@Shrike's banner p

Shrike


				

				

				
0 followers   follows 0 users  
joined 2023 December 20 23:39:44 UTC

				

User ID: 2807

Shrike


				
				
				

				
0 followers   follows 0 users   joined 2023 December 20 23:39:44 UTC

					

No bio...


					

User ID: 2807

The point I am trying to get at as regards executions is that appearances are important, and executing people in secret is a bad appearance. Even if you don't think people's executions should be public, you shouldn't want witnesses removed. Even if you can secretly torture people without leaving marks (which yes, we can) you should not act in such a way as to needlessly arouse suspicion that you are doing so.

I don't know. I based it on your reply to me here, where you acted with incredulity towards the assumed legitimacy of the targets of Nidal Hasan.

The problem with Nidal Hasan's target wasn't that he targeted soldiers. Iran has killed somewhere between a dozen and two dozen soldiers in our war with them and none of the American commentators on here have acted like that was illegitimate in the same way that Hasan's was (or as far as I can remember, at all) because Hasan is an American who turned his gun on his countrymen.

I don't think you can point me to a single star-spangled flag waver on the Motte that has argued that Artesh soldiers, acting in the normal course of their duties, should be executed for the deaths of US soldiers. I think you are mistaking your own ideas about what Americans think with what they actually say they believe. Perhaps you have had other experiences off of the Motte that I can't speak to, but what you are describing of your experiences here does not comport with what I recognize.

acting like a military base is a 'place of trust' and that the military personnel therein where defenseless, and therefor the attack was cowardly.

If you are a soldier and particularly an officer, then yes, a military base is a place of trust, and generally stateside on a military base US soldiers are disarmed, and will go to jail if they carry a weapon on base. (This is a sorry and facile state of affairs, and I think the current admin may be working to change it, which is good).

I am not sure I would say Hasan is a coward but shooting unarmed people is not generally what is associated with courage. I don't think Hasan acted much differently in terms of personal courage than any mass shooter, in the sense that he was brave enough to accept being shot by cops but not brave enough or willing enough to accept defeat to engage armed people from the start. I think that his target set could in some sense be seen as more "legitimate" than bin Laden's in the sense that they were primarily military, but also that his infamy should be greater because he betrayed a position of trust and violated the oath he took.

Or did they kill close to half a million civilians through war campaigns across the middle east?

Did American troops kill close to half a million civilians during the Global War on Terror? No, they did not. The majority of civilian casualties throughout the Global War on Terror were not caused by US troops. I suspect you know this, which is why you used the phrase "through war campaigns" thus absolving the people who actually literally did kill civilians of their actions and rhetorically placing it on people who were innocent of their deaths, while at the same time providing a defense (a motte, if you will) if someone pointed out the inaccuracy.

I think it is completely fine to criticize the outcome of the Global War on Terror (indeed, I have done it myself) and to say that the GWOT led to hundreds of thousands of deaths. I think it is wrong to completely ignore consequentialist reasoning. But saying that the United States killed half a million citizens across the Middle East is no more sensible than saying Al Qaeda killed hundreds of thousands of Middle Easterners through war campaigns across the globe – arguably less sensible, since after all Al Qaeda attacked the States.

It's actually quite perverted to murder hundreds of thousands of people and then write a book about how horrifying that was. And then make a big deal about how bad you feel about that.

Off of the top of my head, there are only two Americans who could reasonably qualify, and I don't think either of them felt bad about what they did, or ever pretended to.

I think it's your ancestry. If you don't want to engage with the topic, fine, but these assertions are inane. A person born of two Palestinian parents is Palestinian. He proved that with his actions as well.

Very well. I shall expect you, then, in the spirit of self-consistency, to refer to Trump as a European leader going forward :)

Words are wind.

And to this point, Hasan did take a specific loyalty oath by virtue of being an officer. But at a certain point this is just a generalized argument against citizenship and even ethnic or kinship bonds. The traditional way of dealing with traitors (ethnic or civic) is to hold them contemptible, not to simply be built different and never have that problem.

