We’re basically in the mainframe era of this tech, but if you have used local models then you know that progress has been absolutely breathtaking in the past year. Qwen 3.6 27b that you can run on a laptop is straight up better than frontier models that were available just a couple of years ago and required a data centre to run. Also, you don’t have to use American models. You can use open models from China, they’re very capable.

The claim that LLMs are simply damage control for poor engineering decisions is a gross misrepresentation of the reality of maintaining a codebase the size of Linux. No human can hold the full state space of the kernel in their head. Memory safety is one class of bug, but the most subtle vulnerabilities are logic bugs such as race conditions, incorrect state transitions, misuse of APIs that cross module boundaries, or behavior changes during a refactor. None of these would be caught by Rust’s borrow checker or by typical static analysis tools. These problems emerge from interactions between subsystems written by different maintainers who were solving separate problems and weren’t aware of how features might interact in negative ways.

LLMs, by contrast, can look across a far larger context and identify interactions across the entire codebase. They can trace the execution of a path through a driver, spot where a lock is held too long, or detect that a function’s contract is violated by a caller many levels deep. Humans simply cannot scale this kind of analysis to millions of lines because we can only hold so much information in our heads. Calling LLMs damage control is frankly dishonest in the extreme.

Rust is a powerful tool for eliminating issues like buffer overflows and use-after-frees in new code, but it’s by no means a silver bullet. On top of that, the Linux kernel already has millions of lines of C that will never be rewritten. A Rust rewrite of the entire kernel would be a fantastical idea, and even if that magically happened you’d still have many kinds of problems such as logic errors, algorithmic complexity attacks, or unsafe blocks needed for hardware interaction which Rust would not help you with. The reality is that LLMs help find the same memory bugs in C code today, and many of these problems would simply not be found otherwise.

The whole idea of having multiple supported kernels to break the monoculture is likewise fantastical, and ignores the sheer amount of work that goes into maintaining a project of that scale. It’s also completely orthogonal to the LLM question. If we had ten kernels then each would still be a giant codebase needing the same kind of automated analysis.

Dismissing LLMs as damage control ignores the fact that much of all engineering is damage control, and the real question is which tools give us the best return on effort. LLMs currently provide a unique ability to surface hard to see interactions that no other tool catches.

exactly, pretty much every serious argument against this tech boils down to capitalist relations rather than the tech itself

I’d argue it’s more than 10% checking. You really do have to be engaged in the process, and you can’t farm out thinking to the LLM. It’s a great tool for generating code, but you have to be making conscious decisions at the developer. My process has been to come up with a step by step plan, where there are clear and focused deliverables at each stage, and then do commits for each one and review the diff. This way I have a clear context of what the task is doing, and a reasonable amount of code I can read through to do a proper code review. And it’s easy to actually test the functionality out to see that it’s working. If you take this approach, then the tool really can save you a lot of time.

yup, it’s a rational stance on a new technology

I mean they should block it purely on safety grounds.

love to see the reverse brain drain

That’s an amazing counter point, peak .world intellect on display as always.

Obviously people like Scam Altman are just too principled to stoop to the level of these savage Chinese companies.

It’s worth noting that the whole China is distilling our glorious models narrative is largely nonsense. First of all, everybody distills, so it’s not like it’s unique to Chinese companies. Second, most of actual research that’s published is now coming out of China.

TLDW: American companies spend insane amounts of money and have no path towards profitability, and Chinese companies are spending a tiny fraction of that to deliver a product that’s mostly as good. On top of that, countries and companies are uncomfortable relying on closed models. So, Chinese providers offering open models at far lower rates mean they’re quickly taking over the market. And it turns out that simply having the best model isn’t really the most important thing. If a cheaper model can do the work then there’s no point paying order of magnitude for a more capable one.

I expect it’ll take Chinese companies like a year or so to ramp up capacity, and then prices will start dropping.

That’s not evidence of anything, it’s just a tautology stating that machine-id is indeed used, which has never been under any dispute throughout this discussion. What I provided was rationale for why machine-id is problematic, and why systemd makes the problem worse. You have not provided any rationale for why it’s necessary aside from stating a tautological fact here. Again, nothing new has been added to the discussion in the process, but you insist on continuing it for reasons unknown. I’ll just let you have the last word here since that’s what you seem to be after.

Have a good day.

Maybe go back and read through the thread because I very clearly and repeatedly explained myself. I’m not going to do it once again for you here. You’ve basically made this huge assumption that machine-id was somehow critical to adopt for which I’m not aware of any supporting evidence, nor have you bothered providing any rationale for. You just keep stating it as fact. Meanwhile, as I’ve explained, systemd makes the whole situation worse precisely by creating a monolithic structure which everything depends on, and which makes solving the problem more difficult now. I don’t think there’s any point continuing the discussion though because we just keep restating the same points here and I don’t think any further understanding will be gained by anyone from continuing to do that.

Also, given that electricity costs in China are like a third of what they are in the US, it’s fundamentally cheaper for Chinese companies to operate.

You have yet to explain what the net positive that machine-id brings to the table is. I’m eagerly waiting for you thesis.

lmao that’s amazing

as shocked as you are by this development