This is an example of an AI model that is useful for work. OCR traditionally only works well if you have the right fonts and a clean scan. An AI model trained to recognize handwriting can OCR to much better accuracy than old OCR systems, which were much more rudimentary and only trained for specific typed fonts, making bad scans and handwriting all but impossible to OCR.

The reasons one would use OCR is typically when scanning documents into pdf. OCR makes the document searchable when viewing it. You might think “well I can use my eyeballs to search it”, but that is only good if the document is less than 3 pages long. If it’s a 50 page document, you are definitely going to want to use a search feature.

Also a blind person scanning a handwritten document to read it. Not everyone that can’t see has super touch and can feel the ink on paper.

Food for thought before you start shooting the downvotes because it’s an AI model.