The shift in how we date pottery shards feels like night and day from 15 years ago

I used to spend hours matching rim profiles to published typologies from the 1960s (which never quite fit), but now I just snap a photo and run it through the ceramic database from the University of Arizona's lab. Anyone else notice how machine learning has basically replaced the old comparative method for common types?

2 comments

2 Comments

nina1802h ago

Man that's exactly what I found switching from old rim charts to the digital database. It saved me so much time on basic type identification once I got the hang of the photo matching. But I still keep a copy of the old typologies around for the weird edge cases that don't fit the machine's template.

reese_bell2h ago

And here I thought I was the only one still clinging to paper like a caveman.