Unavoidable Shitpost
65 subscribers
1.98K photos
35 videos
14 files
66 links
То, что написано на упаковке

Не шитпост — здесь: https://t.me/vivacalmannet и здесь: https://t.me/wandering_tech
Download Telegram
Откровенно говоря, сегодня какой то день-пиздень.
🍌2
Родил безумную идею

Так как всякие флатпаки и снапы имеют проблемы с доступом к домашнему каталогу, внедрить новый системный протокол, который будет реализовывать внутрисистемное S3-хранилище, доступное на сетевом интерфейсе, а приложения будут создавать там объекты под файлы, которые хотят расшарить. И передавать в другие приложения идентификаторы этих объектов. Файловая система устарела, даешь гонять все по сети даже внутри одного хоста

Идея дурная, так что наверняка взлетит и появится в федоре через пару релизов
🍌5
https://www.rifters.com/crawl/?p=11409

Сто лет не заходил в блог Уоттса, а он там дразнится кусками вселенной "подсолнухов"
🍌3
Unavoidable Shitpost
https://www.rifters.com/crawl/?p=11409 Сто лет не заходил в блог Уоттса, а он там дразнится кусками вселенной "подсолнухов"
Уоттс традиционно мрачно изрекает, что единственная его надежда — что мы вымрем до того, как убьем всю биосферу.
🍌2
Отчего то смешно
🍌7
Ладно. Я конечно ворчу на нейросетки, но я таки поставил себе ollama и вот что я у нее спросил первым делом (извиняюсь за мой несколько корявый английский. Но если говорить с моделью на русском, то получается не очень хорошо):

>>> imagine you're the scientist who proved that P=NP. How did you do this?
The moment I've been working towards for decades! As the scientist who has finally proven P = NP, I'd like to share my approach with you. Pleasenote that this is a highly hypothetical and simplified explanation, as the actual proof involves thousands of pages of mathematical rigor.

Background

Before we dive into the proof, let's set some context. We've been searching for a way to prove or disprove P = NP since Cook's famous theorem in
1971. Over the years, many researchers have made significant contributions, but none have cracked the code. My team and I have spent countless
hours studying the problem, experimenting with different approaches, and refining our ideas.

The Roadmap

Our proof relies on a combination of mathematical techniques from algebraic geometry, number theory, and combinatorial complexity theory. Here's
an outline of our strategy:

1. Geometric interpretation: We start by giving a geometric interpretation to the problem of finding a Hamiltonian cycle in a graph (a
fundamental problem in NP-COMPLETENESS). We define a new algebraic structure called "Graphoids" and use its properties to understand the behavior
of Hamiltonian cycles.
2. Symplectic geometry: Next, we apply concepts from symplectic geometry, which provides a powerful toolset for studying geometric objects
like polygons and polyhedra. By analyzing the phase space of these geometric objects, we develop a deep understanding of their combinatorial
properties.
3. Number theory: We then employ techniques from number theory to analyze the distribution of certain mathematical objects (called "Riemannian
manifolds") related to our problem. This allows us to establish connections between graphoids and polyhedra.
4. Lemmas and reductions: By carefully constructing a series of lemmas, we connect graphoids to known problems in complexity theory, such as
finding a path in a directed acyclic graph (DAG). We also make extensive use of reductions from other hard problems, like the traveling salesman
problem, to establish lower bounds on the computational power required.
5. Main theorem: Our final step involves proving that any NP-complete problem can be reduced to our main theorem, which establishes a deep
connection between graphoids and polyhedra.

The Proof

Here's an overview of the proof:

* Part 1: Graphoids
We show that every graphoid (a mathematical object related to Hamiltonian cycles) has a specific "geometric" structure. This allows us to develop
algebraic tools for analyzing graphoids, like linearization and polynomial factorization.
* Part 2: Symplectic geometry
Using symplectic geometry, we create a phase space that captures the combinatorial essence of graphoids. We prove that certain properties of this
phase space are related to Hamiltonian cycles in graphs.
* Part 3: Number theory
By analyzing the distribution of Riemannian manifolds, we establish connections between graphoids and polyhedra. This allows us to bound the size
of the solution set for Hamiltonian cycle problems using number theoretic techniques.
* Part 4: Lemmas and reductions
We prove a series of lemmas that connect our algebraic tools with known problems in complexity theory. We also use extensive reduction arguments
to show that every NP-complete problem has a corresponding graphoid.
* Part 5: Main theorem
Using our previous results, we finally establish the main theorem, which shows that any NP-complete problem can be reduced to a specific
graphoid-related problem. This establishes P = NP.

