He's... hard to talk about.
The critique has long echoed the old Samuel Johnson quote about being "both good and original; but the part that is good is not original, and the part that is original is not good" -- and the man has had a hatedom before 2012, so it's been echoing for a while. Most of the man's more accessible scientific writing is 'just' presenting well-established popsci into a more accessible form (sometimes without sufficient citations), while a lot of his predictive or even cutting-edge scientific analysis has to get suffixed with 'from a certain point of view' at best and 'ah yes but' at worst. If anything, that's only become more true over time: both The Golden Age Sequences and HPMoR have long relied on some of the sociology research that's be found the most wanting under the replication crisis.
Yudkowsky's been moderately open about some of this stuff, and his pro-AI, AI-is-easy, AI-is-hard, anti-AI changes have been a part of his whole story. I like that more than the people insisting they've always been right. It's still not something everyone likes, or that he can do consistently. There's never been a good retrospective on how MIRI's output was so absolutely bad on both the academic paper and popular-reader sides for so long, or the time they had an employee embezzle (tbf, not an unusual thing for new non-profits to have hit them), or yada yada.
But that's a bit of a victim of own success thing. Yudkowsky can't claim the whole replication movement anymore than he can claim the whole effective altruism one. He's at least been in the general vicinity too early to have jumped in front of the parade post-hoc, though. "Map is not the territory" and "fake answers" might have been well-known and obvious before 2008, but it wasn't until after that anyone put them together to actually poke at the core tools we thought we were using to find deep truths about reality. And these movements have been a large part of why so many of the older posts have aged so poorly, though not the only part.
((Although he's also a weird writer to have as big an impact as it seems he's had? The Sequences, fine, if good blog should change people's minds, it's a good enough blog. Why is HpMoR a more effective AI Safety program than Friendship is Optimal? Why is the Sword of Good so much more effective than a lot of more recent attempts at its take?))
... but all that's kinda side stories, at this point. Today, if you care about him, it's the AI safety stuff, not whether he guessed correctly on Kahneman vs Elisabeth Bik, or even on neural networks versus agentic AI research.
Which gets messy, because like Reading Philosophy Backwards, today, all of his demonstrated successful predictions are incredibly obvious, his failed ones ludicrous-sounding, and only the ones we can't evaluate yet relevant. Why would anyone care about the AI Box experiment when corporations or even complete randos are giving LLMs a credit card and saying have fun? (Because some extremely well-credentialed people were sure that these sort of AI would be perfectly harmless if not allowed access to the outside world, even months after the LLMs were given credit card info.) Why would anyone be surprised that an AI might disclose private or dangerous information, if not told otherwise, when we now know LLMs can and do readily do those things? (Because 'the machine will only do what we program it to do' was a serious claim for over a decade.) Who could possibly believe that an LLM couldn't improve code performance? (Uh, except all the people talking about stochaistic parrots today, and convinced that it was philosophically impossible for years before then.)
And the big unsolved questions are very important.
But in turn, that doesn't make his proposed answers better or useful. Say what you will for the ethos of singularitarity races, but at least they have something more credible than the 'you can't just tell people not to do something' guy telling people not to do something, and ultimately that's all that policies like an AI pause boil down to. The various attempts to solve morality have made some progress, despite my own expectations. It might seem like the difference between timeless decision theory and functional decision theory is just huffing fumes, but it does have some genuine benefits... and we have no way to implement them, and no way to validate or even seriously consider whether we're even looking at the most important measures. We don't know what the system they'd need to be implemented on looks like, and it's speculative (though increasingly likely) there will even be a system, and it's not clear the people building that system will be interested or even aware of the general AI safety issues.
So there's big unsolved questions that have been largely left unasked.
gattsuru
2d ago
Given that teenagers have been charged with the production, possession, and distribution of CSAM for sending nudes of themselves, CSAM charges in this case don’t strike me as anything close to nuclear, assuming the police can recover the images from Snapchat
I'm mostly using "nuclear" in the sense of "the biggest available weapon, and its resulting proportionality concerns". Those style of prosecutions happen, but they're pretty uncommon, even though there's good evidence to think the chargeable conduct happens more often than anyone wants to think about.
The only thing I’m not certain of is whether they actually broke any CSAM laws. Is it actually illegal to draw a photorealistic, but fake, image of a nude minor?
In the US, it's a federal felony under the PROTECT Act, unless the content also has serious literary, artistic, political, or scientific value. Enforcement is pretty rare, though, since the feds don't want the law to get another challenge like Ashcroft.
gattsuru
2d ago
The FAA has promoted a RemoteID system in the United States, and I think TheNybbler's commented on it and its problems before. Most modern drone receivers that fly-by-GPS have a fallback to safely land if GPS is interrupted and have some coded-in-keepout zones, but fly-by-eye or by-remote typically don't (or, in extreme cases, might be intended for use in GPS-less environments). And there's an absolute ton of old ones out there. And disabling an outbound antenna isn't that hard. And a lot of important keepout zones aren't permanent. And the software controller is largely not hard part of the drone software.
