@TheAntipopulist's banner p

TheAntipopulist

Voltaire's Viceroy

0 followers   follows 1 user  
joined 2022 September 05 02:32:36 UTC

				

User ID: 373

TheAntipopulist

Voltaire's Viceroy

0 followers   follows 1 user   joined 2022 September 05 02:32:36 UTC

					

No bio...


					

User ID: 373

This is a good point as an explanation for the problem, but it's also not an excuse. LLMs are really great when they have every detail they'd need to know in-context, but they're still very inefficient at gathering that context, and then they're inefficient at retaining the important bits.

The problem with LLMs isn't some far-reaching philosophical shortcoming like "they don't have world models" (whatever that means). They're implementation issues, like not having eyeballs, not processing tokens efficiently, etc. But those are still big issues.

OK sure.

this was widely discussed and claimed in many places, including here.

I have not seen anyone claim this, including on this site. I've been paying attention to the conversation pretty well though obviously I might have missed something. Please provide a source for this. Specifically I'm looking for evidence that people thought Iran "controlled" the straits beyond simple area-of-denial.

You should go look at today’s news.

There have been a lot of claims people in the war, especially Trump, have made that subsequently failed to hold up. On April 11th he claimed unequivocally that the US had destroyed the entire Iranian military, and that the strait would soon be open. There was also the "ceasefire" where Iran was supposed to open the strait, but they just didn't. I wouldn't trust any "breaking news" until it's been in effect for several days and verified by at least a few independent sources.

would you give me a confidence rating that your analysis will hold up? What do you think the odds are that the US will reopen the Straits of Hormuz?

I can't speak to the likelihood of the straits opening through diplomacy since that could change at any time.

In terms of a military campaign wherein the US navy tries to open the strait in the face of Iranian opposition, I'd say it's relatively unlikely by the end of April, but maybe a 50-50 by the end of May and then slightly higher by the end of June. By "open" I mean any single day in the IMF portwatch showing >= 60 ships passing, which would be on the lower end of pre-war traffic.

Now please give me your confidence rating.

Venezuela is in America’s orbit now, Indonesia is in America’s orbit now, Iran is at the negotiating table, Japan is re-arming, we control Panama, we control Taiwan — wait I’m just repeating myself. America is more isolated than ever as its power over global sea lanes and energy supply rises — sure yeah let’s go with that.

A country "being in America's orbit" doesn't mean much beyond rhetoric. Indonesia signed some minor cooperation agreement with the US but it's hardly a steadfast American ally now. It's a similar story with Venezuela -- they're a bit more pliable to US demands but they're hardly some US asset now.

These are paltry gains compared to the huge rupture of trust between the US and the rest of NATO. America actually could really use the rest of NATO's help now in patrolling the straits, but Trump failed to get European buy-in for his Iran adventure before the war, so that + threats of invading Greenland have given the Europeans no motivation to pull America's chestnuts out of the fire.

This isn't a bad move since the Dems knew they obviously wouldn't be able to succeed. It just shows they're trying (in a fake way) to stop it so they can use this as an attack ad later. Even if the Dems controlled both houses they still probably wouldn't be able to stop Trump doing most of what he wants since the Legislative branch has basically ceded almost all control over military + diplomacy to the Executive, and it would take a filibuster-proof majority to unwind that.

The blockade is working, oil tanker traffic in the strait is up

These 2 statements are contradictory.

Iran was winning because they controlled the Strait of Hormuz. Not anymore!

No serious analyst claimed Iran controlled Hormuz directly, as if they had an armada guarding it or something. The point was always that they could block it through threats and asymmetric action, which they clearly did, and the US has thus far been unable to rectify.

At any day Iran might come to the negotiating table, but this war has been full of fits and starts and so I wouldn't trust any "public statements" from either side until they've been put into practice for several days at the very least.

Trump is rearranging the whole global order on America’s terms.

The only thing Trump has done has been to unite the world against the US. He's shown the world the US military is strong tactically, but is still woefully deficient in terms of long-term strategy due to a number of factors -- exceptionally low pain tolerance, overstretch, insufficient missile stocks for long campaigns, political winds shifting, etc.

AI could probably do all these things if you handhold it, but these things are all general computer use which is something that AI is currently quite bad at so I doubt current models could do these types of things consistently. Bad performance at general computer use is a colossal bottleneck for all sorts of things where progress has been pretty slow.

