An Ocean of Data

By the time I finish writing this sentence, vastly more data will have been generated around the world than was produced in the entire 18th century, indeed in the entire history of humanity prior to around 1900.

This sounds like some crazy thing someone might say just to make small-talk at a party, but it’s literally true. In the late 1990s, all the books in the Library of Congress were estimated to contain around 20 terabytes of data (20 million books of roughly 1MB each). In turn, a figure I have come across repeatedly estimates that there were around 337,000 books published between 1700 and 1800, though it is not clear whether this figure includes books published in East Asian and South Asian languages, or only in the languages of Europe and its colonies. So let’s go ahead and double it, bringing the total number of 18th-century books to 674,000, at around 1MB each (let us say). And let’s double that figure in turn to include all the letters and notes and church baptismal scrolls and so on, so that there were, worldwide, a total of 1,348,000 MB of data, or 1.348 terabytes, produced in that hundred-year period.

In 2023, there were about 328.77 exabytes of data produced per day (the figure is surely much higher now, almost three months into 2024). One exabyte is equivalent to 1 million terabytes. At the average 2023 rate, this is the same as 13.69875 exabytes per hour, or 228.3125 petabytes per minute, where a petabyte is equivalent to 1000 terabytes. So, every minute in 2023 there were about 169,371 times more data produced than in the entire century that gave us, among other great achievements, the Encyclopédie.

Do you understand what this means? I am, right now, pissing into the ocean. Sure, I create a little cloud that brings about a few seconds of alteration in this nano-region’s pH balance. But the ocean doesn’t care, and almost immediately restores itself to its previous homeostasis. And if the ocean is increasingly threatened with a loss of homeostasis, this is not because of my own feeble stream, but because human industry is pumping massive quantities of plastic and toxic waste — the counterparts, in our ongoing analogy, to the massive quantities of AI-generated sludge that now constitute the great bulk of the data I attempted to measure, no doubt imprecisely, in the previous paragraph.

The figure of the “writer”, as opposed to his earlier ancestors such as the “scribe”, probably only emerged around the end of the 18th century. But obviously —obviously!— whatever we imagined that figure to be doing at that moment just cannot be what it is today, any more than pissing into the ocean can be reasonably interpreted to be the same sort of act as pissing into the bathtub.

Writing is dead. Dead, dead, dead! It had its moment, but now humanity is moving on to new and different things.

In other news, I’ve been relatively inactive here recently only because I have finally managed to get past my task-specific writer’s block, and seriously to get to work on my book, the manuscript of which is due this summer. It’s going smoothly, I feel inspired, my mind is racing with ideas, I’m in my zone, etc. This is good. Creation is still good. In fact it’s all we’ve got left. My book is going to be good. And anyhow I still get paid, no matter how much data will be circulating out there —much more than today, certainly— by the time it goes to press in early 2025, thanks to the zombie-like afterlife of a publishing industry that has not yet fully faced up to the world-historical transformations the data-revolution has wrought.

Have a great week!

—JSR