Stay safe

Sorry for giving you the wrong impression - I'm not suggesting that *at all*. I know exactly how stretched most academic libraries are.

However, digitisation will still happen, obviously not on the vast majority of material, which as you say will remain undigitised.

*corpora*.

Balls.

What resources I suggest? I'm sorry - I'm not getting your point. I'm just suggesting areas of current academic historical practice that I think will (and are, tbf) start using AI or AI-enhanced tools soon.

The researchers in this thread? Sorry - I was just responding to the original post.

But cool - historians should absolutely be able to use whatever sources, digital or not, ideally. The problem is always time, expense and access. So we see an awful lot of academic research using digitised sources.

...if these valuable non-digitised materials need to be archived (and there are plenty of historians working on these projects, for preservation/access purposes), AI will be part of the historian's toolbox.

It's not how it should be, ideally, but the profession is biased a lot toward material available online.

Having said that, back to my first point - with the ubiquity of very good photographic tech, AI transliteration tools will still get better results than straight OCR/HTR so..

For absolute sure. And this stuff is valuable and crucial to historical research. But with the massive loss of funds and time pressures on modern academic historians, most rely heavily on digitised sources, which are quick and relatively cheap to access.

Finally, we're already seeing a ton of AI exploration in Digital Humanities and DH forming a growing part of UG/PG training. So AI tools for parsing, analysing and presenting large corpera to enable new research, methods practically impossible through traditional methods.

Of serendipity here, but I'd imagine traditional Boolean-based search will still be available. And academics aren't currently great at effective archive searches (I've spent a long time training them) - so a locally-trained AI agent might find them more and better results.

Then, I'm assuming that most commercially available digital archives, most large open repositories and even most smaller boutique collections will include some sort of AI tool, a RAG or similar, trained on the corpus, to facilitate exploration and discovery. Sure, there will be some loss

I'm assuming that most digital archives will employ AI tools to transliterate the physical materials, as by then, these tools will have a character accuracy rate that far surpasses cutting edge OCR/HTR. Not strictly a tool in use, I know, but AI benefiting historians with better digital surrogates

Really great stuff - thanks, Windy.

We were sick of speculation. So thanks to the 1,100 people who shared their data with us, we were able to show how TikTok keeps users hooked on the app. Over just 5 months, users appeared to use the app more compulsively.

Stay tuned for more stories off of this incredible dataset!

We definitely tried to attack in places - not as much as I'd like, but still. It's not like we're an elite passing team but I felt you could really tell they all felt uncomfortable on the ball, and it disrupted our rhythm in possession.

I'm not going to judge them on this game. I think the pitch is such an advantage for them and really makes it tough on the opposition playing with the ball. It bounces and rolls funny!

He's missed some crucial windows to invest and build on success. The Redknapp team with Bale and Modric, and we bring in Saha and Nelsen. Not backing Poch and letting that team stagnate. Not paying wages even half of what top 6 rivals were in the past decade.

He wasn't Satan, but he wasn't a model

I am the Mob. Best delivered first line in pop history