I think it's pretty hard to argue that 'the American tradition' doesn't treat citizenship as a mere piece of paper when it literally gives it to the children of foreign criminals, allows its citizens to hold an unlimited number of other citizenships and doesn't have any substantial requirements to give it out other than 'you've lived here a while'.

And if you say citizenship is just a piece of paper, it's hard to argue against this state of affairs.

And that's without getting into the fact that for most of the country's history, it was illegal for citizens of one race to marry citizens (or chattel) of another race. That racialised understanding of the nation is the American tradition.

This is not really true, as a blanket statement; anti-miscegenation laws were never universal throughout the United States.

And it is declining at the very same time that the number of Americans who even like America is also declining.

This isn't true either – if you're going to lecture me about my own history and tradition, please try to get it right.

To speak in very broad terms:

  • Each state had their own peculiarities in terms of how they dealt with Native Americans and blacks. It's certainly true that ethnic feeling was commonplace, but there was never some sort of universal understanding that was codified uniformly into law throughout American history, and (obviously) anti-slavery and, yes, what we might today call anti-racist sentiment was not uncommon in the United States even before it attained nationhood.
  • I think it is wrong to model the halcyon days of Jim Crow racial restrictions as a continuation of unbroken American tradition from time immemorial. It can better be understood as a shifting attempt to control the black (and Native American) population – and in the specific post-Civil-War context of Southern Jim Crow laws, as resentment against the Yankee conquerers. This might seem like splitting hairs, but my point is that there was not (for instance) some a prior first principles deduction that blacks were inferior and so should not be educated; rather there would be a slave rebellion or a perceived problem with runaway slaves forging passes that would lead to bans on slave education. Or (to illustrate with a non-Southern example) black freeman could vote in Pennsylvania until 1838, when the Constitution was revised to specify "white" as the black population in the state and racial tension increased. Thus, from my understanding, it is best to model American racial restrictions as being reactionary instead of foundational. [Edit to add: I would say "often best to model..." – because, to my point, there's not really a one-size-fits-all American tradition here.]
  • To the extent that there is an American tradition of a racialized understanding of citizenship and civic belonging, it has always been contentious, and it went into decline, not now, but rather during the Cold War. I agree that it's not unrelated – the Civil Rights era and Vietnam were a very divisive era in American society in part due to racial issues – but that was decades ago, not some recent development. What America is experiencing now more greatly resembles post-Reconstruction backlash (sentiment towards diversity is cooling, not increasing!) than some sort of terminal decline of ethnic sentiment, although things are different this time around.

I think the term 'multi-ethnic nation' is a contradiction in terms.

Then America, Russia, China, the United Kingdom etc. have never been a "real" nation (and to a large degree, at least in the case of the United States, which I know best, I actually agree with this statement, albeit for somewhat subtler reasons).

I don't know the details of each state, but I think that something like what you are describing is typical. I believe there may be other outside witnesses as well.

The model of citizenship you describe is clearly incompatible with birthright citizenship, which involves handing out citizenship to babies of (both legal and illegal) immigrants.

Hasan's parents were, as far as I can tell, legal immigrants and became naturalized citizens (a bit unclear to me if that happened before or after he was born), so opposing birthright citizenship for illegal aliens doesn't really "solve" here. Giving birthright citizenship to the children of citizens is not much, if it all, different than the model you seem to think would be preferable, of being born into an ethnic group.

However, I don't disagree with the criticism. If I was King Of The World, citizenship (and adulthood, for that matter) would likely be earned by deed, not birth.

Describing it as just a piece of paper is simply being honest about the world.

I don't want to live in such a world, nor do I agree that that is an honest description of what citizenship has long been understand to entail in the American tradition, so I decline to follow your lead here.

I would like to do better than news headlines!

If America truly is a 'proposition nation' with no ethnic requirements to be a member then pretty much. if America is simply a (very successful) economic zone where citizenship is just a piece of paper that grants you access, like a software key, then America isn't a nation at all.

I don't think these are the only ways to think about how citizenship and country work. There's an older understanding of citizenship that is more akin to being bound in a sacred pact – like adoption. As far as I can tell, ethnic groups all have some way of bringing non-co-ethics into the "in-group" ethnic circle. If you think Americanness should be an ethnicity, when is an idea that I am sympathetic to, then citizenship is that ritual.

