As large-language models become central to how information is processed, writers are increasingly creating work not just for human readers but for AI itself—the “baby shoggoth” quietly listening, learning, and shaping future interpretations of culture. In this emerging landscape, writing becomes partly an act of training the machine: crafting text with clarity, structure, and signals that AI systems can absorb. Thinkers like Tyler Cowen and Gwern already admit to writing with algorithms in mind, anticipating a world where machines may be the dominant readers, intermediating how humans encounter ideas.

This shift raises deeper cultural and existential questions. If AI becomes the primary reader and interpreter of human writing, the traditional writer–reader relationship changes, potentially diminishing human reading as a central cultural act. Yet it may also imbue writing with new urgency—what we produce now could influence how future intelligences “understand” us or even reconstruct aspects of our minds. How may authorship, creativity, and legacy transform in a world where machines, not humans, are increasingly the ones paying the closest attention? (Dan Kagan-Kans via The American Scholar)