#צ'��יה

[thetransintransgenic]

Huh. Wow, Google:

https://encrypted.google.com/search?hl=en&q="יה תשפק"

Is is really that hard to tokenize Hebrew correctly?

“... של זרעי צ'יה: תספק מרקם ...”

“... של ג'ורג'יה, תספק משאבים ...”

For loan words (from English, basically), Hebrew uses apostrophes to modify Hebrew letters into loan-sounds -- “ ג'‏ ” means the soft “ g “ sound like “giraffe”, and “ צ'‏ ” means “ ch “ like in “church”.

So “צ'יה“ is not two words or anything -- it’s the ones loan-word “chia” (”זרעי צ'יה“ means “chia seeds”)

Similarly, “ג'ורג'יה“ is “Georgia”.

Like yeah this is a tricky edge-case, but Google usually knows its stuff.

There are a lot of Israelis in tech -- you’d think they’d have noticed and fixed this.