The hardest thing about being good isn’t untangling complex moral dilemmas, but holding to simple principles consistently and unyieldingly, despite the world constantly insisting you would be savvy to abandon them

I think that both the helldivers reference that calls the target a fascist and the reference to the antifascist song found on groyper playlists could be from either side. So suspect, yes. But it’s also reasonable to suspect that it was motivated by antipathy from the right.

Populism is you govern by elite-driven policy making

The editorial board saying that Zohran “seems to lack the political savvy and instinct for compromise” that Sanders has, after complaining that Sanders “boasts that compromise is anathema to him” in their 2020 endorsement, is quite revealing in my opinion.

AI can never be held accountable, so I actually don't think that AI should report anything, including stuff that is actually bad.

It’s interesting because on the surface, reporting what another news source reports is commonplace. But on the other hand, if every person got their news this way, the entire industry would collapse.

There are 4 distinct acronyms in that thumbnail

What I'm learning from these papers is that people are bad at convincing other people of things, rather than that LLMs have incredible persuasive skill.

You are never beating the “fetish for getting yelled at” allegations

If you don't want to be judged by tossed-off responses to tossed-off trolls, then don't make tossed-off responses to tossed-off trolls.

Seems to match my own intuition on how these perform in R

That being said, food prices have gone down 30% when adjusted for median income, so I guess you win some, you lose some. But I definitely spend more on rent than food, so if I had the option, I'd switch them.

Oh god, paper searching is a fool's errand with existing AI. I don't even think the OpenAI "$200 a month-100 uses per month-5 to 30 minutes per task" tool can do it right from what I've heard.

Skill issue