@BretDevereaux@romanhelmetguy We should found a club of People Wishing Polybius Had Expressed Himself a Little More Clearly.
(My long-standing peeve is προεκθέσεις in 11.1a)
@Athens_Stranger Studying things doesn’t mean ‘tearing them down’.
The solar system is no less great than the Iliad, but you don’t expect astronomers to spend their time waxing lyrical about how great it all is, or accuse them of ‘tearing it down’ if they simply get on with studying it.
@AntiqueThought Sorry for the belated reply. I log into Twitter every once in a blue moon.
The talk will be Zoomed: DM me if you’d like the link. But it won’t contain much that you don’t already know.
@AntiqueThought Coda on running heads: Philodemus P.Herc. 558 and, especially, P.Berol. 9780 recto (Didymus) and verso (Hierocles).
Full disclosure: I’ve already published much of what I’m going to say about Didymus in Fausto Montana’s 2024 volume.
@DavidAstinWalsh Depends on the field. In my field it’s the standard; it’s the opposite that’s unusual (risks Qs concentrating on some papers and not others).
This is undoubtedly true. This captures a reality that many people haven't yet come to terms with: artificial intelligence is not merely an assistant for writing—it may become a default writer. And the implications of this are far more profound than most assume. People who haven’t seriously experimented with AI tools don't realize just how powerful and seamless they already are. They still imagine the clunkiness and hilarious shibboleths of older iterations of ChatGPT rather than a system capable of managing nuance and register, producing emails, marketing copy, technical documentation, code that is indistinguishable—sometimes superior—to what most humans can produce, essays and even moving artistic prose.
This gap in understanding is one of the most dangerous aspects of the current AI debate. Much of the public discourse about AI in writing and education remains caught up in surface-level concerns: plagiarism, productivity, efficiency. But what is really at stake is cognitive atrophy. Writing is not just a way of communicating ideas—it is a tool for forming them. When we write, we clarify, confront contradictions, wrestle with uncertainty. We slow down and think. The act of drafting, revising, structuring—these are exercises in reflection, logic, imagination, and self-discipline. When AI takes over that process, those mental muscles atrophy.
What we’re likely to see is a split, not unlike the one we've seen in other domains touched by automation. Just as very few people today understand how their smartphones work, or how to fix a car, or how to do long division by hand, very few people will understand how to craft a persuasive argument, or a nuanced narrative, or a logical sequence of ideas. Writing will become a specialized skill, like glassblowing or Latin verse composition—taught for cultural reasons, practiced by a few artisans, but functionally irrelevant for the average person. And this is not just about losing a skill. It’s about losing a way of being.
Even in education, where writing is likely to remain a requirement for some time, students may come to see it as an alien ritual like doing long division, something artificially imposed on them in contrast to how things are really done in the world. After all, if every adult they know uses AI to write, why should they view writing as a personal or valuable act? The result is that the skill and the habit will gradually disappear. And with it, the kind of independent, critical, recursive thinking that writing fosters.
This worry is well-founded. More people should be taking it seriously. The issue isn't just whether AI is “cheating” or whether it can help us write faster—it’s whether we are prepared to give up the kind of thinking that only writing teaches us to do.
Every sentence before this one was actually written by an AI. I'm not kidding. This paragraph is the only thing I'm typing myself. Think about that for a moment.
Non si può pretendere che un editorialista verifichi tutto, anche quel che proviene da una fonte che si potrebbe ritenere attendibile.
Ma un quotidiano di un certo livello un po' di fact-checking lo potrebbe anche fare.