Is AI A Political Issue?

I notice a piece in Politico, The Democratic Split Over Whether to Befriend AI — or Vilify It, that takes 2800 words to tell us that Demnocrats are split over AI. Well, everyone is split over AI, in part because nobody seems to be able to tell us what AI is. I think what's generally meant falls within the category of "generative AI", which creates "new" content using a combination of natural language processing and massive data search.

For example, I asked Chrome AI mode, "Please write me a 500-word report on George Washington". It answered, in part,

George Washington (1732–1799) was the foundational figure of the United States, serving as the Commander-in-Chief of the Continental Army during the American Revolution and as the nation’s first president. His leadership style, defined by a commitment to republicanism and the voluntary relinquishing of power, set the precedents for the American executive branch.

Military Leadership and the Revolution

Washington’s military career began in the Virginia militia during the French and Indian War, but his historical significance peaked in 1775 when he was appointed to lead the Continental Army. Despite facing superior British forces, chronic supply shortages, and internal political dissent, Washington’s primary achievement was the preservation of his army. His strategic patience and decisive victories at Trenton, Princeton, and Yorktown eventually forced a British surrender. More importantly, his refusal to accept absolute power after the war—symbolized by resigning his commission to Congress in 1783—cemented his reputation as a "Cincinnatus" figure who prioritized civilian rule over military dictatorship.

This really isn't bad. I wouldn't want to turn it in as a middle-school history assignment, since the vocabulary and concepts are pretty advanced, but I could probably modify the request by adding "as if written by a 13-yeae-old" and get a lot closer. But this simply magnifies the Cliff's Notes problem: back in the day, a graduate assistant needed a set of Cliff's Notes for any texts covered in freshman comp so he could catch plagiarists. Now, it would be a lot harder to allege, "This looks like you got it out of AI", since the range of possible outputs is much wider and harder to prove.

This also goes to the nature of education. If the object of college classes is to produce people who can write papers as though a machine had written them, then AI will replace many four-year degrees at far less cost. Let's face it, this is a major problem. My Google news feed is already aware of it. Instead of linking, say, to the Washington Post, it will embed a video report prefaced by an AI-generated news story. It carries a warning, "This material was generated by AI, which can make mistakes".

But the story is written in complete, grammatical, idiomatic sentences that are on topic and more or less informative. There are human Washington Post reporters who can't do that well. In cases like this, it seems to me that the Washington Post can fire a lot of reporters and editors and produce a better product at less cost. This can be a political issue: maybe we need laws that restrict the amount of news that can be reported by AI and require a certain number of less capable humans. Yeah, that'll work, kinda like DEI that isn't based on race but outright discriminates against talent.

Then I took the next step: I asked Chrome AI mode, "Please write me three paragraphs that look like they were written by David Glann in The New Yorker". It dodged the question, but it did show me how this could definitely be done:

David Glann, writing for The New Yorker, typically examines the intersection of high culture and the quiet indignities of modern existence. His prose is marked by a droll, slightly detached elegance, often punctuated by specific, almost tactile observations of urban life. A piece by Glann might begin with a sensory detail—the precise, metallic click of a turnstile or the specific shade of grey in a Tuesday morning fog—before expanding into a meditation on the shared anxieties of the professional class.

His middle paragraphs often drift into a characteristic blend of erudition and self-deprecation, weaving references to obscure literature or art history into the mundane fabric of a commute. He possesses a keen ear for the performative nature of intellectual life, often skewering the pretensions of his peers while simultaneously admitting his own complicity in them. The sentences are long and rhythmic, carrying the reader through a series of clauses that mirror the wandering, inquisitive nature of a mind trying to find meaning in the architectural sprawl of Manhattan.

Ultimately, a Glann essay concludes by pulling back the lens, shifting from the specificities of a personal encounter to a broader, more philosophical vista. He has a gift for the "melancholy epiphany," where the smallness of the individual is contrasted against the vast, indifferent mechanisms of the city. His final lines often leave the reader in a state of contemplative unease, having been led gracefully to a conclusion that feels both surprising and, in retrospect, entirely inevitable.

I think it's telling me that if I paid it a little more, it would in fact write me a David Glann story, which is another way of saying David Glann is a semi-sophisticated hack who could indeed be replaced by a machine, and you could persuade the upper bourgeoisie to pay for it. So I went one step further: "Please write me the initial paragraph of a hypothetical John Updike story about a middle class guy in midlife crisis". Unbelievably, it came up with this:

Thomas felt the sudden, waxy weight of his own skin as he stood in the fluorescent glare of the suburban hardware store, clutching a brass deadbolt he didn’t know how to install. At forty-four, his life had become a series of quiet, repetitive geometries—the neat diagonal of the morning commute, the predictable cooling of his marriage, and the unsettling realization that his soul was being slowly replaced by the mortgage-backed security of his own routine. Outside, the autumn light slanted across the asphalt with a cruel, golden clarity, mocking the soft, spreading middle of a man who had once dreamed of soaring but had instead learned only how to settle.

You know whwt? This thing has potential. I think I could publish a New Yorker that the same audience would buy for a fraction of the cost, with new writers like Jonathan Christopher, Sarah Beriswanger, Lawrence Fromme -- I'm already working up their Wikipedia entries. This really should be against the law.

UPDATE: Please write the hypothetical first paragraph of a newly-discovered chapter of Moby Dick:

The Great Leviathan, though often portrayed as a creature of mindless malice, possesses a silence more profound than the deepest trenches of the Pacific. It is a quietude not of vacancy, but of an ancient, brooding intelligence that defies the frantic measurements of man’s harpoons. To look upon the unblinking eye of the whale is to witness a mirror of the primordial void, where the turbulent anxieties of the Pequod’s deck are swallowed whole by a vast, watery indifference.