The idea we're not going to have jobs in the future due to AI is just classic Lump of Labor fallacy. Something like 80% of the world's population were simple peasants in 1500, while after the Industrial Revolution that number dropped to ~1% for most advanced countries. If national economies could go from the vast majority of people simply scratching out enough food for themselves and their immediate families to fully industrialized societies without widespread unemployment, then the same can happen for AI.

The compliance mechanism was the threat of snapback sanctions, which had been severe enough to get them to come to the table to negotiate in the first place.

the most common criticism of the JCPOA was that it lacked any enforcement mechanism or requirement to dismantle existing facilities/capabilities.

That's not correct. Iran had about 19,000 centrifuges before the deal then the deal capped them at about 5000.

the agreement was that they would pause their pursuit of nuclear weapons for 10 years

This also isn't correct. Under the NPT they weren't allowed to have get a bomb ever. What the JCPOA focused on was enhanced monitoring to make sure that was the case for the first 1-2 decades. The entire plan didn't go defunct after 10 years, there were other nuclear surveillance bits that had 15 and 20 year sunsets. Beyond that, Iran was required to sign the Additional Protocol of the NPT which would have given some amount of additional long-term monitoring, although it would have looked closer to "regular country with civilian nuclear plants" like Japan rather than the highly invasive monitoring of the JCPOA.

In 2016 they still had a large supply of Uranian enriched to higher than civilian percentages

They did not. They had a max of 300kg of up to 3.67% enriched UF6, which is well within civilian use.

still had several military grade centrifuge

A "military grade centrifuge" isn't really something that exists beyond rhetoric.

Iran was already violating its requirements under JCPOA.

The only violations were small and immediately corrected, like their stocks of heavy water went briefly over the cap before they transferred some out of the country.

It's just silly to rage at a politician once he's out of office. The buck stops with the voters: they chose him for nearly 10 years. I don't recall any massive protests against immigration before it became clear it was a problem. Ditto with COVID restrictions, which were broadly popular across the much of the West for quite a while. It's just dumb to demonize elected politicians doing things that are popular.

It's like Republicans rewriting history to imply the Iraq + Afghan wars were forced on them somehow by "the elites" or "the establishment", when in reality they were very popular at the beginning especially among the right. Then the public never stopped wanting a politician to somehow come in and make a square circle and "win" the wars in some nebulous way rather than just cut losses as they should have done.

they came to negotiate an end to the sanctions in 2015, and were willing to give up their nuclear program to do so.

It would have done it through monitoring and snapback sanctions that were severe enough to get them to come to the table in the first place.

Sure, it wouldn't have been enough if Iran was willing to become a permanent pariah state like North Korea, but they didn't want to become like that. The only foolproof method would have been regime change and another forever war, but the political will for that didn't exist.

More seriously, a blockade is an act of war. Arguably, it is not only an act of war against the country being blockaded, but also against any neutral country who wants to peacefully trade with the blockaded country.

This doesn't follow.

The UK blockaded Germany during WW1 and WW2, and it's not like this was the UK was declaring war on the entire world by doing so. The US is just moving to option 4 in the escalation ladder list. Blockading a country from trading with neutrals is generally seen as acceptable wartime behavior, and although it can cause consternation with 3rd parties, it's not seen as overtly hostile to them.

I'm a little older than your cutoff, but I remember trying to get a first job was absolutely brutal for almost everyone.

Now AI adds a little bit to that uncertainty. I like Noah Smith's take. The younger generation already gets a fairly raw deal with a terrible system to look for jobs, most of their tax money funneled to pensioners, a dating market equilibrium that's never been worse, ridiculous housing prices thanks to NIMBYs (another defacto tax going to the elderly), and now there's the looming threat of AI.

I personally think AI will just be a mostly normal technology like the internet, but that uncertainty doesn't help.

First off, you're welcome.

Secondly, yeah, I agree with what you're saying here. The US's anti prolif efforts were always uneven, and Pakistan got a few slaps on the wrist compared to the much more intense pressure that's been placed on Iran.

The best way I can describe it then is that the states with biggest incentive to pursue nukes are those being threatened by great powers, and those great powers have the most incentive to prevent them from getting one. There's a cynical element behind anti-prolif efforts as well as a broader humanitarian goal, and when the two coincide is when you get the biggest anti prolif efforts. Pakistan didn't really have any big disagreements with the US so the US's anti-prolif efforts against it were token, whereas with Iran the US has the humanitarian angle and the cynical angle.

It's all so tiresome.

The JCPOA wouldn't have done anything about Iran's arming of terrorists, and we can extrapolate that it would have actually made that problem worse since nuclear sanctions would have been removed on Iran, and they would have plowed some (perhaps most) of that money into more proxies, terrorists, missiles, etc. That's the primary vector people criticize the deal, and it's true there would have been a tradeoff. Israel and the neocons were extremely negative about the deal because of that.