It'll probably help (and probably be frustrating) on the margins, but it's just extremely hard to lock things down that aggressively.
I think the only "EMI" weapon that even remotely fits the bill is laser weapons hitting drone batteries or other weak parts. "Some things in this room drone don't react well to bullets directed energy".
Yeah, that's probably true, and also probably something that could be a weapon of its own. I'd hoped that maybe you could treat the brushless motors like antenna, but the math doesn't really work out with modern tech.
It's worth noting that large airports nationwide have full-time teams scaring birds away from airports as aviation hazards. Would an equivalent size anti-drone team even get noticed in the noise?
Fair point, and sometimes those are somewhat hilariously aggressive -- tamed hunting falcons or dogs on one end, noise 'cannons' at the other -- though they're also all from back when we could Just Do Things.
gattsuru
2d ago
Tbf, the PROTECT Act stapled on a Miller test. They're still trying to bypass the 'prevailing community standards' bit, but compared to the pre-Ashcroft version that just pretended the Miller test didn't matter, it's a much wider retreat than, for example, US v. Lopez.
gattsuru
2d ago
I think there's a plausible false light (and defamation per se) claim, given that the images in this situation were being shared and would be themselves illegal for her to produce. Even for deepfakes-of-adults, false claims of sexual promiscuity would fall into these categories. There's some theoretical examples where a Falwell v. Hustler-style defense would be relevant in the case of a public figure where the deepfakes were clear parody, but that's pretty far from the typical case. But from a traditional law perspective you don't have to pull a Gorsuch to find a civil tradition against this sorta stuff.
Useless, though, since the kid who did it's judgement proof. In theory, the state law would allow six months imprisonment per act, but in practice that's really not how the juvenile court systems work, and even an adult doing this to another adult is more likely to just end up with a fine. And while both the boy generating the deepfakes and those passing it around (or even receiving it) could probably charged with federal CSAM stuff, that's such a nuclear option it's extremely unlikely anyone would seriously even threaten it here.
Which is part of why the whole thing is such a mess.
gattsuru
2d ago
Bestiality's a 'funny joke' because as much as people say they care about animals, they don't really care about animals that much unless they're more than a little nuts, and the possibility that someone they know might even consider it is pretty unimaginable. There was a big scandal in the furry fandom a little under a decade ago about a zoophilia-sadist ring (cw: no matter how strong your stomach, you don't want to look to close into this, yes, insert 'beating dead horse' joke here), and it got a lot of critical attention from furries (and even some other zoophiles), but as far as I can tell the only criminal convictions involved literal serial killers of animals or separate possession of CSAM. There was a lot of conduct there that was physically damaging or even likely fatal to the animal, but ultimately, it's something normal people see as gross because of what the bad actors are doing to themselves, less than what's happening to the animal.
Animal protective services aren't going to pull custody from Hassan Piker; that doesn't make putting a shock collar on a kid funny.
Beyond that, a lot of the post-1990s changes to attitudes about abuse of very young children were driven by vastly increased understanding of what psychological impact these actions had on their victims. The Breendoggle or various priest abuses had a number of different reasons they were able to shovel themselves under the rug, but one of the biggest is that it was largely assumed that victims would forget, merely not understand, or at worst become 'precocious': 'corruption of a minor' as a charge was a lot more literally considered than modern readers think. But a significant portion of human victims end up pretty messed up by stuff that doesn't leave bruises or injuries, especially when it's committed by a trusted figure.
[caveat: there's public information on this topic I'm not going to discuss here, and I'm going to encourage anyone who recognizes to not discuss here]
Any organization that can collect and process so much information on drone signals has capabilities comparable to traditional spy organizations.
That's... not strictly true, at least for lower tiers of "drone signals". You can (if you have a ton of money, and are a US citizen, and don't mind getting probed) just buy a commercial 'drone signal tracker' box and antenna kit. My guess is that the smarter ones do a lot of complex FPGA-based TDMA-like analysis work, but there's probably a few cheaper-tier ones that are just boxes stuffed full of SDRs. There are even STC'd variants for shoving into aircraft, though guessing from first principles I doubt they're very good. DJI even makes one for its specific protocol...
Which is the limiting factor: decoding traditional analog control signals doesn't tell you much about DJI's various protocols, which don't tell you about MAVLink, which doesn't tell you about some schmuck hackadayer's DIY version. Protocols are (notoriously) easy to make, and the .mil versions of a drone don't even have to use the same frequency bands as cOTS drones. Then we can throw in encryption, and you're really screwed. The only really universal response to radio control signals is triangulation, and a) that doesn't tell you much more than 'something was here' and b) are the sort of technology that has good solutions old enough to vote.
