iiinova — where your digital self keeps evolving

Jacob Polay @historyjacob.bsky.social · 01/10/2026

The most recent blog post in our series on AI and Collaboration with @activehist.bsky.social. In this post, I look at how AI is changing what tools historians can access, and how we should build and share tools with each other. activehistory.ca/blog/2026/09...

activehistory.ca

Stop Choosing Tools. Start Building Them.

Part of the reason those systems created these kinds of errors was that they were not created for historical research. However, looking at these errors from a historian’s lens allowed me to create …

062

Reposted by Jacob Polay

ActiveHistory.ca @activehist.bsky.social · 25/09/2026

Chloë Farr with Jessica Jack and Jacob Polay explore the use and development of OCR enabled by AI vision language models (VLM-OCR) in galleries, libraries, archives, and museums institutions. This post is part of a series on AI and Collaboration. activehistory.ca/ai-and-colla...

activehistory.ca

The AI-Enabled Boom in Document Transcription

The outputs of many OCR technologies were developed for training LLMs, but that are frequently of little use in the humanities. They can be used by people with specific data mining and data science…

132

Reposted by Jacob Polay

Jim Clifford @jimclifford.bsky.social · 01/05/2026

Talkie LLM is a hot mess. There is a fundamental danger in treating a historical LLM like an archive. When we read primary sources, racism is anchored to a specific context and author. Talkie strips away that provenance, synthesizing the worst pre-1930 biases into a confident chatbot.

195

Jacob Polay @historyjacob.bsky.social · 28/04/2026

This is a fascinating project, though the use of the term "vintage" makes me question the authors' desire for a specific aesthetic over historiographical groundedness. Does this model's dataset replicate the silences of the colonial archive?

000

Reposted by Jacob Polay

Jim Clifford @jimclifford.bsky.social · 09/03/2026

My student @historyjacob.bsky.social has shown the new OCR tools work very well with early modern English print text. We use to struggle with the long s and many other issues. This opens up NER on sources that use to be too messy. He also developed a fine tuned NER model for Atlantic commodities.

162

Reposted by Jacob Polay

Katie McDonough @kmcdono.bsky.social · 06/03/2026

Join the Lancaster-Manchester Environmental #DH Seminar on March 11 @ 3pm UK (online) for a talk by @jimclifford.bsky.social & @historyjacob.bsky.social: "Solving OCR: Using olmOCR to Follow Commodities across the British World" www.eventbrite.co.uk/e/solving-oc... #dhist #ocr #envhist 🗃️

eventbrite.co.uk

Solving OCR: Using olmOCR to Follow Commodities across the British World

Join the Lancaster-Manchester Environmental DH Seminar for a talk and discussion with Jim Clifford and Jacob Polay (Saskatchewan).

21510

Reposted by Jacob Polay

Hank Green @hankgreen.bsky.social · 10/09/2025

It’s ok to put the phone down.

1667062769

Jacob Polay @historyjacob.bsky.social · 14/05/2025

Stories from the archives #1: "You can't put the documents on the ground just because the lighting is better down there"

000

Reposted by Jacob Polay

Timothy Snyder @timothysnyder.bsky.social · 15/04/2025

3/4. If we accept the idea that moving a person from one place to another undoes rights and disempowers the judiciary, we are endorsing the basic Nazism practice that enabled the killing of millions.

108180954867

Reposted by Jacob Polay

Brodie Waddell @brodiewaddell.bsky.social · 22/11/2024

How long would your commute take in 1680? Mine would be about a day and a half. @camunicampop.bsky.social has a brilliant tool for calculating travel times in Roman, 1680, 1830, 1911 and 2024. 🗃️ www.travelintimes.org/journey/52.2...

47663238