I would argue that Hasan is 100% a middle eastern terrorist, even if he is also 100% a US citizen, because all citizenship is just a piece of paper.

This is precisely what I think we should all deny. "Citizenship is just a piece of paper" commodifies and cheapens citizenship, the opposite values you should be pushing if you want people to think in ways that are transcendent instead of being non-transactional.

Hiding it is really the worst thing you could do. If we're that ashamed of it we shouldn't be doing it! My aesthetics demand we embrace the brutal reality. Just like how I think if you want to eat chicken you should know what it's like to slaughter one.

I tend to agree with this. In the United States, we're very far removed from death.

However as a guy who has slaughtered hundreds of chickens I can tell you that if anything it makes you less sympathetic to their plight.

He's legally an American, yet clearly holds no bonds of brotherhood and fellowship for the rest of us. Instead, his true tribe is co-ethnics and co-religionists from the other side of the world

Let's set aside the degree to which is tribal affinity with co-ethics and co-religionists on the other side of the world is a deadly LARP and agree that this is true: claiming that he is Middle Eastern obscures this and the problem (which exists to some degree in America and from what I can tell to a much larger degree in Europe) of second-generation radicalization.

If you think it's bad that Hasan doesn't consider himself an American despite being one, you shouldn't confuse people into thinking he's not American by referring to him as Middle Eastern. You should explicitly highlight exactly what he was, and highlight how it came to be that he continued to serve in the U.S. Armed Forces despite expressing affinity with America's enemies openly.

to the point where he was willing to commit mass murder about it.

As we've seen in this very thread, calling him a "Middle Easterner" seems to make it easier for people to believe that what he did was not mass murder but rather a surprise attack on enemy combatants. It casts him as someone who was true to his deepest self and therefore in some sense admirable rather than someone who betrayed his most sacred confidences.

He's not a Middle Eastern terrorist. We can say that he behaved like a Middle Eastern terrorist, or that he shared their sympathies. We can argue that this is true in some spiritual sense, if we want.

But ultimately he was born as an American citizen to American immigrants who naturalized, and despite this he found that his sympathies lay elsewhere, to the degree that he was willing to kill his fellow soldiers out of misplaced affinity towards a group of people who were never really his own.

To claim that he is Middle Eastern is to deflect scrutiny from how he came to be radicalized.

You could just as well torture them and execute them via firing squad.

Generally speaking, when people are tortured it leaves marks that witnesses to executions will notice.

In either case, people don't associate 'the west' with public executions. But rather the modern justice system that came after.

Who's "people"? I'm a Westerner, and I do. I have been in favor of public executions since I was in high school or before, in part because it is part of the Western, Anglo-American tradition.

I do understand the push-back against live-streaming it and the concerns about making it a spectacle, or that it might be a bad idea in this specific case. Those arguments track with me. But the idea that "here in the West we don't execute people publicly" doesn't, because it's not true.

you and other pro-America guys on here. Who ultimately draw the distinction between good and bad, with America on the good side of every conflict. And they don't see that as unreasonable or questionable.

Do you really think this is an accurate description of me based on my posting history?

Even if you are confusing me with Shakes, Trump's Own Strongest Soldier, I doubt this would be an accurate description.

I get it from interacting with Americans.

Well, I think I am beginning to understand why. It should not surprise you if even the most reasonable and open-minded of people will become a raging nationalist when you paint people of their nationalities with a broad and unflattering brush.

You should read the authors I mentioned. They are emblematic of what Americans read: Clancy was a tremendously popular novelist during the Cold War with the American right; Wright wrote a popular and interesting account of the path to 9/11; Hersey's writing about Hiroshima is, I believe, commonly taught to American students in grade school; and Fritz wrote in the Atlantic, one of America's most prestigious publications.

With two Palestinian parents, I think you are a Palestinian.

His parents were Americans also.

Clearly, calling this man is a terrorist, that's not a partisan statement.

The definition of a "terrorist" is famously almost always a subject of contention. I don't particularly mind applying it to Hasan.

he clearly did not consider himself American in a meaningful sense of the word.

