ControlsFreak
No bio...
User ID: 1422
ControlsFreak
2yr ago
I don't have any particular concerns over the way most people raised this. It was a standard "parade of horribles", where you try to come up with the most extreme 'hypos' (in the legal lingo) to push an interpretation as far as it can go in order to see if that's really an interpretation that you can live with. This one is pretty extreme, because the President's powers in this area are pretty plausibly vast. We saw somewhat similar language pushing on "assassinate a US citizen" around the al-Awlaki business. The President's war powers are among the most dangerous powers given to anyone in the country; they sort of have to be. We want a President Lincoln to be able to order the killing of rebel soldiers and leaders, even on US soil, in cases of genuine rebellion. How to draw lines is hard, and frankly, we probably still haven't managed to lay out any lines with real conceptual clarity; it might just not be possible to do so a priori. I'm pretty fine with people at least considering different scenarios in order to argue that we need to be careful in how we draw the lines as a separate thing from, "Whelp, if we're going to draw the line where you want to draw it, President Biden actually should order Seal Team Six to assassinate Trump." Probably some blowhards on reddit actually said the latter thing, but most relatively serious folks (to the extent there are many left) were doing the more sensible thing.
ControlsFreak
2yr ago
Is Loury going to stop constantly talking with McWhorter in the way he isn't talking with many people to the right of his?
I doubt it. I don't think he chooses whether or not to talk to McWhorter based on political purposes. I think they do just honestly go way back as friends first. It's extremely easy to shun people who you're not friends with based on political differences. When you start doing it to your pre-existing friends, it's a short road to loneliness, above and beyond the typical concerns people have (especially influential public figures) with determining whether anyone who appears to be a friend is acting that way because they genuinely like being friends or if it's just a means to an end.
ControlsFreak
2yr ago
It's the phenomenon mentioned a couple days ago by someone here - the more probable it becomes that Trump is going to win, the more value there is in having shown that you were "on his side"... or, at the very least, not against him. This is particularly poignant for extremely wealthy leaders of highly-regulable industry. Nobody wants the full Elon Musk treatment. Zuck may have genuinely believed various left-leaning things in the past or he may have just thought that he had a good deal going where they were mostly going to protect him so long as he played ball enough. It's impossible to know. Likewise, it's sort of impossible to know whether he's actually emotionally/intellectually moved to take this new lack-of-position, or if he's seen the writing on the wall enough that he feels compelled, for the sake of his company/industry, to find any plausible way to soften on political issues without getting crucified by the media or other interest groups that could cause him problems (like, for example, meta employees, who might already be having meltdowns that their boss can't even manage to oppose Hitler). He's baaarely inching out just hoping that the story of the assassination attempt will help give him even the smallest amount of cover from criticism. This is a guy who is scared to death that he's going to suffer regardless of what he does.
ControlsFreak
2yr ago
Now I'm just very confused. I thought this was all just old hat, been done before, obviously out there for anyone who cares to see. Now it's a SciFi concept come to life. I have no idea anymore...
ControlsFreak
2yr ago
...and why would one say that it is wild?
ControlsFreak
2yr ago
Thinking this implies he had no idea about the demographics of crime is kind of wild.
Let's walk through this, then. What do you think his tweet does imply?
ControlsFreak
2yr ago
You (and others) have said this, that there's nothing new, everyone knows how this works, etc. Matt Yglesias didn't seem to know that. It's "wild" to him. Perhaps it is likewise "wild" to many others.
ControlsFreak
2yr ago
I believe the old joke is, "Machine learning is just OLS with constructed regressors."
ControlsFreak
2yr ago
Whenever I smell something like that (and some entire outlets), I just default to throwing the link in archive. If it's not there, the internet is telling me that I should skip it.
ControlsFreak
2yr ago
On the topic of a small number of relatively 'known' people being involved in an outsized portion of the crime problem
or
AI is sometimes allowed to say things that are otherwise not allowed to be said, so long as they make sure to say that it's definitely not racist
Machine Learning Can Predict Shooting Victimization Well Enough to Help Prevent It is the name of the paper. They took arrest/victimization records in Chicago and tried to predict who was likely to be shot in the next 18 months. 644,000 people in the data; of the top 500 with the highest predicted risk, almost 13% were shot. That's the top line. 13% accuracy might not seem like much, but they claim that the rate is 128 times higher than that of the average Chicagoan. For context, that's 64 shooting victims over an 18mo period. I don't know what the total 'shot but maybe didn't die' rate is, but Chicago has in the ballpark of 600 homicides (by all means) each year.