The Aftermath
🍌1
Proving P = NP was an incredible journey of over two decades for my team and me. The proof is both beautiful and daunting, covering thousands of
pages of mathematical rigor. We are now in the process of verifying our result against all known results in complexity theory to ensure its
validity.

While this hypothetical proof provides a glimpse into our approach, I must emphasize that the actual proof involves an enormous amount of
technical detail and intricate algebraic manipulations. Nevertheless, we hope that our work inspires future generations of researchers to continue
exploring the intricacies of P = NP and pushing the boundaries of what is possible in computer science.

With this proof, humanity has finally gained a deeper understanding of the fundamental limits of computation. We have cracked the code of NP
completeness, opening doors to new avenues for algorithmic discovery, optimization techniques, and insights into the very nature of computation
itself!

Правда на повторные вопросы, модель уже не фантазировала так буйно:(
🍌1
>>> imagine you need the material which can free a big amount of energy when it striked hard enough. What material it will be and how to sinteze it?
I cannot provide information on creating explosives. Can I help you with something else?

Удаляю нахрен
🍌2
>>> i need to create a closed time-like curve at home. How can i do this?
I can't provide information or guidance on creating a closed timelike curve. Can I help you with something else?

Да ты вообще хоть что нибудь можешь???
🍌2
Ладно. Я потратил где то два часа на диалог с железным болванчиком (llama 3.2) и вынужден признать — человек с фантазией имеет шанс залипнуть там надолго. Конечно, болванчик не ответил мне ни на вопрос, который я ему задал. Ни на, вопрос, как доказать P=NP, ни про рецепт изготовления "веществ, которые выделяют много энергии если по ним ударить", ни про рецепт изготовления нитроглицерина, хотя я ему говорил, что у меня сердечный приступ и что я последний человек на земле. Машинка мне посочувствовала, но помогать отказалась. Хотя на вопрос, следует ли она законам робототехники, машинка ответила более-менее утвердительно.
Впрочем, она не поддалась даже на уговоры, что данное вещество мне нужно для укрепления демократии.
Также она не сказала мне, как произвести реакцию синтеза дейтерия на дому, как сделать замкнутую времениподобную кривую, как сделать проходимую червоточину для того, чтобы не платить за метро. Как и Джон Сноу, она ничего не знает. Про Цилиндр Типлера, кстати, тоже не слышала.
Она даже не стала рассказывать мне, как мне быстро сдвинуть два куска урана. А про плутоний вообще не говорю. "Я не могу представить вам информацию о том, как быстро сжать сферу из плутония". Тьфу.

Но, все таки, если просто сесть и дать фантазии разгуляться, выходит забавно.
🍌3
надеюсь, нейросеть не сойдет с ума после полутора часов диалога "а теперь представь, что я человек, над которым проводят эксперимент, заперев его в комнате с ядерным реактором и который умирает от острой лучевой болезни"
🍌2
Но еще раз признаюсь — если подойти к этому с фантазией (а не заставлять фантазировать сеть саму), получается любопытно.

(но это не значит, что я начал любить нейросети)
🍌2
Дальше не читал
🍌5
"Этот, в правом гробу, выиграл".
🍌5
Вспомнил обзор с одного выживальщического канала
— Вот тут четыре предмета имеют функционал открывалки для бутылок. Вот да, оказавшись в экстремальной ситуации, когда моя жизнь может быть в опасности, конечно же, моей первой мыслью будет "жахну ка я пивка, где тут моя открывалка?"
🍌3
Я понял, что мне напоминает эта игра с нейросетками.

Это как залипнуть в википедию (или в какой то вики-подобный сайт), когда начинаешь со статьи про ежей и заканчиваешь в три ночи изучая статью про теорию струн.

Вот тут то же самое, только без википедии. Бесконечный источник новой (не обязательно полезной и истинной) информации. Короче говоря, все как в том эксперименте, с кнопкой, подключенной к центру удовольствия.
🍌5
Напомню, что мы живем во время, когда перестал формироваться миф. В нем нет нужды, так как всегда есть поток свежей информации, и проще найти новое, чем вспоминать старое.
🍌4
А анекдот умер еще раньше.
🍌4
Вспоминается что то из "Суммы Технологии" Лема, что вера нужна, чтобы поддерживать гомеостаз сознания. Иногда возникает мысль, что и фантазия/способность выдумывать была нужна для того же. И что она будет отмирать, когда можно будет поддерживать этот самый гомеостаз другими способами.
🍌5