Sun Tzu teaches us that war is an objective phenomenon with laws that govern its prosecution. So does Napoleon, Mao Zedong and any other successful strategist. Sun Tzu also taught: War is won by the side which calculates better. The side which makes no mistakes wins without fighting at all. The side which makes fewer mistakes wins when they must fight.
In the Ukraine war we witness many miscalculations. The miscalculations of the USA 20 years ago in over-reacting to a terrorist attack, seeking to remake the mid-east, consequent miscalculations of Putin in similarly launching a series of wars, following Hitler's playbook -- the big lie, and level up conquests like some sick platformer. Chechnya, then Georgia, Crimea, then Donetsk and Luhansk and then finally what remained of Ukraine. Navalny also miscalculated thinking he could return to Russia (Muscovy) and would be another Mandela or Gandhi. It's like no one else notices the Bikos, Lin Biaos, Leonard Peltiers or Mumia Abu Jamals of this world. To me it was obvious the instant Navalny set foot in Russia that he was a dead man walking. Prigozhin also miscalculated. The penalty for error in war is death.
Dialectical materialism teaches that a long series of incremental quantitative changes over time eventually result in a much larger and broader quantitative change, and that "catalysts" can accelerate this process of changes. Revolutions are like that: everyone will seem loyal to Putin, but slowly and gradually he will lose support here, sponsorship there, friends here, allies there too are lost, and suddenly one day all those former friends and allies turn into actual enemies and then --no more Putin.
Putin will appear to enjoy absolute power and unlimited privilege until one fine day when he suddenly and unexpectedly will enjoy neither power nor privilege or anything else for that matter. His retirement plan is his coffin. Dialectical change is sudden, unexpected, qualitative, and can be "catalyzed".
In war lots of people will miscalculate. Most do not know all of history, geography, military science, culture (religion, languages, food, art) economics. Thus various people and groups will take crazy positions. They are setting themselves up for a future which will not happen, because they think it might or is even likely. Sometimes their miscalculations are wishful thinking, or even magical thinking, mostly its due to a lack of deep thought on a broad range of topics (history culture languages geography economics law climate military science). The war is already decided and China knows it. Most of my strategic thinking is Chinese (not all of it). Ukraine should start positioning for a post-war world as China's granary. China fears the USA, wrongly, but the promise of limitless low cost grain will greatly motivate the CCP.
Anyway, they miscalculate for a variety of reasons (dualism is another one) and consequently make stupid mistakes which are often fatal, a self-cleaning oven so to speak. Basically 90% or more of the people ("the masses" 老百姓) are ignorant of all the above things, not to forget the fact we all suffer from cognitive biases and psychological challenges. 1% of analysts have all the tools needed for competent accurate predictions. Given the fact of limited knowledge people, probably all people, make mistakes -- but they make mistakes to varying degrees. I'm not infallible I am just less blind than most. No one can have it all, e.g. i have no hebrew, arabic, turkish, no one person can do it all but yeah one person can in fact learn enough to make accurate predictions and strategic decisions to describe and influence the course of any given war. The Jewish Israeli version of "me" obviously Would make much better predictions about Mid East politics, culture is that important, languages matter and so do religions. However, the Jewish Israeli version of me would probably be half blind, over-reacting and under-reacting, to various European fascist movements. Sometimes it's like we are looking into a mirror, and of course the image is reversed, but maybe there is hot steam from the shower clouding our view. The Japanese version of "me" would be much better at Thailand for another example of why regional experts are always needed for the most accurate analysis of international politics.
The real object of Putin’s war was to split Germany and the USA and draw Germany and Russia together. It failed. That is utterly clear as shown by the documents Scholze just signed. A secondary goal was rebuilding the Russian empire. Putin had his priorities backwards. And his methods, repulsive and ineffective, have driven away any hope of Chinese support. Killing Navalny lost the war for Putin. Which is stupid, no one person is that important, but the various elites figure "shit, if Putin will kill someone already in prison for life how long till his government collapses and when am i next on his hit list?"
You may ask me anything.