Blocking the more specialized drones that don't use radio signals gets a lot more complicated. AFAIK, we haven't seen any fiber cable drones in the continental United States, or purely-camera-driven drones... but it's a matter of when, not if.
Any security organization that can destroy, or hijack and control, drones at will, could use those capabilities against the government. Giving everyone ‘robust’ counter-drone capabilities is giving everyone (some) of the pre-requisites for throwing a coup.
Oh, I think it's much worse than that. I'd like to be in a world where the meta favors some type of tightly-targeted EMI weapon that burns out drone motor controllers -- that would still be costly to legitimate drones if misused, but mostly just drones. I don't think we live in that world, though.
Consider what a meta where physical interception with a net by a 300kph low-cost counter-drone looks like, includes, and could in the hands of a bad actor to non-drone known targets, including non-drone targets that are not considered 'critical infrastructure' but would cause tens, hundreds, or thousands of deaths if attacked. And that's still an optimistic case!
When the US opened up the US military Global Positioning Satellite network to global airlines in 1983 following the Korean Airlines Flight 007 disaster, it did not initially provide full capability. There was a policy of selective availability to globally degrade ‘civilian’ GPS signal. However, the Clinton Administration in 1996 made it US policy to provide (free) GPS access and facilitate integration into civilian and commercial applications, and in 2000 removed the selective availability policy.
Specifically, until 2000, GPS degradation was based on an additional psuedorandom delay... that had been turned off over short periods before during periods of military (during the Persian Gulf War) or civil (disaster response) need, and enough had been learned during those temporary turn-off periods that differential GPS had already made it possible to eliminate the noise factor for most applications, and both government and nongovernment orgs were pushing for an implementation. It was still a good thing that Clinton took out SA, but it's also something that was reaching the end of its usefulness as a technology already. The government implementation of dGPS/WAAS meant those higher-quality fixes remained under gov control that could be turned off with a flick of a switch if needed...
Until GLONASS was in good working order in 2011, and dual (or triple-) single-chip GPS/GLONASS/Beidou chips became the new standard (2015?). Now, there's really no way to degrade GPS signal short of just jamming it.
There are about 20,000 public and private airports in the US. Even if you took every single air defense artillery expect from every other military function and spread them around the country, you wouldn’t have enough for one dedicated military air defender per airport.
Tbf, the vast majority of the public airports are tiny, pretty irrelevant, and operate using local contractors for pretty much everything; almost all of the private ones are even more irrelevant to the calculus. To be less fair, you aren't defending the big international airports with one person, or even one person per shift.
Like, say, keeping a national policy that forces local police to become air pirates if they want to throw a jacket on a dangerous drone. Will they get prosecuted as such by government officials? Probably not. Might they get sued as such by private citizens? Maybe not. Could malefactors or ambulance-chasing lawyers sue them to try and coerce a settlement or deter an action? Absolutely.
Eh... anyone can try to sue over anything, but there's not really a lot of grounding for that case here. At least from people in this field, I've seen more concerns about second- or third-party harm (eg, taking down a drone and it landing one someone else).
At least pre-SAFER SKIES, one of the biggest issues was not the FAA or DHS, but FCC -- they really don't like anything even remotely close to signal jamming or devices that can accidentally jam stuff. For a while, the only way the FCC was letting even tests of long-distance drone signal jamming tech happen involved one of their designees standing directly behind the person holding the button down, in person. That's starting to change, but not quickly, and I'm skeptical that it can change as quickly as technology will in response. Dunno if SAFER SKIES changed that, though.
gattsuru
3d ago
The stated purpose of the programs is to teach, and to learn by teaching. Can't say we're always doing that as well as I'd like, but I don't have to seriously consider the Litany of Tarski, either.
But, yes, I'm not optimistic. I'm not even trying to solve things with doves and olive branches; I'm just hoping that having an idea of what the 'other side' even looks like at least could leave us more grounded on actual disagreements instead of several layers of imagined ones. And I'll emphasize the 'hope' on that.
gattsuru
3d ago
Fair, but at the risk of going full Diogenes and regardless of his politics and morality, the practicing attorneys are charlatans, too. The difference between him and a Kennedy School professor and Obama alum that's never practiced law is just a bunch of W-2s, not whether he's more trustworthy or even more knowledgeable.
That might seem, at first glance, like a really expansive claim. But Ken White's a practicing criminal attorney - an ex-federal-prosecutor, as he repeatedly points out to anyone who doubts his bonafides - that has written that "if all else fails, punch him in the balls" is "within shouting distance of prosecutable in the U.S.", who argued that Epstein didn't kill himself by giving a long list of other prisoners almost all of which didn't kill themselves, and argued that the Snowden charges would have required the United States government "to prove that it is harmful to release accurate information about how it is spying on us, and how it is misleading us about spying on us" (when the text White posted of the statute showed they could just prove harm from any of the literal thousands of other specific things that Snowden had already been known to have divulged).
