Paolo Crosetto
@paolocrosetto.bsky.social
Experimental & Behavioural economist INRAE Grenoble • President of the French Association of Experimental Economists • Scientific publishing measurement & reform • Experiments on food labeling - risk - choices • Rstats • Italian Food Police honorary member .. more
Experimental & Behavioural economist INRAE Grenoble • President of the French Association of Experimental Economists • Scientific publishing measurement & reform • Experiments on food labeling - risk - choices • Rstats • Italian Food Police honorary member
Business
25%
Economics
25%
Pinned
Since AI slop is again all over Scientific Reports, a thread on the economics of grey-zone publishing.
Why does slop keep getting published? What does it mean for science? How can we stop this?
Background readings:
Understand the strain: tinyurl.com/2b6wxx5r
Stop the drain: tinyurl.com/3jfscscy
Why does slop keep getting published? What does it mean for science? How can we stop this?
Background readings:
Understand the strain: tinyurl.com/2b6wxx5r
Stop the drain: tinyurl.com/3jfscscy
bsky.app
@hansonmark.bsky.social has a very similar idea to this and floated it around in a thread here some time ago.
It is also one of the potential radical solutions we urge funders to take.
We won't solve all of sicentific publishing in the coming years, but paying less for the same mess seems doable
It is also one of the potential radical solutions we urge funders to take.
We won't solve all of sicentific publishing in the coming years, but paying less for the same mess seems doable
Reposted by Paolo Crosetto
This is potentially significant. The CNRS are one of the best in the world. Reliance on metrics controlled by private companies is a problem for science. We need the powerful organisatios to take the lead.
Understand the strain: tinyurl.com/2b6wxx5r
Stop the drain: tinyurl.com/3jfscscy
Understand the strain: tinyurl.com/2b6wxx5r
Stop the drain: tinyurl.com/3jfscscy
thanks for this
Not that I know of, but good idea.
With @hansonmark.bsky.social @pagomba.bsky.social & @danbrockington.bsky.social we briefly looking at country-specific publication incentives & how they translate into % of MDPI in that country. Ancdotical evidence checks out, but we did not go further (so far).
With @hansonmark.bsky.social @pagomba.bsky.social & @danbrockington.bsky.social we briefly looking at country-specific publication incentives & how they translate into % of MDPI in that country. Ancdotical evidence checks out, but we did not go further (so far).
This is the sort of action we urge funders and scientific governing bodies to take: change incentives, shift financial flows away from big publishers and fund open alternatives instead.
Well done CNRS, let us keep up action to stop the drain of scientific publishing!
Well done CNRS, let us keep up action to stop the drain of scientific publishing!
Reposted by Dan Brockington, Paolo Crosetto
Great thread by one of the folks who brought you “The strain on scientific publishing”
direct.mit.edu/qss/article/...
direct.mit.edu/qss/article/...
Elenchi di riviste classificate
www.anvur.it
www.anvur.it/it/ricerca/r...
I'm not saying it's a bad idea, but I think the devil is in the detail and any shortlist / whitelist approach will be met with the sort of scheming and maneuvering that it met in Italy. People will complain X, Y were excluded and lobby till they're in.
Cycle repeats.
I'm not saying it's a bad idea, but I think the devil is in the detail and any shortlist / whitelist approach will be met with the sort of scheming and maneuvering that it met in Italy. People will complain X, Y were excluded and lobby till they're in.
Cycle repeats.
2. colleagues looking for promotion / validation targeted the weakest links in the list massively - which is a totally rational thing to do. Some MDPI journals had made the list, and the result is that Italy has the highest number of MDPI papers per researcher per year among big countries.
Tricky.
Tricky.
This was tried in Italy, with the "classe A" journals, that were considered "equally excellent".
What happened was:
1. lobbying from different disciplines / people led to inclusion in the whitelist of journals that served the interest of proposers; [continues]
What happened was:
1. lobbying from different disciplines / people led to inclusion in the whitelist of journals that served the interest of proposers; [continues]
Reposted by Paolo Crosetto
I know of few people who get the economics of scientific publishing better than @paolocrosetto.bsky.social This thread on the incentives that are driving bad publishing practice is golden.
