Hacker Newsnew | past | comments | ask | show | jobs | submit | Ygg2's commentslogin

> Upvoted. I think that said humans should get used to the idea that their intellect is special only insofar as it was first

And used to build the LLMs of today. We'll see how that evolves once LLMs have to feed LLMs with their own "intellect".


What did humans build off of but themselves?

LLMs are not humans. It's good to remember that. They're also not God.

It's the difference between watermark or marked bills.

One is proof of authenticity, other is tracking tool.


> To name it "hallucination" is an euphemism

I agree. It's biased language. When talking about AI remember:

- hallucinated -> made it the fuck up

- thinking -> pseudo-randomly guessed

- escaped containment -> (we) need money

- we need regulation -> our competitors are catching up! Help us Prez!


That's not how it was usually run, but say in Ancient Athen's the council that decided was whoever was at the council at that time.

> People least suitable to wield power are most magnetically attracted to it.

In the current system.*

There are systems that allow you to distribute power more equally. It requires a different setup, like for example randomised choice. The people in power don't like it. It's much easier to bribe 2-3 career politicians and guarantee their success.


I support installing kill switch in Dario and Sam. If they make another fear mongering post about AI, they die.

> There were forty-four against this project in just the last audit.

Was there an audit against coreutils? If not, it's not really apple-to-apple comparison.


It doesn’t even matter. The sheer disparity in vulnerabilities over twenty years versus one year is impossible to hand-wave away.

We are talking about fourfold more CVEs over a twentyfold reduction in time.


It really does matter. I don't know enough about this specific case, but multiple order of magnitude differences in CVE numbers are frequently explained by different policies towards finding and assigning CVEs in many many cases.

Absent more information the default should be to hand wave it away as probably such a difference. CVE counts are not a even slightly reliable metric.


Most of them are TOCTOU races or improperly following symbolic links. For example, uutils mkfifo(1) would create a world-readable and writable FIFO before using chmod(2) to restrict its permissions. Another user could replace that file with a symbolic link between the mkfifo(3) call and the chmod(2) to change the permissions of arbitrary files [1].

Other ones I find concerning are that you could also bypass '-- no-preserve-root' with a symbolic link to root [2]. Or by using paths equivalent to "/", e.g., "/../" [3]. Historically, GNU coreutils has been pretty good with symbolic links and avoiding TOCTOU races. The only notable one I can remember is a chmod(1) bug [4].

I agree with your general point that the number of CVEs is a useless metric, though.

[1] https://nvd.nist.gov/vuln/detail/cve-2026-35352 [2] https://nvd.nist.gov/vuln/detail/cve-2026-35349 [3] https://nvd.nist.gov/vuln/detail/cve-2026-35338 [4] https://github.com/coreutils/coreutils/commit/425b8a2f534fe0...


When triaging an issue you have to prioritise. Do you fix a problem that affects 2-3 people or one that may affect thousands?

The point is that such bugs shouldn't exist in the first place.

Using recursion on unbounded inputs on a programming language that doesn't support that (which are most) is an extremely classical mistake that really should be known to all programmers, especially those of low level languages that care about safety.

Every time you call something recursively you should be thinking "how deep is this?".


> The point is that such bugs shouldn't exist in the first place.

That's true of every bug, but you don't have infinite time or manpower, so how do you prioritise?


By that logic, why even spend effort migrating from a known-working implementation to one which is known to have outstanding bugs that there isn't enough bandwidth to fix?

Hard to say, but there are legitimate reasons. One of which being they think Rust is growing while C is a shrinking language. Other could be safety (yes, yes the audit did surface bugs). Or even that having a single lib work seamlessly on Windows/Linux/MacOS/Redux/Web.

Fashion.

I think it's something else. Ads are bringing less revenue so they must put more squeeze to get same level (+ growth) of juice.

Like Python 2 vs 3? That's what Editions were created to fix. With backwards incompatibility you will get ecosystem breakage and looming threat of future compilers not compiling your code.

The problem is that assuring no breaks ever with inference means you can't ever improve the inference algorithm nor update the stdlib. This is what triggered the time 0.35 breakage. I have a still incomplete/unmerged rustc lint to avoid the situation that caused that (a useless .into() that didn't get flagged because the clippy lint has false positives so it is not on by default) which should minimize the likelihood of that happening again (once I get off my ass and finish it, it just requires some side-quests to add more accurate tracking of cfg'd out items). This might be able to be mitigated by editions, but in practice crater helps to not need that (yet?).

This is not the only kind of breakage a project can experience. Trying to bring up an older project on a new platform will be a compile error (like building a project from 2018 on an Mx Mac), and updating the appropriate dependency to an appropriate version might become a chore. I have no idea how to improve the situation there to make that less painful, beyond having a simple database of crate-version+platform+rustc-version so that the toolchain could provide better/actionable messaging beyond "shit's broken".


Not sure how that relates to the argument at hand. The argument that you should break compatibility for justifiable reasons.

We know how that looks, and it's not good. See Scala and Python.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: