@RandomRanger's banner p

RandomRanger

Just build nuclear plants!

2 followers   follows 1 user  
joined 2022 September 05 00:46:54 UTC

				

User ID: 317

RandomRanger

Just build nuclear plants!

2 followers   follows 1 user   joined 2022 September 05 00:46:54 UTC

					

No bio...


					

User ID: 317

Even an omniscient AI would still fight a war. War is about using force to achieve a political goal. If you have force and a goal, you can have a war. Rationality has nothing to do with it. Even if a party knows that it will lose a war, they will often continue fighting out of internal political considerations and spiteful hatred.

Hatred is rational. You would rather face a conciliatory pushover than a hateful, spiteful opponent.

Anyway, some AIs will be smarter than others and so they'll be stronger.

But I'm willing to bet that there are lots of people with fewer than Musk's 219 million Twitter followers who you and I would agree are smarter or more reliable guides

There are absolutely people I agree with more often than Musk. There are some people online who I think are very wise and I agree with basically everything they say. Whereas I disagree with many things that Musk says, we clearly have different goals and understanding of the world. So there are people wiser than Musk.

But that doesn't mean they're smarter than Musk. If they're smarter, then why don't they simply implement their visions and smash every obstacle in their path? Musk wants to settle Mars, so he simply takes over the entire spaceflight market with SpaceX. The Democratic Party/decel culture gets in his way, so he moves to smash them with Twitter and Trump. AI coming up sooner than expected, looks like that's important? Why not simply start a frontier lab? Electric cars and robots as well!

These are impressive achievements! It is hard to create things, rather than merely performing a role for someone else like so many. Try starting your own business. It's hard on a wholly different level.

When Bieber demonstrates general-purpose creative ability (at maybe 10 or 100 times his net worth), as opposed to just being a one-trick pony in music/infatuating young women, then I'll defend his general ability. Taylor Swift does the same thing better than Bieber and has basically no political influence (her endorsement had minimal effect), Musk is on a totally different level.

Biden was never smart or capable, his first presidential campaign crashed because he lied about being first in his class and plagiarizing. Pretty poor on 'not sounding like a fool in speeches' and 'avoiding scandals' too. Elon can sometimes sound like a fool in speeches and he is scandal-prone but there are other redeeming qualities that are lacking with Biden.

I don't know what selection mechanisms exist in the Democratic Party for leadership material but the people that gave us Joe Biden and Kamala Harris are clearly not prioritizing skills and ability.

You can be successful without being smart. Clearly I erred in tying 'smart' and 'successful' together when I was primarily talking about the allegations that Elon broke his brain. Most of the time you need to be smart to be successful.

If he's doing those things, then is he destroying his brain with drugs? What about Meta, what have they been doing? Has Zuckerberg been destroying his brain with drugs throwing $20 billion into VR with zero returns and with Meta AI getting defenestrated by Grok 3 and Deepseek, not to mention kowtowing to Trump after he won the election rather than supporting him before, like Elon? Running these huge organizations is difficult. There are ups and downs.

The elite, prestigious sources of information have discredited themselves. They go on and on about climate change (nothingburger) and demand extremely costly and ineffective fixes. They came up with DEI and globalized the whole US race obsession. They swept Rotherham under the carpet and brought us the summer of Floyd. They've damaged relations between the sexes considerably. They cheered for the retard wars in the middle east. They spurred political division by blundering obsessively and then screeching misinformation when anyone tried to point out their inadequacy.

The damage caused by the narratives they put out far exceeds anything Musk and far-right anime profile pics have done.

The US/allied theory of victory is mostly cope, rehashing the Cold War strategy of technological superiority to overcome numerical inferiority.

Only it's hard to retain technological superiority against a state with such a gigantic amount of STEM talent and a non-broken economic system ruthlessly prioritizing capital investment and technological superiority.