For the darkly funny side of this see:
the-strain-on-scientific-publishing.github.io/website/post...
For the darkly funny side of this see:
the-strain-on-scientific-publishing.github.io/website/post...
Tell your boss to tell her boss to tell their funders to stop playing. Stop the drain.
AI slop is just the tip of the iceberg. The system is set up to inch towards the grey zone, till all is grey, & people will ditch it wholesale.
But we *need* publishing. It could be so much better.
AI slop is just the tip of the iceberg. The system is set up to inch towards the grey zone, till all is grey, & people will ditch it wholesale.
But we *need* publishing. It could be so much better.
Second, this generates a huge & increasing drain. A drain of money, but also of time, control, and trust.
It's like a bubble -- and if it bursts, the trust in the publication system will be gone.
Everybody watering down their wine on their side, till no-one trusts no wine vendor, ever, anywhere.
It's like a bubble -- and if it bursts, the trust in the publication system will be gone.
Everybody watering down their wine on their side, till no-one trusts no wine vendor, ever, anywhere.
What is the price of scientific publishing grey-zoning? and who pays?
First, this is costly. *Very* costly. And it is mostly public money. No wonders some big funders, as the Gates Foundation, are calling it quits and refusing to pay.
First, this is costly. *Very* costly. And it is mostly public money. No wonders some big funders, as the Gates Foundation, are calling it quits and refusing to pay.
So this is why we keep seeing AI slop in Gold OA journals.
It's not a bug: it's a feature.
As long as the incentives stay the same, authors pay & few people read, publishers will find ways to allow the grey zone in, while minimizing reputational impact.
It's the game. It's very profitable.
It's not a bug: it's a feature.
As long as the incentives stay the same, authors pay & few people read, publishers will find ways to allow the grey zone in, while minimizing reputational impact.
It's the game. It's very profitable.
It goes deeper -- and weirder.
Sorbillo is a high-reputation pizzeria in Naples, Italy. Go down Via dei Tribunali, you'll find *several* Sorbillo copycats.
Go down your interent aisle, you'll find *several* Springer-Nature branded MDPI copycats.
*Same* name, cheaper APC.
tinyurl.com/3vn6ab5c
Sorbillo is a high-reputation pizzeria in Naples, Italy. Go down Via dei Tribunali, you'll find *several* Sorbillo copycats.
Go down your interent aisle, you'll find *several* Springer-Nature branded MDPI copycats.
*Same* name, cheaper APC.
tinyurl.com/3vn6ab5c
But it's just whack-a-mole.
Incentives to grey-zone your publisher affect the whole industry. MDPI is down, but grey-zone authors & their demand for CV lines haven't gone anywhere.
Picking up the flag where MDPI dropped it, it's just business.
Submissions to other OA mega-journals skyrocketed.
Incentives to grey-zone your publisher affect the whole industry. MDPI is down, but grey-zone authors & their demand for CV lines haven't gone anywhere.
Picking up the flag where MDPI dropped it, it's just business.
Submissions to other OA mega-journals skyrocketed.
Sooner or later the reputation catches up with you. And it spells doom.
At some point in 2024, common knowledge of bad reputation started to hit MDPI.
IJERPH -- then the world's biggest journal -- got delisted by Clarivate. Submissions plummeted there and across all MDPI portfolio.
At some point in 2024, common knowledge of bad reputation started to hit MDPI.
IJERPH -- then the world's biggest journal -- got delisted by Clarivate. Submissions plummeted there and across all MDPI portfolio.
Brand extension: have a very reputable journal? Just conjure out of thin air new journals inheriting its name. Wine watered down a notch, everybody buying it.
Nature >> Nature XY
*Not* in the grey zone, but the strategy is clear: money = reputation + many papers.
There are 72 Nature XY journals.
Nature >> Nature XY
*Not* in the grey zone, but the strategy is clear: money = reputation + many papers.
There are 72 Nature XY journals.
Impact inflation: the industry's main indicators for journal quality, citation counts, are easily gamed.
Easy then to pump them up (self citations, cartels) to attract authors & provide cover for grey-zone ones.
Inflation of citation indicators is up across the whole industry -- more at Gold OA.
