Sunday Book-Thought 51

The reality, of course, is that every writer’s individual habits and practices are deeply personal and idiosyncratic, and it is difficult, if not impossible, to extract patterns in support of generalizable conclusions – beyond the intense intimacy and commitment that the act of writing invariably demands. Some writers dictate aloud. Some write longhand and then type their work on a typewriter or computer. Some compose at the keyboard but then print out their work for handwritten revision. Others don’t need the hard copy. Some writers print everything out, mark it up, and then retype it themselves. Others hand it off to an assistant. A few still revise by (literally) copying and pasting strips of text. Some writers find the computer alienating, intimidating. Others see it as an intimate, profoundly unmediated experience. Some writers value the slowness of the pen. Some value the speed of the keyboard. Some chafe at the labor of retyping, others embrace it. Some writers are enamored by the small rituals of the process, the changing of the worn-out ribbon, the bright bell of the carriage return. Others are unsentimental. Some writers require an absolutely specific instrument – or setting, or time of day, or slant of light, just so. Others write on anything and everything, anytime, anywhere. Work your way through “The Art of Fiction” and its kindred features in the Paris Review – what George Steiner, in his own “Art of Criticism” interview, describes as “the largest collection of insight into this of any publication” – and you can find accounts describing all of these and more. The one inescapable conclusion is that our instruments of composition, be they a Remington or a Macintosh, all serve to focalize and amplify our imagination of what writing is.
Matthew G. KirschenbaumTrack Changes: A Literary History of Word Processing ( Cambridge, MA: The Belknap Press of Harvard University Press, 2016), pp. 22-23


Sunday Book-Thought 50

One consequence of this model of intelligence, especially of any computer implementation of this model, is that such a system runs the risk of being described as only appearing to be intelligent without actually being intelligent. In some sense, intelligence is in the eye of the beholder, and most beholders are prejudiced by skin. If the skin of the teller of the story is fleshy and humanlike, we are likely to consider the algorithm that produced the story to be an intelligent one, except perhaps in the case of the grandfather who we would agree was intelligent but is now telling the same story too often. But if the skin is plastic and we suspect that a computer is inside we are likely to claim that the algorithms being used to produce the same story were somehow just unintelligent retrieval methods.
In the end all we have, machine or human, are stories and methods of finding and using those stories.
Roger SchankTell Me a Story: Narrative and Intelligence (Evanston: Northwestern University Press, 1995), p. 16