Retro_Dev

I like how the plot summary paragraph of the wikipedia article does not contain the letter "e" as well. Somewhat related is Gadsby - another novel without the letter "e" - https://en.wikipedia.org/wiki/Gadsby_(novel)

I'm a bit confused why people are freaking out about these term changes - Meta scrapes my website already on a fairly regular basis (even with robots.txt). Is the point of this change to justify more intense scraping?

It's probably easier to look at the source code than try to decompile it: https://github.com/anthropics/claude-code

How so?

don't give them too many ideas, or my water bill will go up

Others can confirm, but I suspect a lot of inefficiencies in large company programming projects are due to multiple people working on the same code - where you ultimately have to let both above and below average programmers work, and fixing bugs or improving existing implementations is more difficult because of company politics or interactions with others.

Assuming you literally mean "closed-source" (and not just closed weight), then yes I totally agree - it would be really nice if the training data for some LLM were reliably (and ethically sourced), but more important to our assurance of the model that it is byte-for-byte reproducible by us, as to prevent these "sleeper agents" from appearing.

I like using the aria2[0] download utility - I don't think I've downloaded anything bigger than a dozen megabytes through a web browser for the last several years, and this method allows me more protocols, multiple connections, and of course continuing downloads. That said, it's nice someone found a solution to the problem - makes life easier for everyone who doesn't use an external download utility... How do you even go about finding the solutions to bugs like this?

[0]: https://aria2.github.io/

my bad - I missed this. 230 million parameters is very small though.

> 80+ tok/sec on the smallers models.

how does this compare against existing TPUs? Also, how small is "smallers models?"

Excellent idea. Brain surgery would be an ideal use for AI here. Think about how much money we would save if we no longer needed to pay for neurosurgeons, but instead pay the energy cost to run a precise AI-controlled drill, forceps, and suction tube. Better yet, we could massively speed up brain surgery by forgoing tool-assisted LLMs and just have jev-like system-one decision models control the position and function of these tools in real-time. I see no problems here. In the meantime, all those poor students who study so hard to be a good neurosurgeon can instead devote themselves to studying some sort of fungus that takes over mammal brains and makes them want to study fungi... Or wait, is that AI?

https://www.xkcd.com/1664/

Hmm? I always thought it was how much you had to flavor the statement to swallow it.

Nope, but why would it matter? Nobody is stopping you from writing Zig code with an LLM, the policy you are referring to is only relevant to the compiler source code.

Yep. 95% AI and 5% "mixed" analysis. Is it a good article? Well, I don't care to check because it doesn't feel like any effort was put into this. For all I know and guess, this was stolen from real humans.

Well the proper analogy would be that you write your programs in assembly. I think the more impressive thing here isn't the ray tracer, but the C (or was it C++?) to brainf** transpiler. That would be equivalent to you writing both C code and the C compiler. Pretty impressive, but depending on the complexity of the program not as impressive as writing whatever it is in assembly directly. A raytracer would be PAIN to write directly in brainf** - I think I'd want to use fixed precision everywhere rather than emulated floating point. (on second glance, it does look like the author of the article is also using fixed precision arithmetic, but confusing the label for the layout of the number with the type of number representation itself?)

I cannot wait for my cowsay coprocessor.

Yep it's fast alright. I wonder how much faster Mercury would become if hosted on Cerebras.

Isn't the entire deal with jev that it is fast? I'd be interested to know how the energy cost of the Qwen-based model compares with Jev. Of course, Jev is currently locked up so we don't know... "Trust me bro Jev is revolutionary and amazing, pay more money for our inferior product which costs more to run, and of which you need to access by sending us the data"

I'm fine if they put preventative measures in place to protect their work. They already do so. I am NOT fine with their mass manipulation of public opinion to fuel an entirely hypocritical viewpoint. Like, any argument here is hypocritical - but they aren't saying what is REALLY HAPPENING ("distillation steals our work and reduces our profits"), and are actually saying words that make other people fight their battle ("national security", etc).

> Distillation attacks, in which attackers use thousands of fake accounts to extract a model’s capabilities at industrial scale, create safety and national security risks. Distillation allows bad actors to create highly capable models without the safeguards we build into Claude. Our September 2026 threat intelligence report details the illicit distillation activity we’ve detected and disrupted so far.

