Sarah Thelen

doing historiography again

Everyone Is Using A.I. for Everything. Is That Bad?

i just find this so bizarre. searching through texts and doing the reading for myself is how i actually think through ideas! trying to put together a sentence is how i figure out what i actually want to say!

For me, what makes most of the works of history that I consider great are those that offer a glimpse into the unique mind and obsessions of people who have sifted through and curated a large number of sources, and made compelling and often idiosyncratic choices about what to highlight./9

if you ask it for a thing that looks like an answer, it gives you a thing that looks like an answer

if you ask it for a thing that looks like a source, it gives you thing that looks like a source

if you tell it that it's wrong and ask it for a thing that looks like an apology, that's what you get

This means as a historian or political commentator, you spend quite a lot of time having to explain “yes, I know, I’m not saying it SHOULD be like this, I’m saying it IS like this,” unless you couch everything you say in fluffy adverbs (“Unfortunately…”), and often it happens even if you do.

I know! It's so good!

ChatGPT is bullshit - Ethics and Information Technology Recently, there has been considerable interest in large language models: machine learning systems which produce human-like text and dialogue. Applications of these systems have been plagued by persist...

This is definitely worth reading if you haven't already!

link.springer.com/article/10.1...

🎉🎉🎉🎉 Hooray!! Congrats!! 🎉🎉🎉🎉

To wrap up, some key distinctions I use to do my work:

Public vs. audience
Journalism vs. the media
Truth-seeking vs. refuge-seeking
Political vs. politicized
Issues vs. troubles
Ritual vs. transmission
Expect vs. predict
Subscription vs. membership

"When in doubt, draw a distinction." 25/

"Gen AI is early 1900s spiritualism for digital moderns" prediction stays winning

How Irish booksellers feel about the arrival of Amazon.ie: ‘It’s a very real danger to the industry’. Bookshops offer something that an exclusively online retailer cannot, but they may need support to compete against a corporate giant, writes Tim Groenland

www.irishtimes.com/culture/book...

Hi! Are you a US researcher who spent time thinking about humans? Then your work is in danger of censorship and loss. I'm here to walk you through basic self-archiving.

Maybe you think I am being hyperbolic. You only worked on bacteria! Not your problem. Do me a favor and join me anyway.

New chapter out by me where I argue that people who claim to collaborate with LLMs are mistaken about the role of the bots, and that they're weak ass tools rather than serving the role of an actual collaborator academic.oup.com/edited-volum...

Some really great advice here

The HEA has published Ten Considerations for Generative AI Adoption in Irish Higher Education, which provide a foundation for institutions to shape policies and practices around generative AI.

hub.teachingandlearning.ie/ten-consider...

Universities — including the schools that most aggressively shut down student activism — are now fully panicking because the Trump administration is about to shut down all kinds of funding for research and more. Too bad they made it clear that their students shouldn’t speak up and speak out.

I have been soft lobbying for the ALA to change Banned Books Week to Freedom to Read / Readers’ Choice week for like a decade bc of this very impression. Having a banned book is not a positive thing. It’s an attack.

Going to read up on this once the reporting and applications are done:

www.lawfaremedia.org/article/chat...

Fascinating!

I was writing about deterrence today and remembered my favourite way of explaining it

Channel your inner Amelia Bedelia

History is never as straightforward as some people want it to be

Studying and teaching history is a form of resistance. 🗃️

Ahem, *crossing*

Unfortunately, the pedestrian crossing by me is *more* dangerous than cruising in the middle of the street

anyway it's probably fine, it is definitely reasonable to have a machine that lies confidently about things you may not be certain about. a program that's baked into the OS and designed to be turned to for help when you're stuck and uncertain SHOULD randomly make shit up, including references,

From _Clio Wired_ (2011)::

"The danger is ... that as the Web goes down the well-worn road of radio, tv, and cable, in which large 'infotainment' conglomerates come to dominate the wires, 'choice' becomes narrowly defined as the competition between two or three very similar 'products.'"