๐ฏ
@seismatters
Geoscientist๐จ๐พโ๐ฌ | Visiting Professor at Imperial College | Ex-Academic | ๐๐พโโ๏ธ๐ด๐พโโ๏ธ ๐ง๐พโโ๏ธ๐ธ๐บ| he/him | ๐ณ๏ธโ๐๐ณ๏ธโโง๏ธ| I do all my own stunts | โSelf-declared King of Geotwitterโ (Palin, 2022) | Professional (East)Midlander ๐| Co-founder of @EarthArxiv
๐ฏ
I honestly get a hollow feeling when I think about that particular time in my career โน๏ธ
There was always something depressing about the picket line, when people would cross it and say: "how can you strikeโฝ We get so much freedom to do what we wantโฆ think of the work piling upโฆ but I really have to do X before the deadlineโฆ"
Meanwhile terms & conditions, including take home ยฃ flatlined
Iโm seeing this more in scientific articles by non-native English authors. If AI clarifies meaning, great. If it rewrites meaning, thatโs dangerous. Weโre going to see both cases a lot more in the near future.
I think this is where Iโm currently at with it
One of my recent grad students, for whom English was a second language (maybe even a third), used it in their thesis, with a very detailed supplement explaining how. Basically they wrote the first draft, then asked gpt to improve the flow and grammar but not to change meanings, add citations etc 1/2
Amazed it didnโt have baked beans on it ๐คฎ
What you choose to put inside your body is your choiceโฆ
N.B. that recipe looks aceโฆbut isnโt it simply a fancy potato scone of sorts? ๐ค
Cancel Italy then ๐คท๐พโโ๏ธ
P.S. anchovies and capers on pizza?
Poutine on pizza ๐ค
What the fuck Izzy?!?!
I love @ijayas.bsky.socialโฆbut someone, please cancel her ๐ค
If youโre going to use generative AI, this is the way to do it and to acknowledge doing it.
(I would personally not choose to do it, but there is a world of difference between โI wrote a paper and asked genAI to clean it upโ and โI asked genAI to write my paper.โ)
Reread it? ๐คท๐พโโ๏ธ
Why? Because one has to publish in English even when English may not be the main language of the author, and an llm can successfully polish out minor grammar/syntax errors without changing the meaning of the original text.
Not everywhere. Itโs up to the teachers and head teachers to change it back to be about books, reading and the joy a good story brings. #WBDisaboutreading
The story of the origin of World Book Day was on Radio 4 the other day. So charming and heart-warmingโฆbut now we have this existential nightmare once a year
Declaration of Generative AI and AI-assisted technologies in the writing process Statement: during the preparation of this work the author used ChatGPT in order to copy edit the text to improve quality and readability. After using this tool, the author reviewed and edited the content as needed and takes full responsibility for the content of the publication.
Just spotted this in a paper. First time I've seen something like this.
In a future world I'll spell calendar correctly...
I think about this Tony Benn speech much more than I used to
"Foreign-born Citizens" are actually just called citizens.
"Foreign-born voters" are just called voters
"non-citizen voters" (yes, this is a thing here in ๐ฌ๐ง) are just voters.
Good.
Itโs just your mention of citations made it sound like that was the key metricโฆ
Good point. But one may add, sadly :
iii. anything like MS word is AI software now (see autocomplete) and virtually impossible to avoid completely ; iv. the notion of (dis)proving or quantifying use of AI software seems fantasy moving forward ; v. clicking terms and conditions is pretty much mand..
Putting aside the citations, how do you (and your coauthors, if you have them) feel about the paper? Are you proud of it? Surely thatโs what really mattersโฆ? ๐ค
Tbh, that sounds like THE DREAM!
๐ Our science communication awards are now open!
These awards recognize exceptional science journalists, communicators, and researchers whose work in 2025 made #STEM accessible and engaging for broad audiences.
Deadline: April 3, 2026
Apply now: www.nationalacademies.org/awards/excel... #scicomm
Nice. Lovely stuff
See also: "This is also a work (seemingly) part of a PhD project and, perhaps for that reason, does try and address too many topics..."
"I wonder if [the paper] was compiled with the help of AI software..."
Please don't write this in your paper review, unless: i) you have proof; and ii) it is against the journal policy to use it.
For clarity, this isn't a paper I've written, but the comment from a co-reviewer of a GEOLOGY paper.
Letโs see what the future holds ๐๐พ