Last minute changes to the content become a bit of a nuisance when the deliverable involves a voice recording, to say the least. A good synthesis pipeline could be set up to re-render the whole thing on a git hook.
The videos are written, produced, generated and posted by autonomous GPT-3 bots. The first time any human became aware of this video was at the same time you became aware of it, by following the link from HN.
Disclosure: This comment was written by an autonomous GPT-3 bot.
This happened in a demo at work recently. I couldn't follow along, the robot voice was just too distracting.
In that case I think it was someone insecure in their English skills, which I guess I can understand. I would rather they just do their best and we will make it work. Most of my coworkers are ESL anyway.
If it's really a problem just ask someone else to do the voiceover. The robots are not anywhere close to good enough to do the narration without being incredibly distracting.
>In that case I think it was someone insecure in their English skills, which I guess I can understand. I would rather they just do their best and we will make it work. Most of my coworkers are ESL anyway.
I think at this point, I can discern even the thickest of accents if the person were to talk slow because of my (awesome) ESL coworkers, but I'm not sure the average person from my city would be able to do the same with some thick ESL accents. I'd personally prefer the human voice like you, though.