Let's not knit pick through historic change

Many people are awakening to the significant changes artificial intelligence is beginning to have in various endeavors. But most of what we are individually aware of can amount to nothing more than a diversion away from what is actually important, even critical, regarding AI at this point in history. Less attention should be paid to the distraction lest we miss the central reality of the age: it could be wonderful, or it could be catastrophic.

“Since late November of 2022, when ChatGPT was officially released, savvy readers began to notice signs of an emerging AI dialect slipping into pieces that had supposedly been written by humans. A flurry of ‘—’ em-dashes: Countless sentences structured like ‘it’s not x, it’s y.’ A strange fondness for the word ‘delve.’”

That undoubtedly true but equally unimportant observation is from a recent article in Range on how widespread the use of AI is in opinion articles.

Most of the examples in that paragraph regularly appear in my own writing and have for decades. Em dashes (long dashes like this —) were even prescribed by Associated Press Style, for example. The “it’s not, it’s y” construction, though, I have noticed far more than normal in the last few years, but let’s not bash AI work just because it comes from AI.

Far more important in the realm of writing is whether our ability to effectively communicate is increasing or decreasing. To that end, I’ll note that since AI hit the popular desktop, I’ve seen far less need for simple editing of submitted letters and columns. To write clearly is not as common a skill as we might imagine. In my experience as an editor, the vast majority of us have ideas worth evaluating but lack the skills to put them in writing without distracting mistakes in spelling, homonyms (my personal downfall), punctuation, or word choice. The ability to logically structure a sentence began to disappear about the time educators started backing away from one of the most despised drudgeries in the study of the English language — diagramming sentences — decades ago.

It is no longer a rarity when an opinion article shows up in my inbox flawlessly stating the writer’s intent. I strongly suspect AI use. Every app I open wants to help improve my writing. I swat them off like blue gnats in October as I try to peer through their haze to comprehend the absolute revolution that is upon us.

It is a revolution on every front that will be fraught with danger and disruption to society. But to wring our hands over whether a writer employed AI to proofread, make suggestions, research facts, or even help to structure an argument on an opinion page (all of which I have done) is to be concerned with a drop of rain just before the dam is about to break, or, to make the analogy more apt, several dams.

Right now, our country is closing in on an election at a time when AI can replicate unlimited human personas, “agents” that can work ceaselessly at creating and spewing disinformation designed to enrage one part of the electorate against another so fast it could take decades to untangle what’s left of a society rendered unable to trust itself.

Or, it could provide nearly miraculous answers to intractable problems, answers unimaginable by its human creators because we lack the power to understand everything all at once.

Just last month, OpenAI claimed its experimental system of 10,000 AI agents had collaborated to solve a nearly 100-year-old problem that is one of seven for which a million-dollar reward has been offered since the year 2000. They “solved” it in 88 hours, generating counter claims, hard feelings (two human mathematicians were close and suspected their work might have been used), and questions about the future nature of mathematics and even science, itself.

Scott Hunter

Publisher & Editor