Easy then to pump them up (self citations, cartels) to attract authors & provide cover for grey-zone ones.
Inflation of citation indicators is up across the whole industry -- more at Gold OA.
Special Issues: they attract authors, giving them editorial power. Also it's easy for publishers to shrug off bad papers using the good SIs and heterogeneity as cover -- SIs are in when it's useful, out when it's not.
They are so widespread at Gold OA publishers that special is the new normal.
They are so widespread at Gold OA publishers that special is the new normal.
Mega-Journals: build reputation once, exploit it N times. And shrug off the grey-zone papers as "some bad apples".
They used to be rare, but are the new normal.
2016-22, Number of journals by size:
1 paper/week: + 34%
1 paper/day: +99%
>10 papers/day: +1400%
Author-pays means mega-journals.
They used to be rare, but are the new normal.
2016-22, Number of journals by size:
1 paper/week: + 34%
1 paper/day: +99%
>10 papers/day: +1400%
Author-pays means mega-journals.
So *any* strategy that allows to keep reputation decent while accepting grey-zone authors is a win.
Water down your wine, as long as patrons don't notice. Or at least not everybody knows that everybody knows.
We see *a lot* of this since the switch to author-pay Gold OA.
Let's see some examples.
Water down your wine, as long as patrons don't notice. Or at least not everybody knows that everybody knows.
We see *a lot* of this since the switch to author-pay Gold OA.
Let's see some examples.
and remember: the reputation isn't gone until it's *common knowledge* that it is gone.
As long as people think there might be others thinking X is a good enough journal, or think they have plausible deniability of not knowing better, submissions will flock in.
Grey-zone authors just need cover.
As long as people think there might be others thinking X is a good enough journal, or think they have plausible deniability of not knowing better, submissions will flock in.
Grey-zone authors just need cover.
Why would publishers enter the grey zone? It depends on their business model.
If readers pay, incentives are low. Readers see crap, stop paying. Bad.
If authors pay, publishing becomes a service for your customers - authors. Getting them on board is priority #1. Not losing reputation is #2.
If readers pay, incentives are low. Readers see crap, stop paying. Bad.
If authors pay, publishing becomes a service for your customers - authors. Getting them on board is priority #1. Not losing reputation is #2.
Now, why would publishers ever do that for you? Obviously, for your money.
But there's a catch. Nobody wants certificates that everybody knows are awarded only for money.
Outright predatory publishers that fill your inboxes are not a threat: they have no customers.
Bad authors want the grey zone.
But there's a catch. Nobody wants certificates that everybody knows are awarded only for money.
Outright predatory publishers that fill your inboxes are not a threat: they have no customers.
Bad authors want the grey zone.
The authors' part is straightforward.
You want a line on your CV. You concoct a sciency-looking pdf, call it a paper. This takes an LLM & a few hours.
Now the real trouble begins: you know it's crap, readers will too. You need an outlet that *others* will find reputable enough to skip the reading.
You want a line on your CV. You concoct a sciency-looking pdf, call it a paper. This takes an LLM & a few hours.
Now the real trouble begins: you know it's crap, readers will too. You need an outlet that *others* will find reputable enough to skip the reading.
a man with tears running down his face says it was said that you would destroy the sith not join them
Alt: Obi Wan with tears running down his face says "it was said that you would destroy the sith not join them"
media.tenor.com
Scientific journals have always published bad papers. That's not new.
AI just made it faster & harder to detect.
What is new is that publishers don't really fight slop -- but build structures to tame it and lure it in, at scale.
Why? How? Welcome to grey-zone publisher economics.
AI just made it faster & harder to detect.
What is new is that publishers don't really fight slop -- but build structures to tame it and lure it in, at scale.
Why? How? Welcome to grey-zone publisher economics.
The reputation isn't gone until it's **common knowledge** that it is gone.
As long as people think there might be others thinking that SciRep is a good journal, or think they have plausible deniability of not knowing better, submissions will not only continue but grow.
Same as it was for MDPI.
As long as people think there might be others thinking that SciRep is a good journal, or think they have plausible deniability of not knowing better, submissions will not only continue but grow.
Same as it was for MDPI.