The latest human to fall under the spell of Artificial Intelligence is Thomas L. Friedman of the New York Times. "With GPT-4 and other A. I., the future is right in front of our eyes," says the summarizing pull-head in today's column by Friedman.
Or did he himself write the column? If so, why? He had access to technology that could have written it for him faster.
I'm going to skip over the more grandiose claims (it’s going to change everything about everything we do) made for the ChatBot Friedman was introduced to by someone named Craig, who is connected to Friedman’s wife’s museum, which is called Planet World. I am going to focus on this part of Friedman’s column:
"Craig asked GPT-4 -- for which Craig was a selected advanced tester and which was just released to the public -- to summarize Planet World and its mission in 400 words.
"It did so perfectly -- in a few seconds."
Then, Friedman tells us, GPT-4 did it again in 200 words. And then in Arabic, then in Mandarin and "Then in English again -- but in the form of a Shakespearean sonnet."
Friedman does not share that sonnet with us. Don't you suspect that sonnet would read like a Shakespearean sonnet written by a clothing store owner's nephew who was given this assignment because it was the only aspect of running a clothing store — dubious promotion — that he could get the hang of? (Granted, the nephew would have taken longer to work it up, but how much longer? A day? Two days? What's time to a nephew? It was never going to be used.)
Here is the GPT-4 product that Friedman does share with us, in part: the same description of his wife's museum "in an abecedarian verse -- where the first line begins with the letter A, the second with B and so on through the alphabet. It did it with stunning creativity," says Friedman, and he vouchsafes us the first six lines:
Alluring in Washington, is a museum so grand,
Built to teach, inspire, and help us understand.
Curious minds Planet flock to World's embrace,
Delving into language and its intricate grace
Every exhibit here has a story to tell,
From the origins of speech to the art of the quill.
I'm going to be harsh, here: I don't find that stunningly creative. A party trick, maybe, to tepid applause. Nothing that most museums would find much promotional use for, nor that a fourth-rate literary magazine (if there are such) would be even halfway tempted to publish.
It's verse, I guess, but metrically clunky -- and come on, "Alluring in Washington[,]"? "A museum so grand"? What does "Curious minds Planet flock to World's embrace" even mean? That a bot can churn out this stuff may be surprising, but not surprising in the way that good verse surprises.
My sympathies are with the bot. All sorts of language-stuff programmed into it from all sides and it's just getting its feet on the ground, and here it's hauled out by a sudden request: produce verse. And of course it can, after a fashion, produce awkward, forced, slapdash doggerel verse. Who can't? But is this verse good for anything? Can we imagine that the bot that produced this verse is proud?
That a bot can do this is remarkable, no doubt. But . . . how remarkable? As remarkable as the directions we get routinely on our car nowadays?
No.
So. What do we have here? More quasi-responsive lumps of uninspired, soulless words, framed by a peremptory request, not by . . .
Not by love, not by hate, not by desire, not by a need to grasp the vibrations of something halfway seen.
Okay, nobody expects a bot to make art. But I don't see this bot even making catchy advertisements, except maybe for itself and Thomas Friedman's wife's museum.
I'm absolutely certain GPT-4 would write more intelligibly and intelligently than Thomas Friedman can on any topic one might choose.
Doesn’t one wonder who the programmers might be: Soulless, humorless, automatons. The same people speech writing for present GOP candidates. Zzzzzz!