Strongly disagree. There is a tremendous amount of misinformation & public hysteria about this case, so if you think something materially significant has come out of the document releases you should post it. Virtually all of what I've seen going viral on social media is presented mendaciously for circulation, which is why journalists haven't been touching any of it.

I don't think the discourse is getting "smarter at the top and dumber at the bottom". I think it's uniformly getting dumber. Smart people are impacted relatively less by the refactoring that pushed fact-checking towards the end-user, but they're still impacted rather a lot.

Playing warframe today for the first time in seven years and completely unexpectedly encountered the song we sing at solstice. The game puts all of its Good Story Tokens into that one 30s cutscene so this was more shocking than it sounds

Completely different. The software vulnerabilities we're finding with the most recent codex models are crazy. 

Isaac Newton probably sounded like a 9/11 truther in his day

Stumbled upon to this from the ACX blog. I'm flabbergasted that Eliezer believed this book.

I get this feeling using codex like it's deliberately hard to parse it order to prevent me from understanding what's going on

It's similar but not the same. You can correctly compensate people for innovation and this is still a problem.

You are ignoring the point the hypothetical is attempting to illustrate in order to quibble with practical details of the stories I tell in the post. My point is that new businesses (even productive ones, with few negative externalities) typically make money through siphoning the value of other businesses and also creating value. In some cases the ratio is highly positive (i.e. the companies I list), in most cases it is not.

If you need a more "realistic" example of a business that is almost entirely and clearly predicated on redirecting money from other businesses, see the high frequency trading industry.

I agree that that's a benefit of capitalism, but it's a point orthogonal to the one I'm making.