Such a negative tone they put on this. Distillation is amazing, because it means anthropic and openai fail to keep a monopoly. Who even are they who claim it's unethical? If it is truly unethical, then so is the mass data scraping they do on my personal website on a regular basis (without my consent), and all the unauthorized use of content produced by authors, blog writers, wikipedia contributors, and creators everywhere. If it is truly unethical, then anthropic, openai, meta, google... all these companies should have deleted their LLMs long ago. This wording disgusts me.

Heck, it would be amazing if we had more models without guardrails - some of the models that are produced via heretic[1] are actually quite nice to use - in particular, I've enjoyed investigating Chinese censorship by interacting with an abliterated model of Qwen3.8-27b. If security is really a concern, then secure your systems - don't attempt to dumb-down the tools we use. If someone breaks your window, then they are responsible, not the hammer they use to do so.

[1]: https://github.com/p-e-w/heretic

My bad folks - future links I share will be better displayed (as a foot note).

Spymarks just seem like another word for https://en.wikipedia.org/wiki/Steganography. On that note, one way we can prevent it is to assert that all our content is byte-for-byte identical with the last known trusted stage of what we have produced (for example: a camera we are certain does not watermark, an image editor we are certain doesn't watermark, an image compressor we are certain can't watermark, etc). One vector that I am particularly concerned about is social media. Most images and videos uploaded to most social media is re-compressed by the target platform. This is a door to tracking that is far too easy for social media platforms to open. They might rationalize it (if discovered/announced) by saying that our memes won't be reposted, images or work stolen, etc... but honestly I'd rather my work be stolen than tracking information inserted in there. Oh, we also have stuff which is way more secure, like time-stamped cryptographic signatures.

Does it even matter if it distills other models? ethically no, because the information was originally stolen from us users... it is indirectly harming anthropic if it is happening, but who cares? It is better for everyone except anthropic that anthropic doesn't have a monopoly.

A very cool project, but of course not perfect. I'd rather have 30 megabytes of phrases mapped to the perfect and correct control changes in a home, rather than a heuristic built around 30 megabytes. I tried to "warm the house" (increase the temperature of the thermostat), but the model actually turned the lights to a "warm brightness" - reasoninig being `"'warm the house' -> set_lights to warm brightness. No specific room given, so use default 'living room' as default."`

I especially hope this holds true, because I don't want my information being leaked by anyone.

> there’s no such thing as a secure computer system

Where is your source for this? It is entirely possible to make a secure computer system, though it does require effort. The article specifically mentions "up to" 10% and the fines applying to companies leaking data on purpose or through negligence. I doubt the fines will be nearly as high for a company that tries to secure a system (and thus prevents more leaks) rather than a company that does not try to secure a system (assuming that leaks will occur), if the same breach happens.

Computers are deterministic (excluding cases where practically impossible cosmic ray events occur), so while we have the power to ensure system security, we should ensure system security. Heck, even just encrypting consumer information and protecting just the keys to this data would already decrease the effectiveness of many data breaches.

Um, I think it does the opposite of what you are suggesting - this aims to reduce data breaches and incentivize people to prevent these breaches.

A restart of the process does not necessarily mean reverting the model state. I don't know why you would even do that, because you'd lose all the progress you made.

finding simple bipartite expander graphs for better LDPC (expander) codes

this! - the reason Zig defaults to one semantic analysis thread is due to current (to be improved) limitations; some ideas for Zig also will require one compilation unit - the restricted function pointers proposal being a good example: https://github.com/ziglang/zig/issues/23367

The bun team tried to parallelize Zig's semantic analysis a while back - and it DID compile about 4x as fast... YET the builds were no longer deterministic, which is a hard requirement for lots of things. More effort will need to be spent here than naively adding a thread pool. Hopefully Zig's Sema.zig gets faster, but Rust is also super slow at compiling and needs to improve. Oh - one last thing to note, Zig has very fast debug compilation and incremental compilation, it's a real nice thing to work with that you simply can't approach in most other languages right now.