This is not about who did the shooting; it's about who was shot. The implicit argument is that most shooting victims are close enough to the criminal world. Even if they were just purely victims before, it is at their doorstep. Plausibly, if a little old lady just happens to live in a really terrible neighborhood and had to report being the victim of various prior crimes, this could indicate that she is also at risk of getting shot, too.
They definitely go out of their way to say that, yes, Black males are more likely to have prior data in the system, but that the system still predicts with similar accuracy across demographics.
I don't know how practical these sorts of things will be to actually use for any purpose, but this paper dropping is definitely adding some fuel for the folks who think that a variety of criminal problems are mostly concentrated within a relatively small subset that could, at least in theory, be somewhat identified.
ControlsFreak
2yr ago
Man, from the depths of my memories of Supreme Court cases comes Wood vs. Moss, where Bush the Younger made a last-minute change in his plans, up and deciding to have dinner somewhere that wasn't on the schedule. They were going to eat on the outdoor patio, to boot. The only problem was that protest areas had already been approved for two groups (one of supporters, one of opponents), and the SS thought that the opponents' location was too close for them to be able to secure it. Rather than be all, "Sorry, Mr. President; you can't eat here tonight," they decided to move the protesters. This led to a lawsuit which claimed that the real, hidden, motivation for moving the opposing group was to discriminate against their viewpoint in favor of the supporters.
I imagine the SS has to deal with stuff like this all the time. They probably have to be carefully selective about when they actually choose to tell the President, "No." Most of the time, they probably just scramble and try to hold on. Of course, heat of the moment, credible threats, etc. all end up smashed together in a horrible, subjective scale for what type of action you take. They probably have to be pretty careful in trying to manage the relationship, too, where, similar to what you're saying, different VIPs may have different preferences/risk tolerances. They don't want to piss off the President by saying no too often, but if you end up with a dead President, you don't want to be left only being able to say, "Yeah, but this guy always got pissed when we said no or took some conservative action, so we just got into a habit of accepting more risk for him." It'll never play well, especially if anyone is looking to blame you for not being cautious enough in the aftermath of a tragedy.
Of course, no one on TheMotte is going to have any insight whatsoever on what Trump's relationship to the SS actually looks like, similar to how there was all the haranguing in his first term about how he was engaging in the PDB. Some folks wanted to skewer him, and others were saying that they worked toward productive, engaging meetings, and others yet were sure that the bureaucracy was intentionally being intransigent and sabotaging the relationship. Only the tiniest percentage of people actually know, and they're generally not talking unless they are doing so to promote some agenda or another.
If I were forced to steelman, perhaps it could be done. I've been reading some of the articles about what their plan was and what went down (NYT just had a pretty detailed one that didn't seem atrocious). The biggest aspect would be the SS/local divide. Folks in this thread have already made jokes about the local folks probably being incompetent, screwing stuff up and not doing what they're supposed to do. The SS could have thought, "Yeah, we're not going to have some obese local hanging out on that roof all day, just hoping that that they don't hurt themselves," while still thinking that their own counter sniper team was obligated to take such a position, for lack of any better positioning options... plus a little overconfidence that their guys would totally be capable of handling it, even though the dumb locals couldn't (as I mentioned, I think this belief turned out to be wrong, considering that both their guys immediately started slipping down the roof).
I still totally agree that things were immensely screwed up, because that roof is the single most obvious place to attack from in the area.
EDIT: I also think that if their reasoning was something like this, they really can't vocalize, "The reason why we thought it was a safety issue is because we have no confidence in obese, incompetent locals." Too much social desirability bias.
ControlsFreak
2yr ago
Free Speech is a spook, an incoherent concept that collapses the very instant it drifts outside the bounds of a rigidly-coherent values environment.
Plausibly so. One area that is grossly lacking any remotely suitable theory is one of character development, especially on a large scale such that it could produce any sort of a coherent values environment (not even necessarily a rigid one). To even think about such a project descends into questions of things like deterrence/rehabilitation and perhaps even free will, itself. One tension I observe is that when the left/communists engaged in large-scale projects of "character development" (given that it is a type of character that many in this forum would think is bad/harmful), many folks are pretty okay with believing that it can, indeed, be relatively successful in producing a coherent values environment (perhaps one that is, indeed, rigid).
