As a writer, am I worried about this new generation of A.I. capable of human-like language at inhuman speed and efficiency? Yes. Though not because I think it will ever write better than human writers. ChatGPT’s moral void, its fundamental inability to understand the human condition, shackles its meaning-making capabilities. Devoid of the shared humanity between author and reader, A.I., at best, can create only crude imitations of what’s already been written in aggregate.
What actually worries me is that A.I. writing will become so ubiquitous that all writing produced from here on will need be met with wary skepticism, if not outright distrust. Just as Instagram robbed us of our faith in photography as an accurate reflection of reality, perverting our collective sense of beauty through repeated exposure to images of ersatz perfection, the power of A.I. to write so much with such ease may, in time, become irresistible to all but a few stalwart luddites akin to film photographers and vinyl aficionados. And, just as the glut of information made available by the internet led us to prioritize skimming over focused, deep reading (leading, perhaps, to the rise in popularity of Y.A. and short fiction, and even steeper rise in not reading books at all), A.I. writing may become so polished, fed by an ever-expanding trove of data and algorithmic understanding of user engagement, that authentic human writing could come to feel sloppy and unfocused (as humans themselves are). If/when we reach that point, how much longer will it be before we’re unwilling to (or incapable of) deciphering the literature of our own human past without assistance from an A.I.?
Is the master who uses a tool that subverts its master really a master at all?
The Everyday Essayist 