NAVALNY The audio was speech synthesized using piper, a trainable neural network which can be used to generate deepfake audios. I had to use a medium and not high quality voice, because of processing time on my old disposable computer. Later works should feature Putin voiceovers, though I currently don't have the processor capacity needed for training a neural network using python (q.v. whisper).
Remember when I predicted Ukraine would claw back air parity and then air superiority over Ukraine?
-6
I say! I am in the mood for a good old fashioned Train Bombing!
"Where will we get the money" she said
“Told ya i was gonna eat some airplanes”, I said.
One-eyed beautiful ballerina mom civcas might sound interesting until you understand: it's not fiction. If you want true-life horror stories you can find many of them in Ukraine.
I don't just make shit up, in other words lie. That's also why I prefer non-fiction writing.
As predicted here:
Neural Network Large Language Models (AI)
https://perplexity.ai
like we can plainly see the back rooms are powered by python, because of course they are.
What happens in the back rooms?
Sabine Hossfelder is Clickbait scaremongering.
stfu, dnw.
CIA has had predictive political event AI models for at least five years now. AI was generally greeted differently in various quarters, whether with enthusiasm (analysts), skepticism (case officers), or indifference (operators). It's driven from a very very large database, trillion(s) of tokens. Parts of it are neural networks, other parts are expert systems. It's not merely a large language model- Llms have real limitations on visual processing, it uses various modules for different functions. So if, some case officer has obtained great graphics for processing they would of course be well processed, this is such a basic intelligence function. Unit Differentiation has some operational security benefit, but the usual over-confidence and naive blabby mean I expect parameters or even sources of the AI will wind up leaked by idiot insiders or stolen by effective foreign spies, because of course that will happen given past performance.
This is the first sabine hossfelder video I ever watched and I am entirely disappointed. I don't write about things I don't know about. Others should be so wise.
FWIW I have been working on computer science and wargames for the last 40+ years. Again, I've no idea how much expertise Hossfelder has on either of these topics but an llm is only going to do a certain, limited, amount of tasks. So ... this video looks like stupid scaremongering.
I know what is effective and ineffective and what a fucking surprise I am not a blabbermouth.
I also don't know how much game theory or operational research Hossfelder has studied. I mean, if she were at all serious she would have reproduced the prompts.
A basic principle of computer science is: the program produces whatever you programmed it to produce. Neural networks don't change that, they just make the decision processes much tougher to trace and use search space which cannot realistically be procesed by one individual. If you want the llm to escalate rapidly and use nukes you program it to do so, if you don't you don't "the model is not the terrain".
So why are CIA’s AI models useful?
Suppose you have someone like me who actually knows what he is talking about (look back on my strings of predictions which most thought unlikely yet occurred) you then take that person and have him set up the paramaters and add lots and lots of documentation and then you can give a product to company and platoon leaders who will be highly fatigued, physically exhausted, sleepy, wet, cold, tired, hungry, and getting shot at and now they are making correct decisions 90% of the time or more instead of 60 to 80 percent, depending on their physical state, training, intelligence, and leadership qualities.
For example, let's take Korea. Suppose I want to predict what will happen in Korea when the current dictator rolls over and finally dies of obesity. When is he likeliest to die? This one is rather easily predicted, but not by an llm ai. Another question: Will his drug addicted sister take over? If so, will she seek policies of conciliation, liberalization, and reunification? Or would she instead double down on juche ideology, and continue to play "bad cop" to China's "good cop"? These could be addressed by an llm ai: they are the sort of general question a properly trained and prompted llm could in fact answer better than I since it does speak Korean unlike me and my Japanese really sucks. In contrast, if I want to determine rates and targeting of North Korea's artillery park, this is something that would be better treated by diffusion or an expert system. Different tools for different tasks, no tool is perfect, they are tools not magic.
Learn Languages
Duolingo sucks. Don't waste your time on it. Instead use 50 languages, which is also free, but actually works.
Before you dismiss this as crazy or unrealistic Shenzhen really once was a mere fishing village.
Before you bitch about Xinjiang
Turning a salt water lake into fish farming.