HeyJaredComputational History
A clearinghouse for AI methods in History, where historians swap techniques and perspectives
- Indexed issues, last 90 days
- 7
- Latest publication
- Sep 30, 2026
- Audience
- Checking…
- Earliest in this view
- Jul 7, 2026
Latest issues
How AI Saved One Professor from Drudgery (opens the original)
Read excerpt
<img alt="Untangling the paperwork, setting the table" class="sizing-normal" height="485.3333333333333" src="https://substackcdn.com/image/fetch/$s_!3r3m!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fi
AI and History Conference (opens the original)
Read excerpt
Online registration now available!The AI and History Conference is happening October 15-16 at Johns Hopkins, and registration is open.This isn’t a conference about whether historians should use AI. It’s a conference for people, like you, who are already doing the work or want to know how to do the work: building OCR pipelines that beat commercial services, training human-in-the-loop machine learning, linking 2.2 million historical places across languages so historical GIS stops being reinvented
AI in the Archive: Saving Time by Turning Photos into Structured Metadata (opens the original)
Read excerpt
We’ve all been there: you spend a week at an archive taking hundreds, if not thousands, of photos of documents. It’s a great trip: there’s so much useful material for your research, and while you’ve been jotting down notes as you go, you’re going to need to dive into the photos to get everything you need. But first you need to get back to revising that article your reviewers just sent back, or finish grading those final exams. By the time you’re ready to dive into your archival haul a few weeks
The Experimental Method in History: continuous batching with oMLX and vLLM (opens the original)
Read excerpt
Historians are not comfortable as experimentalists. The past, after all, happened just once and we can’t change it. Yet as we learn to use LLMs, we need to embrace the experimental method. Getting the right combination of model, hardware, and configuration for your particular corpus requires playing around. There is no universal solution.I want to convey in this long-winded post, if not the glamour and drama of my recent experience, at least the experience itself. I used experiments to figure ou
Claude on Cluster (opens the original)
Read excerpt
You Don’t Have to Learn SlurmA few months ago, Loren Moulds wrote that you deserve the cluster. He noted that the quiet assumption in the humanities is that high-performance computing belongs to physicists or geneticists but not to us, even though we too have needs for powerful GPUs to do our work (like OCR!) If you’re affiliated with a research university, you almost certainly already have access to GPU-equipped machines that can turn a faded, ink-stained index card into struc
Publishing over time
Last 90 days. Choose a month to open its work.
Recurring subjects
Named in the text we hold. One piece can cover several.
Audience
No verified audience measurement yet.
About this data
Counts cover the work we have indexed. Tone needs enough text and a confident classification. Excerpts and episode notes are not full articles or transcripts.
Identity or attribution wrong? Suggest a correction.