"Slop"
« previous post | next post »
It's WOTY season, and The Economist's choice for 2025 is slop:
PICKING A WORD of the year is not easy. In the past the American Dialect Society has gone with “tender-age shelters” (2018) and “-ussy” (2022). The Oxford English Dictionary (oed) has caused conniptions by opting for things like “youthquake” (2017) and “goblin mode” (2022). If you cannot remember why those terms were big that year, that is the point: the exercise is not a straightforward one.
Sometimes a single suitable word is not at hand, so a phrase is chosen instead; other times the word simply seems jarring. Middle-aged lexicographers are often tempted to crown a bit of youth slang, but such terms are transient and sound out of date before the press release is published.
The Economist’s choice for 2025 is a single word. It is representative, if not of the whole year, at least of much of the feeling of living in it. It is not a new word, but it is being used in a new way. You may not like it, but you are living with it. And it is probably here to stay.
See the rest of the article for discussion of neijuan, TACO, 6-7, brain rot, and so on…
Readers should prepare themselves: they will probably experience brain rot more often, thanks to our word of the year. Our pick’s rise was spurred by OpenAI’s release of Sora, a generative artificial-intelligence (ai) platform that can create videos based on a prompt. Suddenly social-media feeds were filled with such clips. A term that started circulating in the early years of generative AI is now everywhere: “slop”.
Update — for the latest from the Annals of Slop, see Jonathan Oppenheim, "We are in the era of Science Slop (and it's exciting)", 12/5/2025:
The rate of progress is astounding. About a year ago, AI couldn’t count how many R’s in strawberry, and now it’s contributing incorrect ideas to published physics papers. It is actually incredibly exciting, to see the pace of development. But for now the uptick in the volume of papers is noticeable, and getting louder, and we’re going to be wading through a lot of slop in the near term. Papers that pass peer review because they look technically correct. Results that look impressive because the formalism is sophisticated. The signal-to-noise ratio in science is going to get a lot worse before it gets better.
The history of the internet is worth remembering : we were promised wisdom and universal access to knowledge, and we got some of that, but we also got conspiracy theories and misinformation at unprecedented scale.
AI will surely do exactly this to science. It will accelerate the best researchers but also amplify the worst tendencies. It will generate insight and bullshit in roughly equal measure.
Welcome to the era of science slop!
Update #2 — MW's 2025 WOTY is also "slop"…
Victor Mair said,
December 13, 2025 @ 12:19 pm
Nèijuǎn 内卷 ("involution") has been much discussed on Language Log in recent years. For one of the latest installments, with extensive references, see "Death from overwork" (10/5/25)
JPL said,
December 13, 2025 @ 5:41 pm
'slop', as in "AI slop" and science slop": Not a new word (lexeme), nor a new sense of an existing word (lexeme) (i.e., it does not result in a significant differentiation of existing subcategories of sense), but rather a novel phenomenon ("AI slop"), to the description of which an existing word (lexeme) has been applied. "Science slop" has been around for quite a while. Outwardly prestigious science journals have for some years included articles whose research basis could be described charitably as "sloppy". So now there'll be even more of it. So 'slop' is not a new word (lexeme, or sense of lexeme) this year; AI slop is a new phenomenon this year, that everyone is talking about. However, it could yet become a technical term if explicit criteria are developed.
Jonathan Smith said,
December 13, 2025 @ 7:22 pm
"the signal-to-noise ratio in science is going to get a lot worse" / AI "will generate insight and bullshit in roughly equal measure"
so which is it? gonna say the former…
Philip Taylor said,
December 14, 2025 @ 6:22 am
The two statements are not incompatible, Jonathan. If (say) the signal-to-noise ratio in (published) science were 100:1 today, and it were to increase to 50:50 (as per the second statement), then "the signal-to-noise ratio in science is [indeed] going to get a lot worse".
ajay said,
December 17, 2025 @ 8:14 am
"Science slop" has been around for quite a while. Outwardly prestigious science journals have for some years included articles whose research basis could be described charitably as "sloppy".
Sloppy just means careless or bad, in this sense. The image is that someone is doing some task without enough attention and the paint (or whatever) is getting everywhere – it's related to the verb, meaning to splash about or spill.
Slop in the sense of AI slop is from the noun, meaning kitchen refuse used as animal food, and more generally low-quality food; "what kind of slop are they feeding us?" you might ask. Slop is something you give to people you don't respect very much for them to consume.
Obviously slop tends to be semi-liquid and therefore sloppy, so there's quite a bit of overlap. But I think something can be sloppy science without being slop. Things like the discovery of penicillin were due to sloppy lab practice.
There is a completely etymologically separate meaning that relates to clothes – slops were the clothing issued by the Royal Navy to its sailors, and stored in the slop chest.