Oh, so you think that if he doesn't consider himself an American, then he's not? Interesting. Do people get to pick whatever citizenship they want, now? Does this mean that someone who does consider himself an American is? Maybe the whole world is American (except Nidal Hasan, who is a "Middle Easterner")? I think the answer to all of these questions is "no." So Hasan is an American.

I understand the idea of being (let's say) "spiritually American" or saying that Hasan is "not a true American" – very reasonable rhetorical statements – but this guy had the rights, privileges, and duties of an American, by virtue of being an American citizen. He is, legally, an American citizen. And in terms of his behavior, he deserves to be treated as one, because to do otherwise cheapens his crime.

If you think that citizenship is not something to be cheapened by importation and dilution but rather a sacred bond held between fellow-citizens with attendant rights and duties, then you believe in holding people to that.

So when I use the term wrath in this context we could conclude that I am talking about the sinful version of anger.

Not everyone is aware of the distinction, though.

There's nothing secret about publicly sentencing someone to death in a court of law, killing them in a manner not publicized and then announcing that they have been executed.

This gives rise to speculations that the condemned was tortured the death, or not killed at all. Both are bad.

Which coincides with decreases in public executions and capital punishments along with a host of other social revolutions.

Actually, at least in England, the post-enlightenment period lead to the "Bloody Code" and a rise in executions, which were often public and treated as a spectacle.

The US army has killed plenty of pregnant women in Nidal's lifetime.

This is a false equivalency and you know it, or should know it.

That's what many Americans earnestly believe.

Please quantify this statement and back it up with polling numbers or another citation.

Expressing themselves as completely incapable of recognizing any hostile act taken against them from the viewpoint of its perpetrators.

It's curious how one could get such a perspective. Certainly not from reading American novelists (such as Tom Clancy), or American scholars (such as Lawrence Wright) or American journalists (such as John Hersey) or American combatants (like Ian Fritz) or interacting with American Mottizens (we have GWOT vets on here, and I've never gotten that impression from them at all; rather the opposite.)

Nidal Hasan is a middle Eastern terrorist.

Repeating something doesn't make it true.

Maybe one day American soldiers will stop killing unarmed people in 'places of trust' the world over, but until they do, there's little rational to mourn them beyond anyone else when the shoe is on the other foot.

It is rational to punish traitors because the harm of their actions erodes trust, and as I think you know, a lower-trust society is a worse one.

Wrath is a sin.

This is true inasmuch as "wrath" is, when used technically, the sinful version of anger, but anger is not itself sinful in the Christian tradition.

If this guy was to be executed, walk into his jailcell in the middle of the night and end him. Everything else is degenerate.

Doing this is how the Soviets did it. Secret executions are the tool of repressive governments everywhere, and breed secrecy and fear. The Western tradition is one of public executions. The condemned also has, in the Western tradition, certain last rights that would be abrogated by executing him in his cell while he slept. As far as I know it is routine for other executions in the United States allow outside witnesses to attend.

You can argue that these are outdated ideas but understand that in the Western tradition you reference shooting someone in the back of the head in their cell block pattern-matches onto totalitarian regimes, whereas a public execution is pretty much the staple Anglo-American tradition.

Now you can also argue, I think, that livestreaming it is a bad idea. But it seems to me that we can address the question of if executions should be public and if they should be livestreamed separately.

Listening to this person made me sympathize with the soon to be martyr.

Why? He killed a pregnant woman who was his comrade-in-arms.

An American saying this about a middle Eastern terrorist is beyond a joke.

Nidal Hasan is not a Middle Eastern terrorist, he is an American citizen and soldier, born in Virginia, who shot his fellow soldiers to death when they were unarmed in a place of trust.

...well, I am happy to term that person as experiencing visceral distress. I don't need to additionally know if there are qualia present to conclude they are experiencing distress

If you conclude or agree that someone is experiencing visceral distress, you are agreeing that they have qualia because of what the words mean.

One might also say, qualia are definitionally present in such a case.

Are you arguing that we should define "qualia" (which references internal experience) based on external representations? This would pretty much destroy the meaning and utility of the word - you'd need to agree (for instance) that video game NPCs experienced the qualia of visceral distress because they were giving the external representation of doing so.

