs. custom instructions, the segment on building a voice guide from writing samples, and the advice on preventing stale output — each frames what the AI is holding onto and why. - Memory = what an AI writing tool retains from past context between interactions.- It interacts with the writer's setup: style guides, skills, custom instructions, and a voice guide built from writing samples.- Stale output is a memory-adjacent risk, which is why the Q&A discusses refreshing what the AI produces.