nlpparty

I just assumed that if many people, especially not proficient language users encounter this word in the text generated by ChatGPT they would look it up.

If your select only USA, the trend disappears.

Twitter has been accused of being full of bots long before ChatGPT appeared. For 140 symbols, a template with synonyms would be enough to create mass-generated content.

It has been for me the last 15 years like this.

I suppose it is just Amazon problems. I have never lived in the area where Amazon is prevalent. Where I live, search engines still can't find synonyms or process misspellings.

It's just inevitable. Imagine a world where we get a cheap and accessable AGI. Most work in the world will be done by it. Certainly, it will organise the work the way it finds more preferable. Humans (and other AIs) will find it much harder to train from example as most of the work is performed in the same uniform way. The AI revolution should start with the field closest to its roots.

I use i3-wm. I know that there is Sway but it will require some effort to migrate. Also there are fresh reports that Wayland+Sway have problems with NVidia (even worse I have AMD + NVidia). I'd wait till it gets resolved or my current setup stops working.

I don't know what Pipewire is as well. I've used Linux daily for 10+ years, but I've never bothered to use Wayland.

Photos have never been a fundamental proof if the stakes are high or you have an idling censorship institution. Soviets (and maybe others, I just happen to know only about them ) successfully edited photos and then mass-reproduced them.

just some google link about the issue: https://rarehistoricalphotos.com/stalin-photo-manipulation-1...

Russian doesn't have the guttural 'r' of French.