I'll point out that last one not because I had to dig that far back to find a third example (and believe me, it's tempting to just go reverse down his timeline) but because it shows he was fucking it up then. Worse, everyone else -- ClarkHat, Glenn Reynolds, lawfareblog, several actual practicing defense lawyers, me -- bought it. Hook, line, and sinker. Because we were naive, or stupid, or because we liked the politics, and the morality, and the whatever. And at the risk of stretching a metaphor, it's a claim that's equivalent to an IT guy that thinks you use rollover cables to connect two switches to each other, or a pilot that thinks you steer with the yoke on the ground. Maybe just a little concerning if it were an intern's first day on the job, but not the sort of mistake experts should get to make.
There's a fascinating philosophical question about whether he's intentionally turning the stupid on and off like a mask, if he's some pretense of competent when he's at his day job, or whether he's an idiot savant who reverse-rainmans himself by giving shitty legal theories in such a compelling way that observers are persuaded to agree with him. But since he's also the man who codified the Rule of Goats, I don't have to care. Nor is it just him, even if he's a particularly easy one to document.
But worse than that, it's not like anyone in the real world acts like any of these guys are the charlatans they obviously are. William Baude's at the nexus of both stuffed-shirt 'qualifications' and no-neck academic onanism that you justly criticize the Volokh sphere. The paper he's going to be famous for the rest of his life used Vallandigham as an example he picked to explain why the 14th Amendment required removing names from ballots, without ever noting or disclaiming that Vallandigham ran for Senate after the ratification of the 14th Amendment with his name getting on the ballot. No one cares. Not even in a kayfabe sense, where okay MSNBC is bread-and-circuses, but the important resources of governments and courts just throw this in with the pro se sovereign citizens. Ken White wasn't telling everyone Baude was a conman; no one in the Volokh circles cares (even today after Baude lost) that he's a putz. And it's not just the normies or the talking heads who overlook that these credentials are useless. Hell, I brought it up here with AshLael and ymeskhout - a real public defender - and no one of them said 'oh, sure, he's a swindler, why do you care about the Yalies' in response.
gattsuru
3d ago
The Measurements thing is pretty common, but I'll caution that mainstream (even 'mainstream' in the sense of the Playboy 'I can't believe it happened to me' sense) filters male-focused sexuality a lot more aggressively than women writers, even around the same kinks. Tamora Pierce's weirder age gaps predate the real heavy norm shifts in the late 1990s, but there's reason I keep bringing up the Blue Is The Warmest Color problem in romance; it's something you have to actively avoid and block the further you get into fandom spaces. But even well outside of the squicky questionable or overt absolute number age problems, male-written fantasy like what you're motioning around is not just possible but pretty common in fandom spaces.
((And it's just an outright standard trope for M/M stuff, whether the writer or artist is a top, switch, bottom, or straight guy.))
There's just a lot of forces that make sure that sorta writing doesn't come from a man in modern publishing, from the general pressures against men writing romance-heavy stories, the heightened scrutiny if they do, and the wider pressures against hiring men in those roles at all.
Trying some recipes again:
- Savory puff pastry rolls. Dice and partially caramelize a small white onion, about 10 min high heat, 20 min on low. Toss in and brown at high heat a pound of ground / minced lean beef or turkey, drain off any excess grease. Add salt, pepper, garlic powder, chili powder, and cumin to taste, optionally throw in a tablespoon worchestershire sauce or throw in and simmer off a quarter-cup of sherry, port, or red vinegar. Once crumbly and well-browned, turn off heat and immediately transfer to a plate with paper towels on top and below. Prepare two puff pastry sheets, unroll, but keep them as close to fridge temperature as possible. Scoop half the meat-and-onion mix in a flat line along one of the short edges of each puff pastry sheet, at least a half-inch in, and spread it shallow. Apply a line of 2- to 4-ounces of goat cheese, sharp cheddar, or pepper jack cheese, very shallow, along the same axis. Roll each sheet back up. Cut each puff pastry roll into pinwheels, six-eight per roll. Place onto parchment paper on a baking sheet, apply an egg wash. Cook at 450 F for 25-30 minutes. Serves four to six people, best served with a hearty salad, roasted vegetables, and a chutney or mustard sauce. You can put these into hand pies or more complicated shapes, but the roll works a lot better imo and are much easier, even if it's a bit messier. It's not quite as interesting a result as the chicken-and-cheese-and-apple pie, but it's been a lot more acceptable to most other people.
- Bobotie (sp?). Ground beef/turkey, onion, spice as above, toss in some curry powder to your preferences. Add a half-cup chutney or jam, four slices of white bread torn into tiny pieces, and either a cup of unsweetened applesauce or a cored and diced large apple, optionally a half cup of raisins, dried cranberries, or dried cherries. Layer into a baking dish. Bake at 350 F for 30 minutes. When that's almost done, mix three eggs, a cup of half-and-half (or a half-cup of cream and a half-cup of milk), a tablespoon of tumeric in a bowl, and beat the everliving shit out of it. Pour on top of the meat in the baking dish, cook for another twenty minutes. Most Official Recipes say add bay leaf, but even as someone that loves bay for rice, it doesn't seem to do much here. Serve with rice and some greenage, but expect three servings for most eaters.
