Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I don't think it's an MD5 or SHA512 hash, since just changing one pixel would be enough to evade the scanner. My understanding is that it's heuristic similarity detection, which has a much wider footprint for collisions.


It was a hashing technique that would output similar outputs with similar inputs, maximizing collisions rather than minimizing it

That's why people were initially all up in arms about

But they didn't need to do that.

A regular hash where "just changing one pixel would be enough to evade the scanner" is already good enough




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: