[kovek]

For me, I'm an avid browser, lurker, reader of HN, Reddit, and when I come through an interesting paragraph in an article, or a comment in a thread, I screenshot it.

Next step will be to automatically move the screenshot from iCloud onto a computer, run OCR on the screenshots, index them, and make it possible to search through their text with near-equal search terms (eg "cute dog" would match "cute puppy" as well)

Anyone has ideas on how to make that possible?

[richardliutl]

Screenshot + OCR seems tough, but sounds like it could be your preferred workflow. More robust will always be 1. a link to the article/thread and 2. a transcription of the content (esp. tailored to you [1]) imo. Your search term idea is interesting, I know Google is/has been doing this.

[1] https://zettelkasten.de/posts/collectors-fallacy/

[blandflakes]

I've considered similar but thought that ingesting the actual text from a link would be easier.