- This recipe modifies to work in a rice cooker surprisingly well -- I basically just pan-fried the chicken in its marinade instead of partially-cooking it in an instant pot -- though I don't think it would have passed an authenticity test. Still pretty good, and outside of the (looooong) marinade time, a lot easier than I expected.
- Lazy man's burritos mix. Pan-fry a pound of boneless skinless chicken, allow pan to cool, then dump in drained pre-cooked kidney or pinto beans and a cup of salsa (salsa verde and peach salsa work weirdly well, but even your standard gringo salsa does fine). Simmer for thirty minutes covered on low heat, adding water if it starts to dry out. In a separate pot or a rice cooker, cook three cups rice in four cups water without rinsing the starch off, drain, and then dump the chicken-salsa-bean mix. Mix aggressively, let sit and cool for ten minutes to let the rice absorb as much as it can and get as sticky as possible. You can go really nuts with this recipe in a slow cooker, and it's great and gets rid of the pan-fry step (and can let you use bone-in chicken)... but it takes six hours and a crock pot. Either way, serve on a tortilla or lettuce wrap with iceberg lettuce, a sharp cheese, top with sour cream or greek yogurt and a drizzle of the remaining salsa.
gattsuru
4d ago
Partly, I'm just incredibly petty. Partly, because Gerstein's pivot to Project Veritas is unusually naked and galling - even as a defense of 'principles', both Gerstein and his cited experts quickly pivoted to defending the Times in their attacks on Veritas. Gerstein himself would later imply that Veritas only avoided criminal prosecution thanks to Trump winning the election, and it's possible he's not even wrong, though I'd need to check deeper before I could commit to it.
Partly, because writing up this stuff helps me remember the event. Partly because I've got some port I'm trying to work through before it goes bad, and while I can't drink enough at once to get really drunk, last night it was apparently enough to get me spicy. Partly, because if I don't write up un-newsworthy stuff, no one else will.
((and, I will admit, partly because I can't resist an easy "own", even when it degrades my writing. Can't claim in vino veritas as a defense on that.))
But mostly because, while Gerstein doesn't matter, the system does, and you can't meaningfully talk about a system without mentioning the completely replaceable gears. If I'd sed -i 's/Gerstein/Politico/g' that post, it'd have gotten criticized writing a smear; if I'd done them same to switch Gerstein's name for the entire leftist sphere, I'd be inviting moderation action, probably correctly. It's only that he's going to keep getting supported, employed, unarrested, and unmolested that even lets me make this into a broader critique (and, coincidentally, leaves a testable prediction... and one I've very, very, very seldom had to eat crow on.)
((bonus prediction, SHA256 for three days: 7d4ed1e8c28b93c95e7b533870073df64ca533c9174589124b5b145fd818a611.))
It's not very productive to unload both barrels at a particularly squeaky replaceable gear, even if it can be a little cathartic. Doesn't mean the system's in good shape. This guy shows a lot of the ways its defenses aren't in good shape.
Josh Gerstein is a Terrible Legal Expert
A man wrote a bad post on twitter:
This isn't so much a statement as it is a bunch of Implications given current events, but whatever extent it's implying anything, it's worth spelling out that it's fractally wrong. Minnesota is not a Stand Your Ground state, Stand Your Ground laws don't allow anyone to just blammo someone for knocking on a front door of a public business, there's no bonus tag on amateurs, Minnesota has an unusually constrained castle doctrine, so on, so forth.
At his claimed defense, this is just snark about his imagined false beliefs about the law resulting in the deaths of his hated enemies. False beliefs that Gerstein has actively promoted, and which he does not correct here even twenty hours later to dispel, taking full strength "90% of posters quit digging right before they're about to get out of the hole". Which would be funny, were Gerstein just another incompetent internet lawyer: it's not like I'd have space to complain just because someone's wrong, there's a surfeit of people wrong on the internet, and I'm not exactly bar-certified myself.
But no. He's an expert incompetent internet lawyer. Gerstein is, as his bio helpfully points out, a Senior Legal Affairs Reporter at Politico. He doesn't need X payouts to be compensated to be this wrong: he's a professional!
There’s things to be said about trust in expertise, but Ive said them before, on clearer matters; there’s little I can add to his presentation than shouting “the Cognitive Elite” at the end like the punchline to an Aristocrats joke, and that’s applicable nearly everywhere anyway.
To provide a real steelman, he doesn't want me dead. He doesn't even know me! He might even pretend to be apologetic were Shirley shot tomorrow. It's just be hilarious, the day after.
