Has what they (AWS, Google, and also Apple) did to Parler already been memory-holed, or are you just doing the classic Niemöller thing of being shocked when it happens to you?

their own responsibility as an arbitrator of trust and safety; 

ultimately reduced public trust and safety.

Oh come on, surely pretending to actually believe the name the censors give themselves is laying it on too thick?

I've actually never heard of a US court issuing a secret order like that.

Would you expect to?

I believe there's a line from the Art of War^[1] that translates loosely to "know victory," as in "be able to identify it when it occurs, and distinguish it from defeat."

Like much of the book, the advice seems so obvious that it's silly to state until you see someone making precisely the mistake warned against. If your goal is raising awareness of the risk of AI causing human extinction (or some comparable catastrophe), having a quarter (!!) of the US public list it in their top three reasons means it's just about accomplished.

1. ^{^}

   This is from memory. It's possible the quotation is spurious. 

Try PowerToys Advanced Paste. Its "Paste as Markdown directly" works fine. I see it has a "Paste with AI" feature that looks cool!

I found an AutoHotKey script that's supposed to to do this here (with minor modifications to the pandoc command I copied from the normalize-clipboard script), but couldn't get it to work right. Maybe you'll have better luck.

It's not quite the same thing, but clipboard2markdown might work too.

where did the power go?

I don't think this is a helpful frame, but if you want to use it, the power went to the people – mostly due the democratization of information, accelerated by the internet, social media, etc. – and they then turned it on each other, wasting most of it and leaving little for the people they elect to govern with.

These are the circumstances in which one engages in kettle logic, so I wouldn't read too much into any of their arguments.

This is the core of the dispute between the USPTO and OpenAI over their (failed) attempt to trademark the term in the US, so citing their papers doesn't help resolve this.

GPTs. Yes, it's still an initialism like "LLM" but it's much easier to pronounce ("jee-pee-tee") and you can call them "jepeats" (rhyming with "repeats") if you want.