LessWrong 2.0 Reader

Things are getting scary with the Trump regime.

Things got scary November 5 at the very latest. And I haven't even been in the US for years.

The deportations, both the indiscriminate ones and the vindictive ones, represent a very high level of lawlessness, one that hasn't been seen in a long time. Not only are they ignoring due process, they're actively thwarting it, and openly bragging about doing so. They're not even trying to pretend to be remotely decent. The case you mention isn't even close to the worst of them; that one could at least theoretically have happened before.

The deportations were also a campaign promise. Actually the campaign promise was even more extreme.

It's part of a systematic plan. There've been a lot of administrative and personnel changes obviously designed to weaken institutions that are supposed to prevent things like that.

ICE has always had a reputation for a relatively thuggish, xenophobic organizational culture. It was already primed to get worse. As soon as Trump signalled aproval, it did get worse.

Bad conditions in detention centers are nothing new. There's never been the any willingness to spend what it would take to do them right, or to put in the kind of controls you'd need. It's politically risky to act like you care about "illegal immigrants", whereas it can be politically rewarding to "get tough". The 2020 "kids in cages" scandal was a rare case of something that got some traction. But, sure, I imagine that the newly emboldened ICE is even more indifferent to bad conditions, and may even be actively trying to make them worse. And of course if a center is already bad, putting more people into it and moving people through it fast is only going to make it worse.

They're weird: input and output in the same jack. They're for connecting to external effects, often through a cable that splits TRS to dual TS.

do you already know that a piezo signal is much improved by a preamp with >1 meg ohm input impedance?

Very much so, yes! And input impedance this high pretty much requires an active circuit.

I have a second-handed source hearing this view from a theoretical physics 4th phd student at Stanford - he believes less breakthroughs nowadays as the field becomes more and more established, and this was exactly why he was a bit discouraged/sad. Not sure if things has changed, and that may or may not be his personal view.

The Krome thing is all rumor - looking into it, you see numeric estimates like

>According to its official figures, there are 605 people detained at Krome, although the capacity is 581. While ICE is looking for ways to increase its current detention capacity of 40,000 nationwide to 100,000, lawyers and activists estimate the real number is much higher. Some speak of double the capacity, others of up to 4,000. 

 

“Activists and [activist] lawyers say number is huge” is not news, and shouldn’t dumbfound the reader.

The water claim is also weird. I tried watching one of the instagram links, and it shared so much stylistically with mind-killing videos I remember from the BLM era that I had to turn it off. 

Like, maybe some of this stuff is true. I don’t have evidence against. But when I was deeply involved with the protest scene in 2014-2015, I remember every arrest being an opportunity for claiming major mistreatment. Everything from the way police carried resisting arrestees, to when and if arrestees were made to change into jail uniforms, were spread frantically on social media as clear examples of mistreatment. 

Once, when I was arrested, and we were being transported to the larger jail via van, the other arrestee (to be clear: not related to protests) being transported with me banged his head on the metal separating grate repeatedly, presumably with the idea of later accusing the police of beating him. 

I’d always scoffed at police claims about detainees hurting themselves to get social ammunition, but I’ve ridden in a police van once in my life, and saw this. So now I think detainees often tell very tall tales.

All this isn’t to say “this proves your links are false”. But rather to say this is a low standard of evidence. I think it would be really bad if people started just dumping rumors and accusations on LessWrong whenever those accusations pointed at politicians they already didn’t like.

Social media posts by activists are mind-killing. Like, take a look at previous posts by that instagram account in the post: many are about celebrities, or her breakup, but when the videos are political, they are pretty clearly pro-migrant and anti-trump. “Partisan social media account” is typically not the best information source for rationalists.

Potential token analysis tool idea:

Use the tokenizers of common LLMs to tokenize a corpus of web text (OpenWebText, for example), and identify the contexts in which they frequently appear, their correlation with other tokens, whether they are glitch tokens, ect. It could act as a concise resource for explaining weird tokenizer-related behavior to those less familiar with LLMs (e.g. why they tend to be bad at arithmetic) and how a token entered a tokenizer's vocabulary. 

Would this be useful and/or duplicate work? I already did this with GPT2 when I used it to analyze glitch tokens, so I could probably code the backend in a few days. 

Downvoted. See Burdensome Details [LW · GW]. I particularly dislike predicting "Algorithmic Breakthroughs"

I very much agree with the approach and the values in virtue; in case for humans, we enforce virtues either through empathy or law/punishments (in modern societies); wondering how that can be most effectively translated to machines in a consistent way

... but that means she learned what it was at age 5. I'd assume most people learn between about 4 and 8, maybe 10...

I am aware of it and I regret to say that I've tasted it...