Musings at the British Library

Of authors, espresso and freshly pressed books

Finding refuge in busy London at the British Library. Writing in the café over an iced latte with a strong yet smooth espresso, I am working on a new book proposal and sample chapter. After an hour, I have a wander round and come across something close to the original Gutenberg printing press from the 15th century, tucked away in a corner of the library. I have seen it before but this time I read the inscription. Books were pressed by hand until steam came in. Books as well as espresso!—I realize as I return to my table and order another iced caffè latte.

And then back to the daily challenge… What’s so hard about writing? It’s not just coming up with the right words, but putting them in the right order. My journalist father used to say that what put him off about writing long-form was the paper work. He wrote on a manual typewriter triple-spaced, feeding in scroll paper so he didn’t have to collate his output. Then he edited on the scroll in a felt-tipped pen. In his later years, he would sometimes just send the scroll in a tube to the publication.

In the early Macintosh 512K and Macintosh Plus days, I used to print out my stories on perforated paper, fan-folded into a continuous sheet, using a dot-matrix printer. I liked the way that my articles felt more integral than if they were printed on separate pieces of paper. That also made them easier to edit. I liked being able to fold out and see two or more pages at a time.

And now, most days, I write and edit on a screen. Letters, words, punctuation, paragraphs, pages, chapters, and then books metamorphose from long strings of ones and zeros, yes or no, yay or nay, on or off. Good or bad? Everything gets rounded off in this brave new AI-world. Having “taught” large language models (LLMs) the ways of ones and zeros, we are finding that they can make a kind of sense with words. A kind of sense. Nevermind that these Chat GPT and other LLM writing programs have been devised by invading the copyright of many thousands of authors, the same ones held here in legal deposit in the copyright British Library.

And where does that leave the British Library which attempts to hold every book at least in English, including copies of my various books? That my books are in the stacks here somewhere — that still gives me a comfort and a thrill, even if they are not safe from the jaws of LLMs. It’s awesome to fathom the creative labours that have gone into the total of 170 million items contained in the British Library, at least 14 million of them books. If a book takes let’s say 2 years of labour, including all aspects of production, that’s 28 million years of human endeavour. Chat GPT can write and publish the equivalent of a book in 3 hours. Future chips and computing power may reduce that to 3 minutes, or 3 seconds. But by any stretch of what then might remain of human imagination, those imposter books will still not stand up at all as books. They cannot be copyrighted and have no place among the writers and books in the British Library.