((Although it's worth spelling out that staff writers for a local media org published by a man with fraught ethical ties to the current Minnesota Governor are very interested in the identities of certain 'citizen journalists'. Just "reserving the right", anyway?))
... but that's not exactly news. Indeed, this stuff isn't newsworthy. Gerstein might feel proud about 4.8m (and counting) impressions of being a dumbass, but the reality is no one that matters cares about his entire existence except as a way to make the demons from Frieren look sympathetic and multifacted; if his computer blew up and he was stranded on a desert isle without internet access, there's a hundred thousand Vox wannabes chomping to take his seat. Brendan Eich got decked in a public space by someone specifically targeting him for his political positions a decade ago, you didn't hear about it at the time, I can't find any reporting now, and the only reason I know is that the guy who did it was happy to tell me. A guy who celebrated a political assassination took an NYT slot to debate the corpse, and he's better-known now for (allegedly) electrocuting a dog (cw: music, and be glad that I didn't link the furry version). Ken White has gone from jesus christ territory to full "who needs LLMs when the real hallucinations were with us the whole time". He can't even get an intervention, or for whatever shrink he's paying to pay attention to where his political screed hit actual madness; for everyone else, they're just noise coming from that thing in the corner. There's no shortage of these things; I'll just throw out the link and save the scrolling for everyone else. Even for people here, fish don't think about water.
Ken White is a lawyer in good standing. Punch-happy programmer, afaik, works at a FAANG. RPGnet isn't getting the ARFcom treatment, or men with no sense of humor that they are aware of asking for IP addresses. Discords and fundraisers where people call some political assassinations a good start don't have Harvard-supported orgs delivering hacked personal information to Fox News.
And Gerstein isn't going to lose his job over this. This is his job; he's a Legal Affairs Reporter not someone who actually needs to know what the actual laws are or who to ask about those things. There's no one who can boycott Politico who could care. It's certainly not going to end with fascist jackboots dragging him out of his door at 6AM in his underwear, and the mere idea, were it even imaginable, would be enough to unify a massive array complete with lawyers and judges with little interest or attention for technicalities and no hedging about bad behavior.
That's not a concern, for some.
gattsuru
5d ago
You'll see minheaps a bit under Dijkstra's algorithm in the networking and -adjacent spheres, but it's so well-known as the Default Solution for those use cases you're either going to have it under several layers of abstraction, or you're really doing something weird. I've had to hand-build it once, and I don't think the end product ever actually hit anywhere outside of a toy environment.
gattsuru
6d ago
To a large extent, these tools already exist. They're just limited: SCAIL struggles for movement paths with more than three characters or over nine seconds, ControlNet Pose has to be tuned for each model and sometimes even each finetune, and LoRA can uniquely handle three or four style/character/event/motion per output before they start getting funky interactions.
But even assuming that these problems can be fixed - plausible, but not a given! - there's a fundamental tradeoff between what you let the model do, and what you don't. Sometimes expressed as a double! And still hard to manage.
It's a little difficult to comment with confidence yet, just because there's something like (im)plausible deniability for at least some of the locations, still. Especially near the holidays you'd expect a child care business to be very feast-or-famine, it's not always obvious how closely a given 'business' is tied to a specific number of hours, and a number of violations during an inspection can 'just' point to a small and new business. Even the multitude of businesses with a shared address could mean that the owners are operating out of their homes and have an office park PO box to handle mail -- that's not even particularly unusual for actually-legit service-oriented small businesses.
But there's a lot of stuff that stinks to high heaven. At minimum, compliance had to have completely skipped most of the steps and processes that a normal child care agency had to go through. Even where it's 'real' in the sense that they're doing child care, some of it's probably not real in the sense of paying the claimed fees that justify the various grants and subsidies, and most of it's almost certainly not 'real' in the sense of complying with the long array of standards and regulations.
There's a non-zero chance Shirley ends up facing charges, here, which will be one of the funniest possible endings. ((Of course he's serious, and stop calling him Surely.)) There's a lot of rules about creeps filming kids, with reason.
There's also a >95% chance that there's some org or orgs has been actively farming these businesses or 'businesses' up in exchange for a cut, and is totally within the bounds of the law. Probably has extensive documentation that they cleanly and clearly described each and every regulatory requirement (to people who didn't understand them). The really fun question is how many of them are making political donations. But at best a bunch of particularly shameless small fry might fry; none of the people in the government who should have noticed that Line Went Up will lose their jobs.
Looking at the court records, the sentencing's probably in part due to the 'endangering welfare' bit -- it's not in most of the reporting, but there's pretty overt evidence of both neglect, and directly handing the kid to an incredibly drunken guy who still was more concerned about the child's welfare than the alleged parents.
That said, if I've got the names right, the difference in sentencing may just reflect Urban already being on probation for "receiving" a stolen bike (... I'd be dollars to donuts he just stole it) and a previous domestic violence arrest. Ehlers had a shoplifting arrest, too (and maybe jumped bail?), but that's not going to weigh as heavily in most cases, even ones as weird as this.
