I'm still fascinated by the world encapsulated in this meme:
Instagram has rolled out an integrated A.I. in its comments system: Instead of leaving your own weird note on a stranger’s selfie, you allow Meta A.I. to render your thoughts in its own language. This can be “funny,” “supportive,” “casual,” “absurd” or “emoji.” In “absurd” mode, instead of saying “Looking good,” I could write “Looking so sharp I just cut myself on your vibe.” Essentially every major email client now offers a similar service. Your rambling message can be instantly translated into fluent A.I.-ese.
I can't help but think that the world where you can click a button to generate a detailed email reply or leave a nice comment is a world where such things don't carry meaning any more.
What is the difference between a "like" or a "thumbs up" emoji and pressing a "generate a positive comment for me" button?
Every sentence is gesturing toward some deep significance, but only in the same way that a description of people tickling one another gestures toward humor.
Emojis amd "likes" are "gestures toward meaning. I would expect the generated comments to just collapse back into a simple "like" because who wants to waste the time reading a comment nobody wrote?
But I argued in another thread (#1342921) that things don't have value because of the work put in to them...yet comments may actually be something where the work put in to them is some of the point.
Having someone spend time writing a reply to you feels good, even if the comment doesn't necessarily contain a brilliant insight. It is pleasant when someone takes the time to engage with you.
For some reason a paragraph that more or less says nothing new but that was written by a human means more to me than a paragraph generates by a button -- even if the generated comment says more or less the same thing. What's the reason, though?
Also: we definitely are underestimating how LLMs are going to change our language
In the British Parliament, for instance, transcripts show that M.P.s have suddenly started opening their speeches with the phrase “I rise to speak.” On a single day this June, it happened 26 times. “I rise to speak in support of the amendment.” “I rise to speak against Clause 10.” Which would be fine, if not for the fact that this is not something British parliamentarians said very much previously. Among American lawmakers, however, beginning a speech this way is standard practice. A.I.s are not always so sensitive to these cultural differences.
In 2024, the investor Paul Graham made that mistake when he posted online about receiving a cold pitch. He wasn’t opposed at first. “Then,” he wrote on X, “I noticed it used the word ‘delve.’” This was met with an instant backlash. Just like the people who hang their identity on liking the em dash, the “delve” enjoyers were furious. But a lot of them had one thing in common: They were from Nigeria.
According to the data, post-ChatGPT papers lean more on words like “underscore,” “highlight” and “showcase” than pre-ChatGPT papers do. There have been multiple studies like this, and they’ve found that A.I.s like gesturing at complexity (“intricate” and “tapestry” have surged since 2022), as well as precision and speed: “swift,” “meticulous,” “adept.” But “delve” — in particular the conjugation “delves” — is an extreme case. In 2022, the word appeared in roughly one in every 10,000 abstracts collected in PubMed. By 2024, usage had shot up by 2,700 percent.
This year, I read an article in which a writer complained about A.I. tools cheapening the craft. But I could barely pay attention, because I kept encountering sentences that felt as if they’d been written by A.I. It’s becoming an increasingly wretched life. You can experience it too.
slop4slop?