Plus numerical inferiority will be staggering. Chinese shipbuilding capacity is roughly 52-55% of world shipbuilding, roughly in line with their steel production. We're just not winning a naval war here, it's not going to happen. The little formation that circumnavigated Australia recently has roughly similar firepower (measured by VLS tubes) as the whole Australian navy - Australia is a useful ally for something like subordinating Papua New Guinea or clobbering sand people with special forces but we have negligible competence or firepower in a war of mass. South Korea and Japan are both heavily reliant on food/fuel imports and are de facto islands, they will struggle to sustain a long war against such a big power despite being somewhat serious. And we should assume a long war, all great power wars become long wars.

What stops China executing a full-court press over the Pacific, sweeping elan, training, tactics, fortifications and all else aside with numbers and production capacity just like the US did to Japan? Only temporary factors like the size of the US navy at present, incomplete Chinese autarky... But Chinese autarky is developing and the US navy is still shrinking!

https://www.19fortyfive.com/2025/02/the-u-s-navy-has-a-big-problem-shrinking-away-to-nothing/

What hope do we have when the biggest, strongest power in the bloc is withering away in peacetime!

The worst thing is that the Western world has decided that it's impossible for us to strike first, that's apparently unthinkably unsportsmanlike behaviour. Never mind that we have rapidly diminishing advantages in fleet tonnage, areas of technological superiority and training. We also have to concede the opportunity for the first strike, take another Pearl Harbour. This 'serious and thoughtful' blogger (clearly ex-military or otherwise initiated in these matters) wants to systematically eradicate Chinese military industry and academia but still refuses to consider a first strike: https://navy-matters.blogspot.com/2018/10/china-war-setting-stage.html

The grand plan is to fight the final battle for world dominance with a shrinking fleet against a continuously growing industrial juggernaut... letting them pick the time and place for the final showdown? What have our thought leaders been huffing?

I lay the blame on lib-racism. If you go back and read The Rising Tide of Colour you will ironically find a much more measured and sophisticated analysis of world affairs from a white supremacist writing 100 years ago than soaks through in vibes and in media today. Stoddard believed that the Nordic race was the master race with innately superior martial qualities and yet he still concluded that the Chinese were a serious threat through numbers and sheer tenacity if nothing else. He made the hypothesis that even though European troops would be better at fighting, when it comes to long marches and enduring privations the Chinese could even the odds. And indeed we saw something like that in Korea where clever Chinese tactics, night marches and similar proved highly effective against firepower superiority.

Of course, caring about how tough one's soldiers are in forced marches is suited more for the world of 100 years ago than today. But the abstract logic of looking at the situation as it is rather than as we'd like it to be is rare indeed. Maybe the country that wins all the physics and chemistry olympiads is good at physics and chemistry? Maybe the people that produce so much of the world's manufactured products would be extremely tough to fight in wartime? Maybe the country that's producing the most robots will have highly efficient industry? Perhaps the country that makes the most drones will be advantaged in the drone age? None of this is an extraordinary leap of logic, yet everyone seems to miss it.

People assume that the Chinese fleet is all 'Chinesium', that the concrete can be punched through, it's all slave labour and cheap copying, all their figures are invented and surely they'll collapse soon to the property bubble... Maybe in this fantasy world it is reasonable to let them get in the first blow and control the sequence of events. But we don't live in the fantasy. We can't rely on the B-21 or NGAD showing up to save everything, NGAD may not even arrive at all. There are going to be Chinese equivalents produced at growing speed and numbers (as we are seeing this year) because they have a rich country with vast resources to tap. Meanwhile our resources seem to be shrinking away into the ether and we can't seem to beat Yemen or outproduce Russia. This bodes ill.

I recently listened to a podcast he did in 2021 on the history of technology in warfare in which he seemed like a completely different man. He displayed not only knowledge in engineering, but history, including strategy and tactics in the Second World War. This supports the theory that something in this man’s brain broke around 2022

Who cares what Hanania thinks about human excellence? He has (generously) 1/1000th of Elon's following, maybe 1/100,000 of his wealth. Is Hanania running a viable AGI program? Is Hanania building huge rockets? Are Hanania's opinions relevant in world affairs, does he control key communications infrastructure used by armies? Is he doing anything of importance whatsoever? No. If anything he shot himself in the foot switching from 'I'm a smart tech-right policy guy' to 'let me sneer at all the right-wing retards who are now running the country and are in a position to implement policies'. He's the contrarian rat that jumps on board the sinking ship. What a fool!

Elon may indeed have lost some of his faculties, idk, I've never met the man. I doubt Hanania has either. Armchair psychoanalysis of extremely unusual people is basically just glorified name-calling.

Whatever Elon has lost, if anything, he still makes the rest of the world look like drooling retards. What did I get done in the last 3 years, since 2022? I certainly didn't start an AI company that's outperforming Facebook, Amazon and Microsoft. I didn't build the biggest datacentre on the planet at record speed.

It's perfectly reasonable for us to disagree with Elon's choices or think he should do something else. I disagree with Elon about many things, including his whole concept of what a state is for. But if people want to go around calling him dumb or saying that his brain 'broke', then we'd better have some serious achievements to prove that we know what 'smart' or 'successful' is! Certainly something better than 'I wrote a book rehashing Mearsheimer (nobody cares about it) and blew up my political career' like Hanania.

Why should anyone care what Hanania thinks about politics considering how bad he is at it? He was pivoting away from Trump while Elon pivoted towards Trump... I think it's clear who has better political skills and like everything else between them, it's an orders of magnitude difference.

Deepseek R1 is surely better at maths than most people on this forum and doubtless far superior to this doctor: https://old.reddit.com/r/singularity/comments/1i5r85h/deepseekr1_scored_100_on_a_2023_a_levels/

Maybe she just has really low ability in maths but has otherwise fine working memory and similar.

Some people are like that.

And she can just use AI for maths.

"When in a critical military situation, outnumbered and outgunned, further disperse your forces by attacking elsewhere and expanding the front" isn't going to be making the director's cut of the Art of War.

Especially since the original goal was to use parts of Kursk as a way to get other land back. This clearly hasn't worked.

It defies reason that the best way to defend Pokrovsk would be to attack Kursk. The forces committed to Kursk could've been simply sent to defend Pokrovsk and stabilize the front there.

There are surprisingly reasonable arguments that the US should take steps to break up and annex parts of Canada (or at least intervene with their domestic policies of mass immigration that are undermining stability) to sustain continental hegemony:

https://arctotherium.substack.com/p/the-canadian-question

The primary cause of WW2, as opposed to a German-Polish or German-Soviet war. was British and French leaders being terminally stupid and lacking basic concepts of strategic thought such as 'do not start major wars without offensive capabilities' or 'make alliances with the strongest nearby powers before starting a war'. For some reason they believed that the borders they'd drawn up in 1919 at Versailles were sacred, precious, perfect creations that had to be defended at all costs. Nobody who lived in Eastern Europe liked those borders and they were nearly all later revised by Germany and then the Soviet Union. Poland was very happy to rip some land off Czechoslovakia, Ukraine did the same to Poland later on... It was a feeding frenzy.

The 'Czechoslovakians' didn't even like Czechoslovakia. The country broke up once Germany took the Sudetenland and it broke up again in the 1990s.

And as the cherry on top, European leaders have now totally dissolved the normal meaning of borders with mass migration. The population of London is something like 30% British! It's bizarre to go to so much effort defending Ukraine's borders when it is apparently impossible for the British government to prevent random people coming over the channel and living in their country.

The result of artificially creating and defending this weird equilibrium isn't that it stays perfect and static forever. It's that change happens suddenly and chaotically in a vast storm surge that smashes every bulwark and barrier established against it. WW2 is one example. No static system can survive a dynamic world.

What are the gains from invading the Baltics relative to the risks? It doesn't make sense from Russia's perspective unless NATO dissolves. The botched handling of the Russia-Ukraine war seems to have done a lot of damage to NATO unity but NATO isn't totally broken right now.

Fears about the Baltics from Ukraine are rehashed domino syndrome that makes even less sense.

It seems a lot less destabilizing if "I deserve this territory because our leader founded the cities" and "my nationality inhabits the place" is preferred to "I have this ideology and I want to spread it".

The former cedes Taipei to China. The latter cedes the world to China.

The former cedes Eastern Ukraine to Russia or accepts that it's vaguely contestable (Germany would have a similar claim on Kaliningrad for instance). The latter cedes the world to Russia, or as much as they can get their hands on.

We can't go around attacking random countries around the world for the most abstract, random reasons and then complain when other people do the same thing to their neighbours for much more reasonable causes. The territorial sovereignty of Afghanistan or Iraq or Libya (or Pakistan for that matter, the US freely bombs and sends special forces in there) is totally worthless. We wield arbitrary power over much of the world because we're rich and strong. But others are rich and strong, they can do the same thing as us.

Who made the rule that 'annexing territory is uniquely bad'? Where was it agreed that you can have a war to install a puppet government but not annex? Would it be OK if Putin just set up more puppet governments, more people's republics like Donetsk and Luhansk? No, obviously not. The exact same people were bitching before Russia annexed those people's republics and after, they'd just find different words.

More attractive than anything we can imagine. But their version of kink looks like Event Horizon and the Kama Sutra if it was written by the Dark Eldar.

The best time for Ukraine to restart their nuclear weapons program would have been when Russia defected from the Budapest Memorandum by annexing Crimea

What would the Russian reaction to this be? Would Russia sit around idly while a neighbour with a hostile government nuclearizes? Or would they go in hard and pre-empt nuclearization? One of Zelensky's many bizarre pre-war diplomatic maneuvers was making strange threats about nuclearization. Big nuclear powers tend to get hysterical when hostile neighbours nuclearize or are nuclearized. See the Cuban Missile Crisis for example. The US was hours away from launching a disarming strike on Cuba, they were dropping dummy depth charges on Russian submarines.

Furthermore, the Ukraine war is if anything much less a war of conquest than our Middle East wars. Ukraine is full of Russians and Russian speakers. The commander of the Ukrainian army is Russian, Russian family, educated in Moscow. A significant number of the forces Russia has were drawn from Donetsk and Luhansk which were provinces of Ukraine. Many of the territories in question were part of Novorossiya: Catherine the Great founded Dnipropetrovsk, for instance. Both sides appeal to common historical concepts, calling each other Nazis. The majority of fighting is conventional, between uniformed soldiers.

In Iraq and Afghanistan there was a much clearer division between 'us' and 'them'. Nobody ever found any historical claim for the US to be involved in running Afghanistan or Iraq, such an idea is ludicrous. They're on the other side of the world! The wars were justified via broader universal liberal principles, the need to reshape the Middle East...

At no point was the commander of the Taliban American or British, it was a war between Muslim Afghans/Arabs vs secular European/Americans. There were some auxiliaries drawn from the locals but these proved to be extremely low-quality troops and caused considerable green-on-blue attacks. Western-trained auxiliaries usually disintegrated the moment they ran into any motivated local force (like the Taliban or ISIS) without Western backup. The local population was not really aligned with Western forces and much of the fighting was unconventional with guerrilla tactics and suicide bombings. There was a massive ideological clash in all respects, the forces of Islam vs the forces of secular liberal democracy.

If an alien race shows up and conquers the world, installing strange values like mandatory veganism and bestiality, that's a war of conquest. They can't say 'oh we're just installing a new regime not conquering anything!' when they have no legitimate claim to Earth and only a bunch of perverts and weirdoes collaborating for them.

My point is that we should not conclude that because Russia invaded Ukraine, they will also try and invade Poland or Sweden or Azerbaijan. Ukraine-Russia is a special case where there are a wide range of justifications for Russia beyond 'Russia must grow larger'. The naval base in Crimea, the Novorussia territories, laws regarding the Russian language, potential NATO expansion...

Nor should the rules-based order be held up as this golden age because there was no conquest. The 'rules-based order' directly led to the situation today. Putin has complained repeatedly about the invasion of Iraq, various unilateral actions from the West. China wasn't keen on it either. What were the rules of the rules based order, are they listed anywhere? If we lack the strength to enforce the 'only we can invade countries' equilibrium because we abused it (and failed to even reap any gains from abusing it), then it's time to abandon it and move on without any nostalgia. Rebuilding this equilibrium is not desirable! Lessons must sink in.

If Ivan is dead, then who is making more shells and drones than the US or Europe?

The Russian draftees aren't the primary troops on the front they mostly do rear-area work. Russia has been pumping up soldier's wages looking for volunteers on the front. Ukraine is the side reliant on draftees to hold the line because they have more casualties and less manpower. Which is exactly what we'd expect in a war between a big country and a small country.

Remember when the US pulled out of Vietnam, and this lead to a total collapse of their country?

Ukraine is not Vietnam to Russia. They really, really care about what goes on in Ukraine and will expend vast efforts there, not least because there are a large number of Russians living in Ukraine. Most of the country speaks Russian. The commander in chief of the Ukrainian army, Syrski, is an ethnic Russian who was trained in Moscow Higher Military Command School, his whole family is Russian. This war is halfway between a civil war and an interstate war, despite what the media would like us to believe.

There are serious risks of nuclear use if the Ukrainians somehow threatened to seriously defeat the Russian army, which they have shown little sign of being capable of. The nominal Western strategy is to achieve this though, the 2023 counteroffensive was supposed to do this IIRC. Our strategy is incoherent, there is no clear path to victory.

Russian oligarchs are not profiting from this war. Oligarchs would like to trade with the West, party in the South of France or launder money offshore. Vlad in the armaments factory and textile mill is the one profiting from this war, Ivan living in the formerly-decrepit Urals industrial town that's now a booming part of the war economy is seeing his income rise. The political losers from this war are the English-fluent laptop class who've had many of their foreign assets and IT jobs frozen, many have fled Russia for Turkey or Argentina. The most pro-Western elements in Russia have been politically euthanized while the less-educated patriots are getting more prestige and wealth. Putin is relatively moderate by the standards of Russian leadership, Medvedev is constantly making radical hyperbolic threats and statements online. Other academics have called for a demonstration strike on Poland.

I foresee many nasty surprises waiting for us as this war progresses, it should not be considered in this casual way. It does have a material effect on all of us via energy prices if nothing else.

Interesting stuff, why does India leave the British? Aren't the British still very strong as a power, the Americans weren't slapping them down? Do they pull out because it wasn't cost efficient to hold onto or something?

My general althistory headcanon is that a lot of anti-colonial revolts/insurgencies would've been put down hard if the two superpowers (USA and USSR) hadn't been anti-colonialist. Nerve gas and bombers is a tough combo to beat.

A couple brigades of Wagner are sufficient today to overthrow a mid-sized Central African country, one would think that some big power or other would end up running these places if only to secure gold mines and oil wells.

We already have AI models improving kernels (for AI deployments) today. It's not a big leap of logic that they'll do more and more, achieving takeoff.

https://developer.nvidia.com/blog/automating-gpu-kernel-generation-with-deepseek-r1-and-inference-time-scaling/

I think by 2027. Executives in Anthropic, OpenAI seem to give that as their date.

Defining AGI is tricky. I think the key element is take-off. AI right now is best at coding. You use code to make AI. Recursive self-improvement is the name of the game, AGI will be when you have bots performing the tasks that AI researchers do now in collating data and training models (not in the partial sense like how synthetic data is used today but in a holistic sense where the main intellectual work would be done by AIs). And AGI will be ephemeral because superintelligence will happen immediately afterwards and things will get very crazy very quickly, the world power structure will change fundamentally. Nobody will be asking 'what does this mean for unemployment' because that will be the least of our concerns. We will know AGI when we see it.

Little things like mapping the relations between objects in space don't disprove intelligence. If you presented Pokemon like a text adventure game, Claude would have no problem winning. The intelligence is there, that's what they're working on. Advanced vision isn't there, people don't particularly need AIs to play Pokemon, they're needed for writing code.

And yes I am heavily, heavily invested in AI companies, so I have some skin in the game.

They could get away with a tactical nuke, it's just that doing so would incur various costs. It's just a matter of calculation about risk and reward. If somehow the whole Russian army got encircled in Mariupol, they might well start nuking intensively rather than lose the war. The US considered nukes in Korea and Vietnam but concluded the costs weren't worth the gains.

These weapons aren't unthinkable, that's just a social construct that the US likes to propagate.

Russia has been confident of conventional victory the whole time and doesn't want to irradiate land it wants to conquer, a country they want to vassalize or annex.

In the Cold War the Soviets demonstrated what they'd do if they lost a European satellite - send in the tanks!

LLMs have intelligence, what they don't have is advanced spatial skills and visual comprehension. Claude Sonnet 3.7 is designed to code first and foremost, secondarily as a writer/conversationalist. Game-playing and consoling in Minecraft cathedrals (which Sonnet does quite well) is a tertiary capability that they didn't even aim for but are testing anyway. They didn't try to make it good at this, unsurprisingly it's not that great at it.

I had Sonnet play through a civ 4 game where I implemented its strategy and tactics, it was perfectly capable of reacting to text inputs but didn't really understand the pictures, where units were in relation to eachother. If you give it the inputs in the medium it understands best, text, it's pretty capable. When these AIs struggle with strawberry, that's because their tokenizer can't properly count letters. They never see a single letter.

Have a look at what they do with code in minecraft: https://youtube.com/watch?v=FCnQvdypW_I

It's a long video but you can take a general look at what they build. Can you build a house by coding it in? OK, they're not that great at stairs. But they are directly coding things in. I bet 98% of the planet couldn't do this specific niche skill that humans have no aptitude for, coding in architecture in minecraft. That's not a tool in our arsenal. It doesn't discredit our intelligence because we're bad at things we're not supposed to do in mediums unintuitive to us.

Now consider Alphago. You can't beat it, nobody can. A few people can beat Google's 7 year old AI Starcraft pro, not very many though (and this is the version downgraded to non vastly superhuman speed).

That Sonnet can kind of play Pokemon is proof in my mind that AGI is imminent. It proves that Sonnet's intelligence generalizes out into domains it wasn't remotely designed for, even crippled by the visual input barrier. Combined with the specialized bots, we have both 'extremely broad' and 'extremely capable'. All that remains is marrying the strengths of both approaches together, scaling up, working on visuals and tweaking.

Consider just today another AI agent arrived, it seems pretty capable: https://x.com/LamarDealMaker/status/1898454061277458498

But sending out tanks does gain you things? The Russians have secured a swathe of territory in Donbass, they took Mariupol.

We're agreed that the rules can't be unilaterally changed but I think there must be some concrete reason why all the powers invest so much into conventional forces. Nukes are very powerful but not appropriate for all conditions.

Even in the Cold War everyone was stacking up huge columns of troops in Europe, along with masses of nukes. Nukes held the line for the Western bloc up till about 1978 when they started to gain a conventional advantage. But people were still interested in conventional weapons.

It's Uganda though, the home of Idi Amin and Joseph Kony. It seems pretty reasonable to conclude that women have a tough time there, as does everyone else. Shouldn't our default expectation for Uganda in all departments (and especially the quality of the population) be 'very bad'? I thought Uganda was where they were killing bald men for gold in their heads, turns out that was Mozambique... Anyway, they have plenty of problems:

https://en.wikipedia.org/wiki/Child_sacrifice_in_Uganda

If there's widespread child sacrifice, why not wifebeating too?

There's nothing wrong with preferring easy wars to hard wars ceteris paribus. The costs have to be proportionate to the gains.

There's nothing wrong with focusing on primary threats, as opposed to secondary ones.

There's nothing wrong with seeing a conflict overseas and doing nothing about it since it's not relevant to your interests. Plus it usually causes all kinds of flow-on problems if you do intervene.

Colby is no isolationist, if you read his book 'strategy of denial' he says that the US goal should be to back up frontline allies in Asia to prevent Chinese hegemony over this very valuable and important region. He judges that Russia is not powerful enough to threaten hegemony over Europe, the Chinese are the primary threat to US power and so there needs to be a substantial US presence in Asia, he wants to maintain alliances. It's a judicious, strategically justified rationale.