I looked at 24 words that were identified as distinctively LLMish (interestingly, almost all positive) and checked their presence in full text of papers - four showed very strong increases, six medium, and two relatively weak but still noticeable. Looking at the number of these published each year let us estimate the size of the "excess" in 2023. Very simple & straightforward, but striking results.
I have a preprint out estimating how many scholarly papers are written using chatGPT etc? I estimate upwards of 60k articles (>1% of global output) published in 2023. https://arxiv.org/abs/2403.16887
How can we identify this? Simple: there are certain words that LLMs love, and they suddenly start showing up *a lot* last year. Twice as many papers call something "intricate", big rises for "commendable" and "meticulous".
Second favourite bit is the point where our protagonist has to improvise a political speech without knowing precisely a) who he is meant to be or b) whether the meeting was for the Liberals or the Conservatives https://youtu.be/aalwXQINtio
NEWS:
Due to problems with our site upgrade, we are dropping the paywall, allowing full access to all our material.
https://www.british-history.ac.uk/
Additionally, due to problems with search, we are temporarily making available the old version of our site.
Several of us overly online biologists spent years quietly doing an experiment on Twitter, trying to find out if tweeting about new studies from a set of mid-range journals caused an increase in later citations, compared to set of untweeted control articles.
Turns out we had no noticeable effect; the tweeted papers were cited at the same rate as the control set.
Our paper, headed by Trevor Branch, was published today in PLOS One:
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0292201
Great story about how YouTube helped with moving away from IE6.
"Our most renegade web developer, an otherwise soft-spoken Croatian guy, insisted on checking in the code under his name, as a badge of personal honor, and the rest of us leveraged our OldTuber status to approve the code review."
I swear that wasn't me. Although I would have loved to do it.
This is a really interesting (if obviously very bleak) piece on state compensation for deaths, and I realised while reading it I have absolutely no idea what to think about a lot of this. https://www.theguardian.com/uk-news/2024/mar/09/london-bridge-terror-attack-why-did-victims-families-get-wildly-different-payouts
"The most important lesson to figure out is why it is taking so long to restore services. That will tell us how to prevent such a calamity in other vital national institutions."
On the matter of the British Library cyber incident https://ciaranmartin.substack.com/p/on-the-matter-of-the-british-library
Enjoyed this: the world of amateur photography in the late USSR. https://kosmofoto.com/2021/03/amateur-photography-in-the-ussr-part-one/ & https://kosmofoto.com/2021/04/amateur-photography-in-the-ussr-part-two/
(I think my favourite detail is that the vast majority of flash units were wall powered...)
"If we can't pay a fair Pension, then how can we afford foreign adventurism, nuclear weapons, armies, royals, pageantry, MP's salaries, etc?"
Total defence budget: £45bn. So that's just £245bn we need to find by slashing, er, MPs salaries and the royals. Feel like we might struggle to balance the books on this one.
Would love to know how many people who signed this petition also believe the nation is spending far too much and should cut taxes.
State pension spend: currently £125bn. Under this model: £415bn (up £290bn). Entire public spending budget, for everything: £1,190bn.
So implementing this would mean approximately a 25% raise in taxation. Can't see that one being very popular... https://petition.parliament.uk/petitions/654389
I really need to find something better than a flatbed for this, but it's definitely working.
remember, to mitigate the effects of security breaches, change your Olympic host city on a regular basis https://www.theguardian.com/world/2024/feb/27/bag-containing-security-plans-for-paris-olympics-stolen-from-french-train
(in my notes this is "Dud shot at f/16 1/200", I had assumed it was underexposed and written it off. Phone metering remains a bit challenging.)
Second 120 film today - spooling a little fiddlier but came out OK. Quality looks excellent for a 70-year-old camera & lens! Going to be a pain to scan, though...
Librarian and occasional researcher. Opinions of course my own. Scholarly communications, historic MPs, Wikipedia, inter alia other things. Misplaced Scot.