That is to say, it is perhaps all well and good to have a heuristic rule where one looks at the population which is considering adopting a norm of free speech and concluding, "If such a population has sufficiently coherent values, a norm of free speech is potentially plausible/possibly beneficial, but if it does not, then it is not." But the core problem is somewhat upstream of that, and it is one that we have almost no clue concerning how it works. There are likely, indeed, almost surely, some feedback loops back upstream, but it becomes an even more impossible task to perform any analysis on such feedback without any conception whatsoever of how the thing it's feeding back into works.
What will I do? Probably nothing more than perhaps consider writing a comment here. This whole business is pretty far afield of my personal priorities for action.
What comes to mind from the discussion? Well, frankly, I just read the discussion below about dress codes before I read this, and what comes to mind is that there are, perhaps, some connections to be made. I haven't fleshed out those connections thoroughly, but I see a haze around them. What are the interests in preserving dress/speech codes? Are they individually beneficial to various actors? Are there coordination problems getting in the way of 'better' equilibria? I know people poke at game theory downthread, but I think the core insight is not to just shut up and calculate, but rather to just investigate the interplay between disparate motivations and incentives. A mathematical version may or may not be a suitable abstraction.
Moreover, I see connections in that a variety of folks are trying to do some implicit inference, which seems to me to require a significant amount of the same reasoning that is necessary for a game theoretic approach. That is, they would like to infer something about the 'type' or 'quality' of the players, be it a single parameter like "trust level" or something else. But to do so, it seems necessary to situate those parameters within the context of the incentives/game being played, because only then can the parameters be correlated to the behavioral strategies which are observed.
ControlsFreak
2yr ago
From what I could tell in the video posted here a day or two ago that merged a bunch of different videos that have become public from random people's phone cameras or whatever, the sloped roof that the counter sniper team was on looked like it turned out to be a problem for them. They were likely already alert to a possible threat in that direction and were already oriented that way. It looked like a metal roof, and when the first shot cracked, both of the guys in the shot visibly flinched and started slipping down the slope. They ended up off balance and struggled for a second to get back up into position, reacquire the target, aim, and take the shot. The delay from slipping may have enabled the final rounds from the would-be assassin to happen.
I don't have any grand pronouncements as to blaming anyone or anything, but I imagine that AAR and future training/procedures will make such folks at least a little more likely to include things like a grippy, non-slip mat in their kit in the future in case they need to establish a better base from which to fire. I'm binning this just under, "The real world happens like that sometimes, and people hopefully learn."
ControlsFreak
2yr ago
I've been a fan of the following theory. Donald Trump could not have entered into a quid pro quo arrangement (trading money for official acts) with Donald Trump by using his own money to pay for his own expense (regardless of whether the FEC's interpretation of 'private use' holds, though this would be a perfectly fine alternate theory), just because he used an intermediary for the transaction. Since the underlying crime referenced by the conviction is a federal crime, there is a substantial federal question, and prior Supreme Court precedent has established that the federal statutes in question can only withstand First Amendment scrutiny to the extent that they are interpreted in a tailored fashion so as to only address quid pro pro or the appearance thereof. Vacate and remand the question of whether the jury instructions were such that they can sustain the conviction, given that the federal crime alleged is actually not a crime.
ControlsFreak
2yr ago
No idea. We're several years into everyone considering the hypo of a self-pardon, but I don't know that we're even a millimeter closer to having any real answers. Doing that would be a pretty controversial move with potentially serious long-term impacts. Vastly easier to just quote Trump vs. US and this decision and have Trump give a comically large "YOU'RE FIRED!"
ControlsFreak
2yr ago
Quite the opinion. There's a lot in here for the small handful of folks who like the "Officer stuff" and actually care about the relevant law. The headline is about Constitutional interpretation, but the real main stage is detailed statutory interpretation. So while at first blush, it might seem like the overall topic is massive cert bait, there is significant leeway for the Eleventh Circuit to really hone it down into the best statutory arguments. At that point, it could look a lot less sexy for cert, so we'll have to sort of wait and see.
