@sotolf may I ask to not further mention me in such posts, as I get notified every time. And I don't think, the tone is appropriate.

There was a survey, we got some interesting details about the implementation, we can draw our conclusions. At least, that's what I did.
No need to personally attack some dev folks in the Fediverse.

@gregtatum many thanks for the insights. Very helpful. @firefoxwebdevs

@firefoxwebdevs if I better understood what "downloaded-on-demand ML models" means in practice, I'd feel more qualified to participate. No external service involved in any way, right? How resource hungry was/is training that model?