Hope Shinn
@hopeshinn21.bsky.social
She/her. Interested in (almost) everything, including open access and librarianship.
Reposted by Hope Shinn
@aryazandvakili.bsky.social states that the fake title isn't imported into the library catalogue: bsky.app/profile/arya... I'd guess that the library's link resolver which is registered with Google Scholar determines that a full-text should be available based on the journal/issue information only.
I was confused since I see this page on my uni's library when linked from Google Scholar, but could not get to it by searching the libraries article catalog. I realized that the link URL from Google Scholar contains the citation info - so library page is generated programmatically from the info
December 21, 2025 at 8:22 AM
@aryazandvakili.bsky.social states that the fake title isn't imported into the library catalogue: bsky.app/profile/arya... I'd guess that the library's link resolver which is registered with Google Scholar determines that a full-text should be available based on the journal/issue information only.
If your library has ExLibris Primo and uses it to make Google Scholar citations discoverable in your catalog... then there are fake, AI-hallucinated articles in your catalog 😬
When viewing the fake article in Google scholar on my university network, there is a link to access the article via my uni's library. That link sends me to a library page that makes fake article appear real... Turns out library page is made programmatically from info on Google scholar 🤦
December 21, 2025 at 5:19 PM
If your library has ExLibris Primo and uses it to make Google Scholar citations discoverable in your catalog... then there are fake, AI-hallucinated articles in your catalog 😬
Reposted by Hope Shinn
I think this post nails the actual problem, for researchers at least—AI hallucinations would simply not be a problem in academic work if we’d not normalized citation-as-signaling rather than actual engagement—you can only cite a fake paper if you’re not in the habit of reading the papers you cite
Do not cite an academic paper unless you’ve read it
AI Is Inventing Academic Papers That Don't Exist -- And They're Being Cited in Real Journals
Academic articles from authors using large language model are creating an ecosystem of fake research that threatens human knowledge itself.
www.rollingstone.com
December 19, 2025 at 7:01 PM
I think this post nails the actual problem, for researchers at least—AI hallucinations would simply not be a problem in academic work if we’d not normalized citation-as-signaling rather than actual engagement—you can only cite a fake paper if you’re not in the habit of reading the papers you cite
Reposted by Hope Shinn
I think “speak to your fake dead relative” might be the most evil application of “AI” i’ve seen so far
December 10, 2025 at 7:56 PM
I think “speak to your fake dead relative” might be the most evil application of “AI” i’ve seen so far
Reposted by Hope Shinn
The Librarian and Archivist Lament - a site is neither truly discoverable nor preserved unless you've really really put the work in (or work with those who have).
This has me reflecting on born digital history projects. In recent weeks, I tried to access a few fantastic ones—all less than a decade old—and was met with dead links.
With the death of the open internet and humanities funding more generally, what is going to happen to these projects?
With the death of the open internet and humanities funding more generally, what is going to happen to these projects?
Talking to somebody yesterday about a Cdn oral history archive that disappeared - turns out LLMs were scraping it so much it crashed their servers so it’s offline. Vile vulture technology.
December 9, 2025 at 4:16 PM
The Librarian and Archivist Lament - a site is neither truly discoverable nor preserved unless you've really really put the work in (or work with those who have).
Reposted by Hope Shinn
As we know I have Strong Feelings about academic journals expecting free peer review, charging people to publish, and charging people to read. This article framing that as part of a 'triple tax' on scientific research is really interesting!
The Triple Tax on U.S. Scientific Research
Opinion | The hidden economics behind federal research funding causes Americans to pay three times for the same body of research.
undark.org
November 20, 2025 at 6:55 PM
As we know I have Strong Feelings about academic journals expecting free peer review, charging people to publish, and charging people to read. This article framing that as part of a 'triple tax' on scientific research is really interesting!
Reposted by Hope Shinn
Our team developed an open-source WordPress plugin that lets you pull in items from your ORCID profile and display them on your site. Very cool!
Schopieray, S., & Eben, G. (2025). Linked Open Profiles - ORCID Global Participation Fund Final Report 2025 (p. 3). MSU Commons. doi.org/10.17613/ypc...