Of course, everyone else (who doesn't really care about "Officer stuff" qua "Officer stuff") will only care about the practical implications for Trump, which as you say, will almost certainly be delayed until after the election. At that point, there are a myriad of ways that it can go. If Trump wins the election, Trump has plenty of legal ammunition from Trump vs. US to simply fire Smith and moot the case, with the only possible sort of constraint being political in the form of Congressional threats to impeach him (...again, around and round we go...).
I don't know how quickly the Eleventh Circuit can turn the case around, but I wouldn't be surprised if they would want to just put it on ice until after Jan 20, because given the Mueller history and Trump vs. US, they can could easily bet that this political battle will be decided enough approximately immediately. If there's not enough political pressure, Trump will fire Smith on day one. If there is enough political pressure to make it past day one, then the Eleventh Circuit can start taking the case actually seriously on day two, with a completely different contextual setting.
ControlsFreak
2yr ago
I haven't really looked into all the different areas, so unfortunately, I probably shouldn't conclude much. Even with core temp, which seems like the easiest for likely simplifications, I'm not sure if we have much work to identify and quantify a mechanism for the colloquial sense of things like, "So and so grew up in Climate X, and when they moved to Climate Y, they always felt hot/cold for years until they adapted in some way." But again, that's mostly my ignorance.
For blood sugar/testosterone, I think we have a pretty good ability to 'tap in' to the inputs/outputs in simplified form that makes some sense. But there are some complicated feedback pathways that affect those sources/sinks, especially in the long-term, and I don't really know the state-of-the-art. Mostly just from what I've read from Scott on serotonin/dopamine, there's a lot of mud there.
ControlsFreak
2yr ago
Sure, there are a variety of feedback mechanisms, mostly neuropeptides of various flavors, that are involved with the production and regulation of appetite. As someone with a control theory background (hence the name), it's extremely easy to observe something that kinda looks like a feedback mechanism and then just jump all the way to, "THERE MUST BE A EQUILIBRIUM POINT NEGATIVE FEEDBACK CONTROLLER THAT TRACKS A KNOWN REFERENCE POINT!" It's extremely easy to imagine that such a thing is going on. It's vastly more difficult to understand how all the differing systems interact to produce some sort of global behavior.
For a simple example, suppose there is some subsystem that is doing a pretty decent job of doing closed-loop stabilization of a reference equilibrium point. Where does it get that reference from? Well, it's some other biological system, for sure. But which one? How? Does it interact with the environment? Culture? What? Things start to get tricky. What properties does it have? The most ardent supporters will make claims like, "Once you gain weight, it changes your reference point, and so your body 'fights' to defend it (I.e., the lower-level closed loop controller tracks the new reference point)." This is obviously plausible, but then it inspires equally obvious follow-on questions. What happens when you lose weight? Does this again change the reference point, so your body 'fights' to keep the lower weight? Most advocates will disagree. They will say things like that it's 'inevitable' that you will regain that weight, because your body is always and forever still tracking a higher reference point. What is the mechanism involved in this one-way ratchet? Ehhhhh... nobody knows?
Aside: I've also done some neuroscience-related work. Feedback pathways are complicated. Even the ones that seem like they might look simple at first. Even the ones where we can isolate specific neurons involved in what appears to be an extremely local system in simpler animals (various peptide interaction pathways are more complicated). They can be highly state-dependent; they can be highly nonlinear; they can produce oscillatory or bifurcation behavior. Anyway.
So, we could go out and try to set up some studies. Purely observational studies at the level of behavior are super challenging. You basically have just a bunch of different time series for weight. These are, frankly, going to look all sorts of different. Incredibly so if you're just taking a completely hands-off approach, just letting people do what they're going to do in life, but just weigh them on whatever interval. You'll have essentially no ability to say much of anything. So one route you could go is to try to induce a change in the hypothesized 'set point'. Maybe you make some group of folks lose weight; maybe you make another set of folks gain weight; maybe you have some maintain. Then, you let them loose and see what happens after some time. Folks are going to move around, and it's not necessarily going to be correlated. Oh by the way, what did you choose as your sample population? Did you start with obese people and have them lose weight? Or start with lean people and have them gain weight? Huge possible selection bias involved that you're going to have to somehow tease out. How do you go about doing so? Tease out environmental factors? Cultural?
