01:04:28	Ondřej Dušek:	Q1: Did you have any access to speakers of those low-resource languages for assessments -- how exactly did you check the meaning equivalence?

Q2: Did you also evaluate ChatGPT in the other direction? Your table only shows translations *into* the low-resource languages for ChatGPT, right?

Remark: for the dissimilar examples in DE/HSB, I can still see some level of (abstract) similarity -- "And that's not all" (HSB) vs "Tomorrow is a day too" (DE) both suggest there will be something to follow.