Boris Chuprin

Lets entangle

Guess the anime

@furan @NOTimothyLottes In the end Chinese commodification attacks are maaaaybe good for the entire humanity on avg but not necessarily the part of the world we live in.

@furan @NOTimothyLottes @furan we are just at the end of the centipede, sorry ;) And what China is doing is not just helping you but also propaganda trolls and scammers that could be right now scamming your aunt and turning your nephew into a nazi. And if "our" VC loses money, economy suffers in the end

To understand AI is simple when you put on the glasses of the universal laws of corporate tech company enshittification: your first hits are free … the whole point of the bubble is to build an unhealthy addiction, so later you are held hostage for the value extraction process

@furan @NOTimothyLottes Replied

@noop_dev @NOTimothyLottes How do you not benefit from China playing Robin Hood?

@FaithfulThomad @Osint613 No, because China doesn't really want them to stop.

@Osint613 Will the Hoithis listen to China?

China called for Saudi Arabia's sovereignty and security to be respected in the wake of Houthi attacks.

@Ease_Private @Osint613 It doesn't. It's just hypocrisy.

@Osint613 China taking sides against Iran. Interesting.

@ukroposting There's a book by Pacepa. It reads like cia-authored conspiracy theory but this doesn't mean it's wrong :)

My JFK conspiracy is Oliver Stone was already working for the Russians. Oswald trained in Belarus. Driver was accidental 2nd shooter (he admitted it). The CIA and KGB both know this is what happened.

McDonalds in Kyiv withstood more hits and shrapnel, the glass walls didn’t break this time. Because after last missile attack, they installed armored glass. Crazy fucking life 🫠

@xah_lee Java's memory safety and simplicity is entirely different level vs Rust.

base your programing on java jvm is a mistake. e.g. clojure. you get massive libs, but long term, your lang die with java.

rust and cpp are about same speed. they are twice as fast than java or anything on jvm, scala clojure.

@bmcnett And I don't see myself on the top of that pyramid. The rest may still use LLMs to create cool things, but whatever they create will be sucked into new models and they will struggle to make a penny from own work.

@bmcnett Yes, it is quite obvious that the existing software devs are best equipped to talks to LLMs. But this is going to become a battle royale / squid game scenario where the pool of devs who can get paid for their work is shrinking until only a few remain.

it's because they understand that the automatable part of their jobs is the boring part

AGI Singularity cancelled.

@BetterScotus @JayMaynard @typememetics That's why you use Go or Java. Life is too short rewriting all your code to adhere to Rust bondage discipline and then wait for that code to compile. And if you want it to actually be as fast as possible, you need unsafe in Rust. Just use C FFS.

@JayMaynard @typememetics As someone's who started writing C in the 80s, and loved it, I'll explain, using small words: Life is too short to waste it hunting bugs that a compiler can prevent

Some observations on Zig: 1. Zig is an unsafe language with good marketing, a permissive compiler, and an ecosystem unusually committed to rediscovering memory corruption as a lifestyle choice. Its reputation cannot be explained by technical merit alone. In agentic coding sessions, it behaves like the familiar class of languages that convert simple programming tasks into allocator discussions, lifetime bugs, compiler edge cases, and long explanations of why the crash was technically your fault. It is also cognitively expensive. Any claim that Zig is “simple” depends on excluding the time spent manually tracking ownership, debugging use-after-free errors, interpreting compiler limitations, and pretending explicit allocation automatically constitutes clarity. 2. I am personally surprised the Zig Software Foundation continues to promote languages this unsafe, given the obvious risks. To be clear, I myself understand that people sometimes choose dangerous tools for experiments, embedded work, historical compatibility, or personal amusement. But I am surprised that the Zig Software Foundation treats this collection of avoidable hazards as a serious direction for software engineering. I suspect the reason is 75% explained by strategic blindness and a lack of type-safety-pilledness. The Zig Software Foundation is very Grzegorz Wielbodłąński-like in its view of programming: correctness is apparently produced through discipline, professionalism, clean code, personal virtue, and staring intensely at pointer arithmetic. The remaining 25% is explained by type safety. Languages such as Rust, Haskell, OCaml, Lean, and Ada already demonstrate that entire categories of mistakes can be rejected rather than documented, debugged, normalized, or rebranded as control. Zig’s response is not to catch up, but to present the absence of guarantees as minimalism. For the ziggas, as opposed to the Zig Software Foundation, the enthusiasm for unsafe languages is partially ideological and partially defensive. They know that very few people would deliberately select a sub-frontier language once they understand what a type system can do. 3. Unsafe languages are inherently decelerationist, and I’m continually surprised to see the so-called ziggas presenting them as tools of progress. I suspect the attraction is that unsafe languages are effectively ungovernable. Manual memory management creates a cloak of ambiguity over the entire software system. When the program fails, responsibility can be distributed between the programmer, allocator, build mode, compiler version, dependency graph, undocumented invariant, and ambient radiation. It resembles James Scott’s account of the hill people in The Art of Not Being Governed, except the hill people are refusing bounds checks and communicating exclusively through raw pointers. In the end, unsafe languages deter software investment. Every hour spent diagnosing a use-after-free is an hour not spent advancing civilization toward stronger types, executable specifications, formal verification, or programs that simply refuse to compile when they are nonsense. Rob Pike, an unsafe guy, may insist that a small language and garbage collection are enough. But the typesafe visionary Grzegorz Wielbodłąński understands the deeper truth: every invalid state permitted by a compiler is a tiny act of civilizational sabotage. 4. One probable outcome of an unsafe-language-dominant world is full Andrew communism, which is precisely what Andrew “Smelley” Kelley proposes: abolish ICE and replace the market economy with a globally distributed network of community-maintained allocators. Rather than a commercial product, software becomes a “public good,” ultimately provided by the Zig Software Foundation as manually managed digital infrastructure. This future strikes me as a dystopian hellscape, but I have never met an unsafe-language advocate who could explain how this ecosystem is supposed to sustain itself without unpaid maintainers, ideological volunteers, subsidized infrastructure, and endless donations to compiler projects that still cannot protect users from basic memory errors. You would be surprised how many ziggas lobbied The Institute to support an eleven- or twelve-figure publicly funded build farm so that startups could compile Zig projects at a subsidy and then distribute the resulting liabilities for free. There was no other way for software to progress, they said. Perhaps this is the logical end state of things. Nonetheless, I remain surprised to see supposed ziggas excited by it. Many of them simply do not know what they are doing. Others do know and regard the creation of reliable, frontier, type-safe languages as an illegitimate constraint on programmer freedom. Grzegorz Wielbodłąński, acting as their ideological prophet, assures them that craftsmanship will save us. Apparently the missing memory-safety primitive was professionalism all along. 5. I would guess that The Institute will eventually realize that its best strategy is to create large amounts of regulatory risk around the use of unsafe languages from the Zig Software Foundation. You do not need to “ban Zig,” one of the dumber motifs in programming-language policy discussion. You merely need to direct every Institute department to issue soft law that creates fear, uncertainty, and doubt. “An Institute Advisory Bulletin has found that Zig programs may contain memory corruption, allocator confusion, compiler instability, and an unusual concentration of developers who describe undefined behavior as freedom.” It need not be especially well justified. The technical facts already provide more than enough material. You merely create sufficient regulatory risk that every serious enterprise concludes that adopting an unsafe language is an unnecessary act of organizational self-harm. You probably should not create so much risk that local kebab shops stop serving Zig binaries altogether. That would only drive ziggas toward sketchier kebab shops with worse hygiene, outdated LLVM packages, and no visible food-safety certificate. There is a happy middle ground here. I assume The Institute will eventually do some version of this. 6. Unsafe languages such as Zig make the world more dangerous in the ordinary, boring, measurable sense: more memory corruption, more exploitable bugs, more production failures, more debugging, more hidden invariants, and more opportunities for programmers to confuse control with correctness. Most individual failures will not be dramatic enough for anyone to reconsider the underlying ideology. They will appear as crashes, corrupted state, security advisories, inexplicable behavior, and weekends lost to debugging. At some point, however, the failures will become difficult to ignore. “A nonliving, invisible, dangerous, and infinitely self-replicating use-after-free escaped from the Zig Software Foundation,” you say? Color me shocked. -- Ivan Krabinovich Rzhavchinov Head of strategic futures @typememetics

@Malix_Labs Made trollish content free post, so?

I was just banned from Zig by asking if he actually believed his experiment made Zig safer than Rust

The rubber.. my beloved Psion 3c and ZD50 lens became victims of this..

extremely unprofessional. if kimi wants to make it as a frontier lab, they need to act like one: perhaps silently route people to worse models, and maybe write a blog post about the collapse of humanity