I can't see anything about the disposition of the kid, but I'd be very surprised if the kid didn't go straight to child protective services. The six-pack of beer wasn't really part of the whole 'adoption-for-sale' gimmick... but it was part of loaning the kid out overnight, instead. Optimistically, that's just an incredible level of neglect by a man and woman too dumb to realize that concerned bystanders were trying to get the kid into a safer location... but pessimistically, they might just not have cared why people were wanting to borrow a 2-month-old baby for one night.
gattsuru
7d ago
Something like SCAIL and LoRA abuses can probably do that today and is probably already getting used in that sense today, but the current version of the technology goes a little nuts for segments longer than 9 seconds, and it's painful to do even short segments using the existing workflows, on top of being egregiously slow on consumer hardware. I've seen people take it into a couple minutes by doing really aggressive generation of prompts to make a flipshow to start with, but anything longer than that tends to either end up needing to compromise on weird physics or ugly scene changes.
And the current implementations have some limits; pose info can't do talking heads well, going beyond three characters with pose info gets rough, and some particular pose changes can go full-on Exorcist. SCAIL's lipsync capabilities are worse than WAN animate, and while it's possible to combine them, it's even more finicky.
But compared to the cost and unpleasantness of traditional mocap, or even makeup? If you can possibly use this tech, there's a lot of good arguments in its favor.
gattsuru
7d ago
There's some gay guys (and arguably the entire sperm donation industry) that work on paternity roulette logic. Even for gay guys where it's just so they don't really have to think about who's the 'real' dad, though, it's kinda messy, and not just literally. These days, you can figure out the answer for a couple hundred bucks, obviously, but even if everyone involved credibly commits to never doing that (and the alternatives aren't obvious), it's just denying the questions, rather than actually handling them.
These things all have answers. Especially for soccons who care the most about this stuff, there are sometimes even doctrinal answers, but even most gay guys who only truck with the church when nailing complaints to the door have pretty good ideas about what they wished their fathers had been like. Dropping the odds to 1/6th only really gives an excuse to forget about or delay obligations and responsibilities, rather than making them actually not exist.
gattsuru
7d ago
... I philosophically prefer surrogacy where the donors stay in the picture, so caveat that I'm going to be biased in favor of donation, here. That said, potential problems:
- You've already discussed your side of the relationship woes and you've got a much better idea of what they look like than I can guess, but they are a pretty important thing.
- There's a lot of messiness with lesbian/bisexual woman politics, because there's such extreme potential for jealousy, and because a non-trivial number of bi women do either get out of college or just randomly sort into het relationships. Unless she's routinely seeing a guy on the side before you, I'd honestly say you've dodged a bullet not getting your dick wet, here, but if you do this, you can never be just that friend she'd never consider again, either to your own wife, to her wife, or to her. Doesn't matter if the only thing involved was a jar and a turkey baster. You don't have to and probably shouldn't go full Pence rule, but you still should be aware there will be new eggshells around.
- You can't really sign away parental rights/responsibilities; courts routinely compare a child's interests against contractual statements and throw the paper away. That's unlikely to come up, given the background you've mentioned here, but it's potentially very expensive -- and worse, may be something you'll constantly be weighing when considering things like offering to babysit the kid even if you ultimately decide to help out. Informal donation provides less protection, to my frustration. Divorce or death are the 'obvious' sources of problems here, but even something like a surprise illness can end up a big question mark pointed your way.
- ... but you can kinda sign away parental rights, and the couple really should insist on you doing it, and there's a point where that's gonna hurt and you're gonna have to bite your tongue. Maybe the breadwinner of this couple gets a job in another state or country and you go from seeing the kid once-a-season to once-a-year, maybe once a teenager the kid gets into hobbies the parents are okay with and you aren't (or vice versa!), there's a hundred different possibilities. You will be, at absolute best, Uncle Guy. Some guys can handle that perfectly fine, some guys can handle it for daughters but not sons, some guys don't even see how it could be a problem, but it's not an obvious problem until years down the line.
- Conversely, if you do become Uncle Guy, you might find that people you're fine seeing once-a-season are really obnoxious to see once-a-week. (Or even really obnoxious to your wife for them to be good friends with her, and nearly-family with you.)
- It's harder, though not impossible, to get genetic screening done through informal donation. That may or may not matter to you, or to the couple; it can even matter for different reasons for each side of the equation.
All of that said, I've seen it work out perfectly fine for a good few people, and not in the porn premise (or polyamory) sorta way. The problems are downstream of you not just getting a kid, but a whole set of informal relationships, but those relationships remain when good things are happening, too.
gattsuru
7d ago
I've pointed to Cash's Hurt and Air Traffic Controller's Blame, and I'll point to them again. Fastball's The Way is... reeaaaaally fucking dark if you look up the backstory, but it does make the song a little bit more poignant.
