Guilty. I had a few hours to kill on Amtrak... the time will pass anyway y'know

(Plus I didn't know if it would work. The first version was just some Python functions I had to call from the REPL, livecoding style)

If you're really wanting to do perceptual hash based deduplication, use multiple, heterogeneous hash algorithms (phash, dct hash, mean hash, ...) as it is likely that a given hash algo will happily lossily match with very very different images--but if all hashes match, you're much less likely to have false positives.

I wrote up what I do here: https://photostructure.com/guide/what-do-you-mean-by-dedupli...

yeah there's something you can still feel in hand-written code - the variable names show how someone was actually thinking about the problem. AI code is grammatically perfect but quite hollow

Ah yes, artisanal code!

> and the script on github looks hand-written which is refreshing after having been reading AI-written code for months.

We really need a short for "is it AI or not? has entered the discussion".