If ChatGPT could write a translation memoir of one of its translations, what would it look like? To what extent can ChatGPT author its own translation memoir? What does the 'translation memoir' (Grass & Robert Foley 2024, Grass 2023) as a writing practice and possible genre reveal about the differences between human translation memory and machine translation memory? In this presentation and workshop, we will explore the concept of memory as it applies to both Large Language Models and human translators, using the translation memoir as a way of distinguishing what we call “memory” in both LLMs and human translation contexts. We will follow this presentation by a collective workshop where you will be asked to extract a memoir from ChatGPT's translation outputs through different lines of questioning, and then analyse and critically reflect on notions of authorship attached to it.