For video games and anime, I'm a bit of a basic bitch. Gurren Lagann's Libera Me From Hell and FLCL's I Think I Can aren't ultimate songs -- I'd say not even the best songs from their respective shows -- but the ethos they describe and how they reinforce the themes of the shows are extremely powerful. The Chrono Cross soundtrack is from a pretty meh game and there's nothing special about Scars of Time beyond just being good, but I'd listened to it on repeat a lot in a specific time and trance state, so I don't think I can resist reacting to it now.
gattsuru
9d ago
He's also just really scuzzy, even by the low standards of Ohio politicians. Musk can make these sort of arguments stick by pointing at one of several giant factories. Vivek can point to... failed biomedical stuff in an even-more-scammy-than-normal field, a politically-focused investment fund, and partial ownership of BuzzFeed.
None of it's clearly fraud, or illegal, or even likely to get the New York AG going after him, but I wouldn't be surprised if he loses more Red Tribe support than Jay Jones lost Blue Tribe support.
gattsuru
9d ago
Laurenson had no X commentary about either the Jay Jones scandals nor the Charlie Kirk assassination, and as far I can tell, did not anywhere else. Perhaps I’m missing some counterexample, or some massive story from her media deal.
The movement she’s left has changed. It’s not the only thing failing her tests.
gattsuru
10d ago
I'll start off with my condolences to anyone of any gender that had to listen to nRXs for this to happen at the end of it.
I don’t spent as much time in the movement these days; it’s turned into an echo chamber; racism and sexism are front-and-center, and most of my favorite people have left. But there was a fascinating moment in time when it truly felt like the so-called “dissident right” could discuss almost anything in a sensible way.
... I have a thought. That's not fair, I recognize... though I'd like some counterexample showing I'm just cherrypicking.
Talking about anything in a sensible way is not the point of the story; talking about why no one can group can talk about anything in a sensible way isn't part of the story. Laurenson isn't writing for Red Tribers, obviously, nor writing to Red Tribers, or even explaining Red Tribers to Blue Tribers. The point is to explain where Laurenson's coming from to Blue Tribers, and that's fine.
But from a Red Triber perspective -- and I fully recognize I'm far from hydroacetelene's level of 'real' Red Triber -- it's kinda the last part of this story where there's anything interesting. The Now What section is the biggest frustration, since it starts with 'here's the Blue Tribe principle insulting me and here's the Red Tribe random asshats insulting me, tots similar in scope and regularity' and then leads to a trio of revelations that practically come with the punchline 'do you think we don't know that?' But does anyone think there's a literate Red Triber on the planet, and I'm defining 'literate' here by Chicago definitions, that does not already know that a news media environment with any mix, no matter how lopsided, of Red and Blue Tribers devolves into squabbles?
I can make the argument that deradicalization matters, I can and regularly do make the argument that liberalism is dying at this rate, I can and have made the argument that it's really really dangerous. I'd like to solve that! I've been doing the (sometimes literal) Touch Grass thing, and some STEM-focused community outreach, and a half-dozen other programs trying to bring people together without bringing politics to the forefront.
Online, I'm overtly the bi furry gun owner for a reason, knowing how offputting the constant gay or gunnie references are to so many people: areas I'd like to go and philosophies I'd like to let live become 'target rich environments' if the only ones who wear it on their sleeves are the Everything Leftism Coneheads and straight-from-central-casting . I'd hope that there'd be some impact from showing people are people, and that Red and Blue can join together to achieve goals more important than wars of all-against-all over microns of worthless territory, and that at least some goals of those politics are more than hate or rage or malice.
... but I don't know if that will work, either, and ultimately, the Litany of Tarski wins. The STEM outreach program's had a small and subtle cold war over rainbow pins in the local community, the modded Minecraft server I helped with software problems had its owner proudly promote the time he beat the shit out of Brendan Eich, there's many Blue and Red tribe spheres I can't wear anything on my sleeve. Worse, so many high-profile people pretending to straddle political aisles are very clearly not that there's less than zero trust, here.
Everything else is a distraction. Yes, people are hallucinating their own consensus realities, but (as Laurenson points out!) it's not like the normies are doing any better. If we can't even talk about anything in a sensible way, if we can't talk about why we can't talk about anything in a sensible way, trying increasingly complicated and roundabout messaging won't solve it.
Given that they've caused less damage than a single particularly retarded-as-in-dropped-on-his-head arsonist, I'm not convinced that they've been optimizing for damage, whatever thought they've put in. I'm... actually a little unconvinced that they've even optimized well for disruption, but there's a lot of reasons I don't want to talk about that publicly.
Which is one half of the problem in talking about this stuff. If there are red team exercises that can up the high score from the known alternatives, it's... not a very good idea to start talking about them at length in public. The other half's that if you have reasons why a given attack shouldn't work, it's not a very good idea to talk about that at length in public, either.
More options
Context Copy link