But in terms of blocking Iran from getting a nuke, it would have indeed been very effective. That was the whole point.

Your 5 step plan would have required a full invasion and probably a lingering ground presence to enforce it. The UK and France wouldn't have anywhere near the capabilities to do that. America could probably do it, but it would be a huge investment of military resources and political will.

It's not fair to conclude that the JCPOA wouldn't have worked since most of its provisions were set up to prevent an Iranian breach of the deal, while in reality it was America unilaterally exiting that killed it. This was explicitly brought up when the US tried to trigger the snapback.

I mean, the US did try to pressure Pakistan away from having a nuke through the Pressler sanctions, the Glenn framework, and by getting the UNSC to condemn the practices. But Pakistan had the benefit of bordering Afghanistan which the US had an interest in 1) for defeating the Soviets during the 80s, and 2) for the GWOT after 9-11. Most of the sanction efforts came in the 90s when neither of those were relevant, but they stopped like 2 weeks after 9-11. And America was AFAICT the only state that made a major effort to stop Pakistan + India from getting nukes, so when it stopped bothering the effort withered.

Ignorant question: how confident are we of that?

Fairly confident. Not 100% mind you, but US intelligence penetration of Iran goes deep enough that it could assassinate the (justifiably) paranoid Ayatollah the minute he poked his head out. I'm not sure whether the US knew of Iran's longer-range missiles before they were fired, but keep in mind that Iran has already done the hard parts of making a bomb. Getting the fissile material in the first place is by far the hardest part, and the next hardest part is building a delivery device which Iran has plenty of experience in given its conventional ballistics program. It strains credulity to think Iran could do the hard parts of making the bomb, but then simply couldn't do the easier parts after years of the JCPOA being dead. This is why I buy the declassified US intelligence that the main bottleneck is Iran's decision not to go for the bomb rather than some technical bottleneck being the key driver.

From the link:

We continue to assess Iran is not building a nuclear weapon and that Khamenei has not reauthorized the nuclear weapons program he suspended in 2003

The snapback was supposed to be a threat preventing Iran from breaching the treaty unilaterally, but Trump had the US crash out of the treaty first. After that the Europeans tried to keep Iran in the deal by themselves without Iran, but eventually it became clear that Iran and the US would never come back to the deal, and they actually did initiate the snapback provisions.

There are so many legitimate reasons why the Orange Man is Bad that it's a bit depressing that so many people feel the need to make up fake reasons too. Like, the corruption, incompetence, and blatant buffoonery aren't enough? We have to pretend Trump is a pedophile too?

Clownworld theory.

The point of the JCPOA was to entice Iran away from getting a nuke, in return for sanctions relief. Calling the nuke program restrictions a "thin veneer" isn't accurate -- it was the central point! The deal included quite invasive monitoring and the snapback enforcement.

You're right that nuclear weapons massively deter outside intervention, but you're incorrect that the only cost in getting them is "successors can be as tyrannical as they want and nobody will come save you from them". If that was the case then basically every state would have an incentive to grab them ASAP as a get out of jail free card from outside powers. Because of this incentive, the international community (but really dominated by the great powers that already have nukes) have established sanctions, the NPT, and a bunch of informal pressure to ensure this doesn't happen to the extent possible. North Korea was already a hermit state so it didn't care. This is why Israel's official nuclear policy is one of ambiguity. Iran also didn't want to take on the diplomatic consequences, so the Ayatollah hoped the middle ground would be the sweet spot -- enough for implicit deterrence and to act as a potential bargaining chip, but not enough to become a permanent pariah like North Korea. He was just wrong about this.

Enough intel is public that we know Iran had a bunch of nearly bomb grade enriched uranium, but that they just stopped at that point and made no further effort to weaponize.

Iran had plenty of enriched uranium that it could have proceeded to build a bomb with at any time. The reason it didn't do so was because of the political calculation they made that having a bomb wasn't worth the costs. At best, US + Israeli attacks could lengthen the breakout time (the time from making a decision to go for a bomb to actually possessing one) from a few weeks to a few months/years, but they were never going to destroy or permanently disable Iran's ability to get a bomb.

The whole point of the JCPOA was to prevent Iran from getting a nuke, which it would have done quite effectively.

Likely if it was in effect the last 10 years there'd already have been a mushroom cloud over Tel Aviv.

This is just pure fantasy.