Schopieray, S., & Eben, G. (2025). Linked Open Profiles - ORCID Global Participation Fund Final Report 2025 (p. 3). MSU Commons. doi.org/10.17613/ypc...
Linked Open Profiles - ORCID Global Participation Fund Final Report 2025
Our project, Linked Open Profiles, is a WordPress plugin designed to display ORCID profile data directly on websites. With ORCID Global Participation Fund (GPF) support, we moved the tool from a proto...
doi.org
October 27, 2025 at 9:38 AM
Our team developed an open-source WordPress plugin that lets you pull in items from your ORCID profile and display them on your site. Very cool!
Schopieray, S., & Eben, G. (2025). Linked Open Profiles - ORCID Global Participation Fund Final Report 2025 (p. 3). MSU Commons. doi.org/10.17613/ypc...
Schopieray, S., & Eben, G. (2025). Linked Open Profiles - ORCID Global Participation Fund Final Report 2025 (p. 3). MSU Commons. doi.org/10.17613/ypc...
Reposted by Hope Shinn
Yeah well whose fault is that??
September 4, 2025 at 4:39 PM
Yeah well whose fault is that??
Reposted by Hope Shinn
Find all the things NOAA does in your neck of the woods by visiting this page: www.noaa.gov/legislative-...
In Oklahoma, the list is quite long! www.noaa.gov/sites/defaul...
Not as long in West Virginia, but there’s probably more than you think! www.noaa.gov/sites/defaul...
In Oklahoma, the list is quite long! www.noaa.gov/sites/defaul...
Not as long in West Virginia, but there’s probably more than you think! www.noaa.gov/sites/defaul...
February 19, 2025 at 10:35 PM
Find all the things NOAA does in your neck of the woods by visiting this page: www.noaa.gov/legislative-...
In Oklahoma, the list is quite long! www.noaa.gov/sites/defaul...
Not as long in West Virginia, but there’s probably more than you think! www.noaa.gov/sites/defaul...
In Oklahoma, the list is quite long! www.noaa.gov/sites/defaul...
Not as long in West Virginia, but there’s probably more than you think! www.noaa.gov/sites/defaul...
Reposted by Hope Shinn
Good article on LLMs, including why their use for peer-review is wrong (LLM will tend to discard what's novel & concentrate on what's already known).
@himself.bsky.social is good reading but especially this on averaging in LLMs; next token prediction moves (basically) down the gradient of most likely: it is attracted to average; interesting LLM work (Janelle Shane, bsky's own Theophite) often uses old, broken models because they're worse at this
After software eats the world, what comes out the other end?
Malkovich. Malkovich. Malkovich?
www.programmablemutter.com
October 5, 2024 at 6:21 PM
Good article on LLMs, including why their use for peer-review is wrong (LLM will tend to discard what's novel & concentrate on what's already known).
Reposted by Hope Shinn
I honestly believe all LLM results should come with a disclaimer reminding people that the thing doesn’t (and absolutely cannot) know any facts or do any reasoning; it is simply designed to *sound* like it knows facts and frequently sweeps some up accidentally in the process.
I really think many if not most of the LLM fail cases are people trying to use it as a search engine. It is the opposite of a search engine.
December 4, 2024 at 12:53 PM
I honestly believe all LLM results should come with a disclaimer reminding people that the thing doesn’t (and absolutely cannot) know any facts or do any reasoning; it is simply designed to *sound* like it knows facts and frequently sweeps some up accidentally in the process.
This is VERY COOL
A group of researchers have recreated the clothing of medieval Nubian royalty and clergy from the wall paintings in the cathedral of Faras, which sits at the border of modern Egypt and Sudan.
November 20, 2024 at 1:41 AM
This is VERY COOL
Reposted by Hope Shinn
Coelacanths are pregnant for five years, the longest gestation of any known animal, and do not reach maturity until they are around 55 years old.
November 13, 2024 at 5:26 PM
Coelacanths are pregnant for five years, the longest gestation of any known animal, and do not reach maturity until they are around 55 years old.