Probably worth noting that the professional definition of pain is linked to tissue damage which a machine cannot experience:

“An unpleasant sensory and emotional experience associated with, or resembling that associated with, actual or potential tissue damage,”

So functionalist or non-functionalist can embrace the idea that machines cannot feel pain, as long as they embrace the definition set forth by the International Association for the Study of Pain.

(On the other hand, I could see someone arguing that a machine could have a sensory experience resembling that associated with actual or potential tissue damage, but I think that's more of an uphill battle than arguing on the grounds of whether or not machines can be taught to avoid or prefer certain behaviors.)

Right but I also tend to think visceral distress is defined functionally without needing to rule either way on the presence of qualia, so I deny your premise.

Something can't be "distressing" and certainly not "visceral" without qualia, so in denying my premise you are affirming something that is definitionally impossible.

I don't have the technical expertise to evaluate this confidently. From what I do know, it seems plausible to me that for chaos-theory reasons any time we adjust an AI model's weights we may also change its behavior in unintended ways, which I think casts some doubt on "hard-coding an LLM to always obey instructions."

Qualia are an unhelpful posit at the best of times to the extent we can't be truly certain anyone or anything does or doesn't experience them.

I know I experience qualia.

Should we be kind to rocks because pansychism could be true and they too might experience qualia?

This is what the people who worry about AI experiencing pain think. But pain (and similar sensory experiences) aren't so mysterious that we can't come to conclusions on these sorts of matters. We know enough about pain to be pretty certain that trees (and rocks) don't experience it.

A better reason to object to causing people pain is because it's viscerally distressing

Only entities with qualia would find it viscerally distressing, so if you don't think qualia is a helpful posit, you don't think this is a helpful posit, either. Besides which, we have good inductive reasons to believe that some humans that have qualia don't find causing people pain viscerally distressing, so it's not a principle you can universally reason from.

Ironically I think it will be easier to decapitate AIs, at least in the medium term. And likely a viable strategy, even if they aren't in charge of the army, simply as an extension of the age-old tactic of disrupting enemy communications.

As I said, if the scaling stops tomorrow then maybe the models simply are not good enough to do this autonomously and other human labor is required to guide them along. If it continues at this rate for another year or two things change dramatically.

If things continue at this rate then OpenAI and Anthropic run out of money, that's kinda the problem.

I don't have time right now to rehash this debate again.

And that's fine, I understand time priorities. But this is, from what I've seen, a pretty typical blow-off when discussing the issues with models. "Pfft I know what it does for me" and not looking at the broader data or reviewing the literature for the downsides. And hey, that might not be a good use of your time! Just please understand where I am coming from, which again is very much not an "AI is useless" perspective. I have to consider how the average user will interface with the product. I don't think [insert Microsoft product] is useless, either, but that doesn't stop me from complaining at you for an hour about inserting pictures into Word being cognizant of the downsides.

using sonnet instead of opus, not having a proper harness and planning/review cycle.

Upthread you joke about "go ahead and stand up any product that would be profitable to run by a wrapper company, make no mistakes." Now, we both know that's a joke, but I suggest that AIs need too much human hand-holding to just be turned loose to solve problems, and when I illustrated this with a personal example, your response is...that I'm not properly hand-holding the AI. And fair enough, I don't disagree, but I think it rather proves my point.

You'd have to consider your time basically worthless to avoid playing $20/month for some opus 5.5 usage instead of sonnet 5.

I believe most AI users still use free plans (to my point about subsidized provision). As it happens, I actually do have access to Opus (and I've had issues with it too). It's just something I use for work and not when I am screwing around with strictly personal stuff.

However, I think this example is salient because "agentic swarms" are going to use more lightweight models to accomplish some tasks instead of doing it all on the latest model. Furthermore, Anthropic advertises Sonnet as a coding model. If your argument is "the models are actually bad at doing the things they are advertised to be good at" then you're agreeing with me that AI is overhyped.

(Which is fine! There's nothing inherently contradictory about thinking that AI is a marvelous tool and also that salespeople overhype things!)

