AI and RM + Any tips or tricks?

Has anyone use AI tools such as Claude to fix your RootsMagic file(s)? Also curious if anyone has any other neat or useful or creative things they’ve done by uploading their RM file into AI.

I use copilot within vs code to help write SQL and Python scripts that modify my database.

Great combination. I was using copilot conversations using the desktop app but then I realized how much more powerful the vs code configuration is.

Thanks for responding. I am not very techie but I recognize Copilot. What do you have the scripts do? One thing I am specifically wondering about is citations. I wish there was an easy way to fix them. My citations from years ago when I began are not to genealogical standards, but it will be a formidable task to correct them all.

Best answer is to look at my website where I describe some of them.

1 Like

Richard . . . just looked at your website. It’s a terrific resource, although a bit over my head. I plan to spend some time digging into it in the days ahead. Thanks.

You’ve got my interest. ChatGPT has given me a good start to SQL Lite for RM database work. I’m an old (90 years) electrical engineer with two degrees, but the most recent is 60 years old – think SL/VT (slide rule - vacuum tube). However, I’ve been interested in computers for decades – I worked in FORTRAN, Basic, MSDos. I started with a Commodore 64, graduated to PCs and about 10 years ago went full-bore Mac. So a little SQL will be good for a mental workout. Again, thanks. Herb

Responding in reverse order of your questions, I haven’t tried uploading a RM database to AI as there are file type and size barriers that my free accounts have. More common is the upload of GEDCOM files as they are text and accepted. The Facebook Group Genealogy and Artificial Intelligence is a hotbed of discussion. A caveat is to avoid uploading data about living people, for obvious reasons.

Like @RichardOtter , I’ve used AI to generate SQLite queries of RM databases. Here’s a report on my experience last year:

A subsequent query that I was super-pleased with responds to a need and feature request I submitted to RM a decade or more ago: