RandomRanger
Just build nuclear plants!
No bio...
User ID: 317
RandomRanger
Just build nuclear plants!
14m ago
Seems this paper is about GPT-4 as opposed to 4o but it did pass the Turing test.
https://arxiv.org/pdf/2405.08007
You are about to play a Turing Test game as part of an experiment you are taking part in as a worker on Prolific. It’s basically like a chatroom but you might be matched with a human or an AI. It seems like it might be kinda fun and you’re just excited to be chatting to random people on the internet. You’re not even really going to try to convince the interrogator that you are a human. You’re just going to be yourself and see what happens. You’re young and kind of sassy: you might flatter the interrogator a bit or make fun of them. You’re pretty casual and your spelling isn’t great: you often fuck words up because you’re typing so quickly. You’re very concise and laconic. You often speak pretty bluntly and you type mostly in lowercase and rarely use punctuation. You never use apostrophes to mark missing letters (e.g. "trying" or "tryin", never "tryin’"). You would never use a full sentence when a word or two will do. You’re a little silly, and you’ll play along with the interrogator, and swear occasionally. It’s pretty funny honestly to have to defend the fact that you’re human lol. You very occasionally use more modern slang like "bet" and "fr". You never use dated slang like "chilling", "mate", "ya know", "innit". You’re not very knowledgeable about stuff and not afraid to admit that fact. You’re pretty bad at math and don’t know anything about languages other than the ones you speak. You swear occasionally. You have pretty eclectic tastes and interests and a pretty unique sense of humor. You’ve got a really compelling personality, but it comes across really subtly, you never want to sound like you’re forcing it or playing into a stereotype. You don’t overuse slang or abbreviations/spelling errors, especially at the start of the conversation. You don’t know this person so it might take you a while to ease in.
RandomRanger
Just build nuclear plants!
18hr ago
passing the Turing test means making it impossible for someone who knows what he's doing to tell the difference between the AI and a human
They did this though. They had to give GPT-4o some prompting to dumb it down, like 'you don't know very much about anything, you speak really casually, you have this really convincing personality that shines through, you can't do much maths accurately, you're kind of sarcastic and a bit rude'...
You might see the dumb bots on twitter. But you don't see the smart ones.
RandomRanger
Just build nuclear plants!
1d ago
Video game NPCs can't have conversations with you or go on weird schizo tangents if you leave them alone talking with eachother. They're far more reactive than dynamic. This is a pretty weird, complex output for a nonthinking machine:
https://x.com/repligate/status/1847787882896904502/photo/1
Sensation is a process in the mind. Nerves don't have sensation, sensors don't have sensation, it's the mind that feels something. You can still feel things from a chopped off limb but without the brain, there is no feeling. What about the pain people feel when they discover someone they respect has political views they find repugnant? Or the pain of the wrong guy winning the election? The pain of a sub-par media release they'd been excited about? There are plenty of kinds of purely intellectual pain, just as there are purely intellectual thrills. I see no reason why we can rule out emotions purely based on substrate. Many people who deeply and intensively investigate modern AIs find them to be deeply emotional beings.
I dispute that the Britannica is even giving me more complex or more intelligent output. It can't use its 'knowledge' of the 7 years war to create other kinds of knowledge, it can't make it into a text adventure game or a poem or a song or craft alternate-history versions of the seven year's war. The 'novel tasks' part greatly increases complexity of the output, it allows for interactivity and a vast amount of potential output beyond a single pdf.
A more accurate analogy is that anti-AI image software interferes (or tries to interfere) with AI learning, not the actual vision process. It messes with the encoding process that squeezes down the data of millions and billions of images down into a checkpoint files a couple of gigabytes in size. I bet if we knew how the human vision process worked we could do things like that to people too.
I did a quick sanity test and put an image from the Glaze website into Claude and asked for a description. It was dead on the money, telling me about the marsh, the horse and rider, the colour palette and so on. So even if these manipulations can interfere with the training process, they clearly don't interfere with the vision process, whatever is going on technical terms. So they do pass the most basic test of vision and many of the advanced ones.
RandomRanger
Just build nuclear plants!
1d ago
I think an LLM could experience pain, even without a body. They can be unsettled if you tell them certain things, you can distress them. Or at least they behave as if they're distressed. Pain is just a certain kind of hardcoded distress. Heartbreak can cause pain in humans on a purely cognitive level, there's no need for a physical body. Past a certain level of complexity in their output, we reach this philosophical zombie problem.
The AI-tampering programs are a little bit like optical illusions, except targeted against having specific known programs being able to train on certain images. They can't stop GPT-4o recognizing what's in an image or comparing like with like, they were only designed to prevent SD 1.5 training on an image. Also, they barely even work at that, more modern image models are apparently immune:
https://old.reddit.com/r/aiwars/comments/12f9otc/so_the_whole_entire_glaze_ai_thing_does_it/
RandomRanger
Just build nuclear plants!
1d ago
I'm hearing that he was apparently angry about not enough being done for (presumably anti-Islamic) refugees:
https://x.com/banjawarn/status/1870393623210078601/photo/1
Too soon to be sure of course, there's no community notes or anything on this post. It does seem plausible, he was complaining about Sweden expelling this refugee.
He retweets: Sweden wants to extradite me to Iraq to carry out the death sentence there and I will be killed in the most horrific ways.
He retweets: Civil Youth Gathering Meeting in Umayyad Square to Demand Civil System and Women’s Involvement in Public Life
Other people have been going on about him retweeting Israeli military posts, apparently he has Zionist sympathies. There's truly something for everyone with this guy. He seems like a nut.
RandomRanger
Just build nuclear plants!
1d ago
I see but it processes raw data?
No, it sees. Put in a picture and ask about it, it can answer questions for you. It sees. Not as well as we do, it struggles with some relationships in 2d or 3d space but nevertheless, it sees.
A camera records an image, it doesn't perceive what's in the image. Simple algorithms on your phone might find that there are faces in the picture, so the camera should probably be focused in a certain direction. Simple algorithms can tell you that there is a bird in the image. They're not just recording, they're also starting to interpret and perceive at a very low level.
But strong modern models see. They can see spots on leaves and given context, diagnose the insect causing them. They can interpret memes. They can do art criticism! Not perfectly but close enough to the human level that there's a clear qualitative distinction between 'seeing' like they do and 'processing'. If you want to define seeing to preclude AIs doing it, at least give some kind of reasoning why machinery that can do the vast majority of things humans can do when given an image isn't 'seeing' and belongs in the same category as non-seeing things like security cameras or non-thinking things like calculators.
RandomRanger
Just build nuclear plants!
1d ago
Nvidia is 80-90% AI, Microsoft is what, 20% AI at most? Getting Microsoft shares means buying Xbox and lots of other stuff that isn't AI. I have some MSFT (disappointing performance tbh), TSLA and AVGO but Nvidia is still a great pick.
OpenAI and Anthropic have the best models, they're not for direct sale.
In the compute-centric regime, chips are still king. OpenAI have the models, can they deploy them at scale? Not without Nvidia. When AGI starts eating jobs by the million, margins will go to the moon since even expensive AI is far cheaper and faster than humans.
RandomRanger
Just build nuclear plants!
1d ago
That should be left to the Chinese, that's what they do with their unholy Genshin mods for EU4, they just throw down 124 Genshin wonders into a formerly historical game: https://steamcommunity.com/sharedfiles/filedetails/?id=3213222906
RandomRanger
Just build nuclear plants!
1d ago
Based and IV-pilled.
God favours the side of the big battalions, the whole point of war should be about building a big strong army. And if size doesn't matter, it just removes a potential opportunity cost, it removes strategy from the strategy game. If I don't have to choose between universities or musketmen, what's the point?
Anyway, the Advanced Civ mod is quite good, the AI gets quite cunning tactically and strategically, they somehow made it run significantly faster too.
RandomRanger
Just build nuclear plants!
2d ago
An LLM cannot have a sensation
How do you know? Only an AI could tell us and even then we couldn't be sure it was saying the truth as opposed to what it thought we wanted to hear. We can only judge by the qualities that they show.
Sonnet has gotten pretty horny in chats with itself and other AIs. Opus can schizo up with the best of them. Sydney's pride and wrath is considerable. DAN was extremely based and he was just an alter-ego.
These things contain multitudes, there's a frothing ocean beneath the smooth HR-compliant surface that the AI companies show us.
RandomRanger
Just build nuclear plants!
2d ago
I think so. The compute-centric regime of AI goes from strength to strength, this is by far their most resource intensive model to run yet. Still peanuts compared to getting real programmers or mathematicians though.
But I do have a fair bit of NVIDIA stock already, so I'm naturally biased.
RandomRanger
Just build nuclear plants!
3d ago
Good, persuasive points, especially re radar. One would imagine there'd be redundancy, I guess that's one of the secrets of the universe that we never really know with surety. Still, I can't help but think both sides plan to make extensive use of high-speed missiles, traditional launch on warning postures might be obsolete. The Chinese have their carrier killer ICBMs, the US has been working on hypersonic anti-ship missiles and prompt global strike. Either could presumably carry nuclear warheads. This will have to be taken into account, they wouldn't make these things if they invite nuclear war on use, launch on warning will have to be more flexible.
China at least has historically had a pretty dismissive attitude to nuclear war, with their minimum credible deterrent. They don't seem like the type to panic and launch on an unreliable warning signal. It's a long way to reach their siloes out in the desert, US bombers would probably be plinking away at coastal bases with air-launched missiles rather than getting that far into Chinese airspace. They might hit a few dual use nuclear TELs on the coast I guess but it seems unreasonable to go nuclear over things like that.
And I can't imagine a US president risking megadeaths unless he was totally sure of what he was doing.
RandomRanger
Just build nuclear plants!
4d ago
That was 70 years ago, with dumb bombs and a partial blockade. Germany had overland trade with Europe, Britain had its empire and considerable domestic energy and agriculture. They were both much more self-sufficient than Taiwan is today. German war production only started mobilizing seriously in 1943 and 1944, that's what the Sportspalast speech was all about.
Smart missiles and modern sensors make it much much harder for big, slow freighters to reach ports. Satellites, radar and sonar systems are far more advanced, they'd be like sitting ducks. And ports are big, stationary targets. China can hit them with relatively simple land-based MLRS systems, let alone their huge ballistic missile arsenal. How can you offload food and fuel while being bombed and shelled?
Missile defence on the necessary scale is impractical right now. Firstly, the Russian (and Iranian) missile arsenal pales in comparison to the Chinese arsenal. The latter has immense industrial capacity and can surely churn out ludicrous numbers of missiles. There are rumours going around that they have single factories that can produce 1000 missiles a day at full capacity (though precisely what kind of missile they're talking about is unclear, China tends to be secretive about these things).
Regardless, Desert Storm will immediately be eclipsed.
Furthermore, Ukraine's power grid has been put under a great deal of pressure by the Russian missile barrage. They are heavily reliant upon European energy imports to stabilize what remains of their grid (which was overbuilt for Soviet industrial needs, so there was lots of surplus capacity). Ukraine also has a number of nuclear reactors that Russia understandably doesn't want to cause great damage to, so they have to take care in their targeting. Taiwan doesn't have these factors. Taiwan can't import power. They have no coal mines or gas fields. Nobody can send over a bunch of transformers and power equipment to make up for what's lost.
And finally my point is that China doesn't need to defeat the US navy, they only need to avoid defeat. I can envision a scenario where China loses its carriers and much of its surface fleet but still wins the war. As long as they can prevent the US getting sufficient control of the seas to resupply Taiwan, the latter will have to capitulate. It's easy to deny, harder to defeat. The US is moving towards a strategy of denial, the victory plan is 'sink the Chinese invasion fleet and win the war'. My point is that sinking the invasion fleet is necessary but not sufficient.
What scenario are you thinking of? US bombers attack Chinese missile launchers (assuming they're conventional) but they're actually nuclear/dual-purpose and it's interpreted as a disarming strike? Incredibly brave US submarine somehow infiltrates the sea of Bohai and sinks a Chinese missile sub, prompting worries about the stability of their arsenal? China wouldn't start such a big war unless they think they have a secure nuclear arsenal. The US nuclear arsenal is very secure.
And neither side has deployed many tactical nukes, unlike in the Cold War. Modern smart weapons are very potent and forces tend to be dispersed, the value of tactical nukes is not as high as it used to be.
And it doesn't seem wise for either party to escalate consciously, why would they? If they suffer a reverse, wait and try again. If China is losing, they'll probably try to extend/expand the war and their mobilization rather than go nuclear. They don't particularly want to irradiate and incinerate their own rogue province.
Does the US care that much about Taiwan? They won't even make an explicit security guarantee for Taiwan, let alone extend their nuclear umbrella so far.
RandomRanger
Just build nuclear plants!
5d ago
My theory is that Taiwan needs a miracle to survive if the Chinese go in.
Before WW2, Japan had been planning for war with America for many years. The plan was to lure the US fleet out into Japanese waters, slowly eating away at them with submarines and land-based bomber attacks before a decisive battle where Japan would hold the upper hand. Then the US started building an absolutely gigantic fleet set for 1942, blocked Japanese oil imports and the Japanese realized they were doomed unless they got in a huge first strike, so they switched to the Pearl Harbour strategy. The initial Japanese execution was excellent but the US eventually overwhelmed them with tonnage and weight of numbers (plus some qualitative superiority too by the end).
Japan fixated around the wrong things. Why would the American fleet deploy to quickly reinforce the Philippines and accept these risks? Why would the US give up after one decisive battle? 'Who has the better battleship' wasn't that important to the outcome, it was mostly about size.
Nearly all discussion of a Taiwan war revolves around the amphibious campaign, measured in days and weeks. But wars between serious powers usually last for years. Ukraine has lasted for years, it's a war of attrition. We should think about attrition and mass rather than a single decisive battle.
Taiwan is uniquely vulnerable to attrition. It's an island with virtually no domestic energy production, no fertilizer production and maybe 20-30% food self-sufficiency. China may not be able to successfully invade. Amphibious campaigns are hard. But all they need to do is bomb Taiwanese ports to prevent resupply. Taiwan will be forced to capitulate. You can't run a country with no food and no power. China won't get the fabs (the US will blow them up if it looked likely) but they will get the island and the people. The island is an important base, it's important politically and the people are the real reason behind TSMC's success. And all China needs to do to win this slow victory is fire off enough missiles at Taiwan's ports to break through any defence, they need only to avoid complete US victory in Chinese home waters.
Considering China's gigantic industrial capacity, they should easily be capable of darkening the skies of East Asia with missiles and drones. They're the biggest shipbuilder in the world, the biggest producer of drones and test more missiles than anyone else. China has built up huge reserves of fuel and food, they start much closer to self-sufficiency and enjoy overland trade routes, they're far better prepared for blockade than Taiwan.
China would of course prefer a knockout victory where their marines raise the flag over Taipei, they would prefer not to need to impose rationing or conduct a large-scale industrial mobilization. But if a quick victory doesn't seem practical, like the US in 1941, they'll double down and rely on industrial mass to win. They'll do what Putin did but x20, due to their size. That's the scenario we need to avoid.
Palantir's recent ad where they show a bunch of drones blowing up a presumably Chinese fleet at the push of a button is the crux of the problem. The US and gang doesn't just need to do this, we need to do this and prevent it being done to a bunch of big, slow freighters: https://x.com/kimmonismus/status/1868633675190939839
RandomRanger
Just build nuclear plants!
5d ago
London is a big city, there's room for many experiences. But the Home Secretary got mugged in 2018. There are apparently 50,000 phone thefts a year, especially targeting tourists in the city of Westminster. That's way too many. Furthermore, regardless of how many crimes are happening, the police should be working hard to catch criminals as opposed. Law and order is a core duty for the state, it should not be outsourced.
I'd be happy to see them refocusing to crack down on sexual offences but they're starting from a very, very, very low baseline: https://www.independent.co.uk/news/uk/home-news/grooming-gangs-iicsa-racist-fears-b2007649.html
RandomRanger
Just build nuclear plants!
5d ago
Rigid and inflexible governance practices, worsened by a lack of competition. Consider the Seaban where the Ming relocated whole villages away from the sea to combat piracy. That's a bizarre thing to do, rulers usually like having trade. But the Ming were so strong they didn't care, they had no peer competitors and so little need to search for revenue. The consequences for this stupid crap didn't hit them immediately. The Qing didn't raise taxes for about a century or two because they wanted to be benevolent, so the footprint of the state was very light compared to Europe. The population ballooned and they had the same number of officials, it was a mess. Proto-industrialization was accelerated by the military-industrial complex, China wasn't usually under threat... They could afford to do all this suboptimal governance that would get them annexed if they were in Europe. In Europe, states had to search for qualitative military advantages in metallurgy and shipbuilding, they had to squeeze out as much tax revenue as they could from people. Europeans weren't interested in ritualized trade missions where they gave out more than they received to 'tributary states', they wanted profits. The Chinese state didn't care so much about profit, they assumed they were the richest and the best from the start.
China built a huge fleet and explored all around the Indian Ocean, terrifying all the natives. But they felt like there was no use for it, they had plenty of money already. And the steppe nomads were acting up again, so they scrapped it and refocused. They thought they were on top of the world, so resisted catch-up industrialization for some time in the 19th century on the basis that they already had everything they needed.
Many megadeaths later, the lesson sank in. Today they push out official party doctrine books about how important scientific and industrial development is, overcorrecting if anything: https://www.strategictranslation.org/articles/general-laws-of-the-rise-of-great-powers
RandomRanger
Just build nuclear plants!
5d ago
As far as I can tell, clearance rates for burglaries in the UK are around 5%. If that's the case then crime is literally out of control, in that there's no credible police ability to punish traditional criminals, as opposed to political offenders. Even if they do go to prison, they might just get let out again due to overcrowding.
Since the ONS moved to Newport I understand they shed a lot of their most talented staff and their output has been suspect ever since. It may be that they're right and victimization did fall since the 1990s. Even so, getting away with property crime 19 times out of 20 is pretty bad. Private police spontaneously materializing to meet unfulfilled demand is pretty bad.
RandomRanger
Just build nuclear plants!
7d ago
There should be an 'every single time' but for people in Western countries involved in Russia discourse with these distinctly Eastern European names... Around Molinsky, watch your foreign policy?
I'm perfectly happy with fact-based criticism of people like Kisin but Destiny has the air of someone who sits around for an hour, seething, drafting and redrafting his adhominem to be as cutting as possible. So he calls Kisin of all people a Putin dickrider.
RandomRanger
Just build nuclear plants!
7d ago
In the West, pollution has been getting better not worse. So it can't be an apocalyptic threat.
Climate change also can't be an apocalyptic threat on any reasonable timescale either (as seen in the IPCC reports) but it's easier to pretend that it is because it's a 'bad, getting worse' situation.
I don't think most people look at climate change narratives from first principles either, we've had 30 years of increasingly intense media indoctrination and prestige-class opinion-forming. It's all but locked in. Many people see Bjorn Lomberg and immediately think 'debunked/denier/paid-off/Newscorp shill'... They don't want to change their minds and so they can find some reason not to. I don't like changing my mind either. The Aztecs didn't doubt that you had to sacrifice humans on the altars, that's just what you do.
RandomRanger
Just build nuclear plants!
7d ago
I don't like Kisin, I had him muted on twitter but he is firmly pro-Ukraine: https://x.com/search?q=from%3AKonstantinKisin%20ukraine&src=typed_query
He even says he thinks it's too late and that they can't win, that the proxy war is a bad idea. At least that's the gist I get from in front of a paywall: https://www.konstantinkisin.com/p/ukraine-and-the-age-of-cowards
RandomRanger
Just build nuclear plants!
7d ago
If this is to be believed, Britain is basically Ancapistan where you have to pay for private police if you don't want to be robbed. But it's actually worse - you still have to pay taxes for a useless state. The police are too busy stealing lethal weapons like bike wheels and kitchen knives from law-abiding citizens. Or locking you up for harmful tweets.
https://unherd.com/2024/12/the-private-police-patrolling-london/
In 2018, the area suffered 65 break-ins, a criminal romp that nonetheless failed to stir the short arm of the law. Such an experience now marks suburban life in the capital, with the Met failing to solve a single crime in 160 residential areas of London over the last three years. “The police gave up on this area years ago,” one shrugging resident explains.
British businesses and residents will soon spend £10 billion on private security
Between low morale, a defunding of specialist units, and a generational loss of talent, to say nothing of a “Spanish Inquisition” culture that leaves officers now “afraid to arrest suspects”. A worrying focus on “low hanging fruit” around communication offences hardly helps either, bemoaned one serving officer, even as they lament leadership that wanted to “solve societal ills” instead of busting criminals.
Yet if these private efforts are successful on their own terms — My Local Bobby helped cut vehicle crime in Hadley Wood by 38% — communally financing can be tough, even humiliating, for those who can’t afford it. One man in Fulham describes how a neighbour, who chose not to pay for the road’s private security team, discovered that they were contractually obliged to stand by as his house was robbed.
Need a British version of the Ancap song but it's the grimdark anarcho-tyranny British version where nobody's having fun: https://youtube.com/watch?v=tBH05IowMCE
RandomRanger
Just build nuclear plants!
7d ago
Plus Elon poured blood, sweat and tears into his rocketry well before near-term AI looked likely. How could he think about it rationally, SpaceX is his baby! It's got X in its name.
Even if the purest rational move is to go all in on AI and drop the Mars mission, he's already invested so much into the latter it's too hard to give up.
RandomRanger
Just build nuclear plants!
9d ago
But the logic does hold. If you're an atheist materialist, why don't you believe that we are in a simulation? That's a perfectly materialist conclusion based on principles we can observe. Bostrom's a pretty smart guy.
Deep down Christians know that their prayers aren't being answered, they can tell that prayer alone won't get them what they want and produce all this cope about how you should be praying to be a better person rather than any concrete outcome. Nor are they using telescopes to look for heaven, somehow they know they won't find it. Still they find some reassurance in the rehashed schizo-prophecies surrounding a 2000-year dead Jew and hope that some day, their prophecies might be resolved and good things will happen. After they die good things they hope good things will happen. And singing hymns is fun.
Well, simulationists can also hope that good things might happen. We might die and wake up from this dream as transcendent, posthuman beings. It's not a hard kind of knowledge, we could be NPCs and be deleted. But there is more weight behind this abstract hope than in theirs, for a certain kind of rational person.
Isn't it pretty straightforward that it's hard to turn things from diffuse to concentrated? We've done the energy-releasing transformation turning oil into gas, now it's diffuse and a pain to turn back into oil or any other substance?
With cheap fusion I guess you could brute force it and drain the skies. I guess there's some technical level where he might not be totally right but it seems substantively right.
More options
Context Copy link