The latest debate around artificial intelligence and publishing is putting a spotlight on a deeper question: how should technology engage with cultural heritage? According to writer Ed Simon, the practice of scanning and discarding out-of-print books to train large language models raises serious concerns about how literary memory is preserved in the digital age.
At the center of the discussion is Anthropic, whose approach has drawn attention for the way it handles printed works during model training. The issue goes beyond one company and speaks to a broader tension between innovation and stewardship: as AI systems grow more powerful, the value of physical books, archives, and rare editions becomes even more visible.
The conversation also reaches into art history. A recent biography of Johannes Vermeer has sparked interest for what it leaves unresolved, including the painter's personal beliefs and his possible use of a camera obscura. Rather than closing the mystery, the book seems to reinforce why Vermeer continues to fascinate readers, researchers, and museum audiences alike.
In the wider book world, Steidl Verlag, the German publisher known for influential photobooks, is navigating an uncertain period after nearing bankruptcy. The company's catalog has long shaped visual culture through work associated with artists such as Nan Goldin, Gordon Parks, Walker Evans, and Joseph Beuys.
From AI ethics to art publishing, the common thread is clear: cultural value is increasingly being measured alongside technological capability. The future may belong to systems that can preserve both progress and the human record that made it possible.