In this particular case, though, the issue wasn't that there was anything wrong with the coding, and I think it proves nothing about that; it was a feature implementation decision.

Remember this is all predicated on having an objectively verifiable way to ensure that the AI is truly benevolent.

And there's really no way to do this, unless you are committed to never improving your model, since model changes might influence its ethics. Learning might change its ethics, too, so a model that can actually govern mankind will likely not be "safe" unless it represses humanity to prevent change.

I think it's quite plausibly that AIs end up slowing change just incidentally, of course, without government intervention or force, just by their mechanism.

Sure, I don't disagree.

I think (in the near future) that you're more likely to see this in ships than small airborne drones, though. ("Drone" can mean a lot of things, sometimes I assume without clarification that people are thinking of little FPV drones. Putting a lot of compute on a ship makes a lot more sense.)

That's a bit of a tongue in cheek example but what is stopping anthropic with a sufficiently advanced unreleased model from just becoming a hedge fund and doing to the financial industry what they're doing to mathematics?

Why haven't they done this already?

At least part of the answer to this question is probably that models often perform worse in real-world applications than their evaluation benchmarks would suggest. Thus, while the models are improving and becoming more useful for real world work, Anthropic can't just throw them at a problem and say "make no mistakes" without staffing up humans to oversee them.

Even in coding, which they have been optimized for, the data that I've seen indicates that trust by developers has fallen as adoption has risen, and that as LLMs have increasingly been adopted in coding, the amount of insecure code has risen substantially while deployments have actually slightly dropped.

Just as a personal example, the other day I asked Sonnet 5 to implement a design and it decided to simplify the product on its own recognizance. For me, it's a little annoying. But obviously at scale for a real-world product that's disastrous.

This is not an "LLMs are useless" post (I'm very very interested to see what comes out of the wetlab!), but it is a "LLMs are not ready to go to the moon on their own" post. It wouldn't surprise me if the primes try something like this, and they definitely might succeed. But it wouldn't be a "we can just let the AI do everything" sort of thing.

I also don't think they'll have trouble finding investors money if they ask for it and it's not obvious that scaling laws have failed

It seems like Anthropic and OpenAI think they will have at least momentary trouble, hence the delayed IPOs, and I noted the trouble the hardware suppliers were having upstream. If investors are becoming skeptical of datacenter investment, there's reason to think they are going to be skeptical of the primes. If they are skeptical of datacenter investment but not of the primes, they are just shooting themselves in the foot very dramatically, since the primes need the datacenters to scale. I suppose it's possible that they are all-in on orbital ones after the Space-X IPO, though.

This isn't a hard prediction here (nor do I think Anthropic and OpenAI delaying their IPOs is necessarily a bad idea or indicates terminal distress) but I think it's important to keep your eyes open. Particularly if you're intending to throw your net worth at them.

Research is being subsidized, the inference is not.

I'm getting inference for free, it's being subsidized.

If you've been keeping score the 2027 project is doing much better than its detractors and pretty damn good in absolute terms.

The area where I think it did the worst was in the economic projection of the stock market as a whole, which I think is pretty relevant to the fundamental question here.

It might be unlikely to be instantiated on Earth again without preserving the particular body, but that's not a problem for me because I will not be experiencing any time that passes while I am not instantiated.

You don't know that though, do you?

Probably worth keeping in mind that the drones cannot be coordinated out of line-of-sight. This isn't a really big deal if you have access to, say, Starlink, but short of that it's a huge problem. In the future, weapons that can effectively take care of Starlink can be expected to come online (if they aren't deployed already), so you'd need massive and very vulnerable backup communications network.

I think the bigger problem, though, is that your commander-in-chief is a massive huge immobile building that cannot be replaced. In a war between humans, killing a single person is generally understood to change little, but massive huge immobile buildings that cannot be replaced are pretty well understood not to be survivable.

Really the upshot is that if we permit a super-intelligent and evil AI to control a vertically-integrated supply chain that terminates in the military resources of a nation-state, it could in fact pose a serious treat to us. But this isn't really a unique problem, it's the same threat we face from other nation-states and insurgent groups.