I've been sceptical about AI for genealogy but yesterday I came across a blog post
https://familylocket.com/handwritten-text-recognition-by-claude-3-5-sonnet/
I've been trying it this morning and thought others might find it handy.
I've got Claude 3.5 to transcribe several UK 1841 census schedules and then convert them to HTML which I pasted into a GCII note for that particular source.
After a bit of mucking about, I came up with the following workflow. I use Typinator to write the instruction(s) since they are pretty complex.
1. Select and copy the relevant household from the census image and paste into a file saved to desktop.
2. Attach the file to the Claude 3.5 chat and give it the following instruction:
You are an expert genealogist and transcriptionist. Transcribe this document and put anything you can’t read in square brackets with question marks or your best guess. Preserve the original line lengths in the doc.
Convert multiple spaces into a single space and replace that space with \t. End each line with \r\t.
3. Correct any transcription errors, add any headings you might want, and copy corrected output.
4. Give Claud 3.5 the following instruction:
Convert the following text into HTML using \t to delimit cells and \r for a new table row. Use heading 2 for the first line and heading 3 for the second. Precede the first line with the following <div><--name Unnamed html text --> replacing Unnamed html text with the first line. Make the final line </div>.
5. Copy the output and paste it into a GCII note.
The attached, nicely formatted, note took me about six minutes all up. I imagine it'll be quicker once I get the hang of it.
Cheers
Mick