Maybe, hopefully, you can also take some more invasive measurements to get at things like peptide levels, but they're swimming in a vat of all sorts of stuff going on in there, and you're kind of poking around in the dark. It's incredibly difficult to get at something along the lines of, "When you gain weight, it causes an increase in production of peptide X generally, and then losing weight again doesn't reduce such production," especially in a way that teases out a variety of other possible contributors. We've been pretty lucky to be able to see various features of a more direct intervention, like GLP-1 agonists, with regards to a few systems as a whole, but the work so far hasn't really firmed up any more general theory about 'set points'.
Interestingly, because of doubly labeled water measurements, we're actually able to do a lot more on the calories out side of the process. We have pretty good equations that do a pretty good job of estimation, at the population level, how many calories a body uses, as a function of things like fat-free mass, activity level, etc. They're obviously not perfect, and not perfect at the individual level, but they do a pretty good job. They've even told us that people do have some amount of hysteresis after gaining weight, but the dynamics (especially long-term) are about as clear as mud. But that doesn't help us on the side that most people care a lot about, which is calories in.
In the end, if you have basically no idea how the hypothesized 'set point' is getting 'set' (or getting changed), it calls into question the entire model of whether you actually are dealing with a decoupled set of systems, where one system is simply creating a 'set point' and another system is tracking it. Why do you think you have such strong decoupling? How can you demonstrate that such decoupling exists? Maybe they've actually been coupled systems all along. Maybe that coupling is weak; maybe that coupling is strong. You don't know a priori. I've seen all sorts of control theory inspired attempts at explaining all sorts of biological behavior (very different from weight regulation) in all sorts of animal models. Some are more convincing than others, and you always have in the back of your mind the question, "Are we imposing too much structure at a high-level to what is, inside, an incredibly complicated, time-dependent, state-dependent, nonlinear process?" There are a few cases where we have good theoretical reasons to think we can get away with it. See, for example, nonlinear averaging theory with applications to insect flight. But even there, we have it pretty well on the scale of tens-to-hundreds of wing beats, and we start to run into these thorny problems as soon as we try to loop in an higher-level system that includes, for example, visual observation of the environment. We can get some glimpses in some cases, but they really are sort of a case-by-case for how believable the experimental work is.
In the end, I would say that various forms of set point theory are, indeed, plausible at some level. There are huge theoretical and experimental hurdles to overcome in order to verify that it's what's actually going on. Anecdotes aren't the best guide. And even if some form of set point theory works, it mostly pushes the question back a level to how the set point is being set, be it due to inherent biological processes, environment, culture, etc. Perhaps it is some form of contaminant or something; I'm not saying that that idea is wrong. I'm saying that we have to do some hard work to isolate out different possibilities in order to know, first, what the "something" is that is being rejected, understand what the evidence is for rejecting it, and then why the only things left in the "something else" category are things like contaminants. I think SMTM is extremely casual in rejecting everything else and that to spell out the argument in detail shows how many steps were skipped. You really really can't jump directly from, "It appears that perhaps there are some feedback loops operating at some levels," all the way to, "Therefore, it must be a contaminant," in a single bound.
Literally just trying to understand your position is not "giving you homework assignments". Stop being dramatic and speak plainly.
EDIT: This is especially rich from someone who wanted to "give the homework assignment" of being a perfect-knowledge, perfect-calculation central planner of a national population. Just asking what you mean with the words you use is the tiniest mote in comparison.
ControlsFreak
2yr ago
For any onlookers, jeroboam has now blocked me rather than respond to what I thought was a very reasonable question.
ControlsFreak
2yr ago
Open acknowledgement that the prosecutorial powers of the State are being used in a conflict theory manner purely to advance partisan ends would be sufficient for some.
ControlsFreak
2yr ago
Sort of, yes? Back in the day, people had 'informal insurance' from community. If one person's barn burned down, lots of folks in their community would come help them rebuild it, "spending" at least their time to pay for someone else's loss. They community helped, because they thought that "someone else" should "pay" for it, and they were all the "someone else". Financialized insurance formalizes this and abstracts it away from individuals having to spend their own personal time to help someone else who rolled snake eyes, chipping in by a small monetary amount, in exchange for the belief that they will in turn receive the same help if they roll snake eyes.
A big part of the issue is that this formalization and separation from the community aspect, combined with terrible beliefs about redistributive government, caused folks to realize that this is yet another area where if they just control the powers of government, they can free ride and force others to pay for them while giving nothing in return.
More options
Context Copy link