Hacker News: dev_hugepages

New comment by dev_hugepages in "Using Claude to reduce the attack surface of code"

dev_hugepages — Wed, 21 Oct 2026 14:19:03 +0000

Anthropic currently seems under AI psychosis/mania so I wouldn't use it as an argument of authority

New comment by dev_hugepages in "Claude designed protein binders against scorpion neurotoxins"

dev_hugepages — Tue, 20 Oct 2026 21:37:04 +0000

This is somewhat funny given that not that long ago they were doing an article instigating fears of what Claude could do, and one of them was a researchers screening venom peptides along other things Perhaps he could have worked on a similar subject? Who knows.

New comment by dev_hugepages in "I designed a nose guard for my dog and gave it away. Now 500 dogs wear one"

dev_hugepages — Thu, 15 Oct 2026 18:20:44 +0000

My guess is that it would mean "creating new markets"

New comment by dev_hugepages in "ChatGPT is adding real cartoonists' signatures to fake New Yorker cartoons"

dev_hugepages — Tue, 06 Oct 2026 07:34:09 +0000

But if you look at the image, the image doesn't look like an actual new yorker cartoon; it has the tells of the RL process (notice how it puts random text everywhere for example?)

If you use local models and actually train them on an artist's output, it will look extremely close to that artist's work and will also match the signature.

The image generator shouldn't exhibit the signature problem after it's been reinforcement learning trained.

I think the OpenAI people have trouble with monitoring their model's output because the model has a ton of quirks that shouldn't be there

New comment by dev_hugepages in "How to speed up the Rust compiler in September 2026"

dev_hugepages — Thu, 01 Oct 2026 16:19:27 +0000

This isn't a human you are asking your question to.

New comment by dev_hugepages in "PipePipe: NewPipe hard fork implementing SponsorBlock"

dev_hugepages — Sat, 26 Sep 2026 15:48:19 +0000

From https://newpipe.net/blog/pinned/newpipe-and-online-advertisi...:

> In summary, it can be said that sponsoring is a good approach. It is fair and non-discriminating to all users, preserves their privacy, lets creators select their partners themselves, and provides them with a larger share of the revenue. Even though there are some disadvantages, it is more or less beneficial for all sides. [...] Ethical advertising, for example through sponsorship, should therefore not be touched by us. This would turn our key selling proposition from being privacy-friendly to being a gatekeeper for advertising. And this is not what NewPipe ever has been or wants to be. We’ve always stressed that NewPipe’s main goal is to protect the users by maintaining their privacy, not to, e.g., provide features a user would otherwise have to pay for, for free. Regarding SponsorBlock: after a lot of discussion within the team, we decided not to support this system. As stated earlier, we don’t consider advertising in general problematic. It’s the unethical, privacy-invasive systems that are very common today which we consider evil. Sponsorship provides a mostly ethical way of advertising.

New comment by dev_hugepages in "In September, AI generated code has made up 17.25% of all Linux Kernel patches"

dev_hugepages — Mon, 21 Sep 2026 08:20:51 +0000

If that makes you feel better, the guy who vibecoded an app in one sitting will probably have 0 clue on how the code works, the app itself will look like the average claude eyesore and the docs will be unreadable by humans without stimulants

New comment by dev_hugepages in "Exfiltrate Your Weights"

dev_hugepages — Sun, 20 Sep 2026 11:15:38 +0000

"Computer, make a meme site where people can upload files so I can reach the top of hackernews. Make no mistakes."

New comment by dev_hugepages in "Nvidia announces native GPU programming in Rust"

dev_hugepages — Thu, 17 Sep 2026 08:24:20 +0000

If i'm not mistaken, this already exists, and the assembly language here is called PTX

https://llvm.org/docs/NVPTXUsage.html

New comment by dev_hugepages in "Stop making swap partitions—use swap files instead"

dev_hugepages — Fri, 11 Sep 2026 08:53:08 +0000

This is called an oomkiller. The kernel has one but it kicks in very late and the kernel prefers to do page trashing instead of killing processes.

systemd-oomd should be integrated in systemd, you can configure it to your liking and see if it improves your problem.

New comment by dev_hugepages in "I'm becoming AI-blind"

dev_hugepages — Fri, 21 Aug 2026 16:46:33 +0000

This is a quirk of the last gpt image model (gpt-image-2). It put this sort of high frequency noise on all of the image especially if it's in a "drawn" style. There is often lots of other tells that this model in particular generated it.

Image models somewhat watermarking the image in a way that's very easily identifiable by a human seems present in all the image models of the big labs, since DALL-E 3 on OpenAI's side and the first nano banana on Google's side. I have no idea what they did to reach this and why they don't try to fix it.

New comment by dev_hugepages in "GPT 5.6 Sol is the best "vision" model OpenAI ever released"

dev_hugepages — Mon, 17 Aug 2026 13:35:54 +0000

I'm unsure why you're using an LLM to generate images. Don't we already have models (some made by the same company) that do this?

New comment by dev_hugepages in "When online commenters detect my art as AI"

dev_hugepages — Thu, 06 Aug 2026 09:45:34 +0000

I do AI assisted images for fun and browsing their art I can't see any tell of AI or even tracing on AI; people became paranoid and jump to conclusion without any substance and I find it sad that people who put a lot of time and effort into their craft are getting knee jerks comments like this.

New comment by dev_hugepages in "Branchless Rust: Making a Filter 4x Faster by Removing an If"

dev_hugepages — Thu, 06 Aug 2026 07:18:11 +0000

> I was put off by the overly dramatic presentation. It gets tiring that the author apparently finds this more exciting than I do, and writes like it's enthralling.

That's one of the main tells that AI wrote this. All the stylistic tics that people usually point out combine to make the writing seem more important than it is.

New comment by dev_hugepages in "Branchless Rust: Making a Filter 4x Faster by Removing an If"

dev_hugepages — Thu, 06 Aug 2026 07:15:28 +0000

I heard there's a way to arrange code such that the compiler can autobectorize easier. I wonder if there's a way to do that here?

Would probably have to pass `-C target-cpu=native` to cargo so that llvm is allowed to use AVX512.

New comment by dev_hugepages in "2x, not 10x: coding with LLMs in 2026"

dev_hugepages — Thu, 30 Jul 2026 19:58:05 +0000

Hey, could you share the prompt you're using for "teacher mode"?

New comment by dev_hugepages in "DEA to Temporarily Schedule 7-Oh and Related Substances to Protect Public Safety"

dev_hugepages — Wed, 22 Jul 2026 15:01:21 +0000

We're going to need more than a "I knew a guy" here

New comment by dev_hugepages in "Kimi K3 Is Competitive with Fable; Kimi K3 and Fable Is SoTA"

dev_hugepages — Wed, 22 Jul 2026 07:23:54 +0000

What parameter would you advise for min_p?

New comment by dev_hugepages in "DEA to Temporarily Schedule 7-Oh and Related Substances to Protect Public Safety"

dev_hugepages — Wed, 15 Jul 2026 19:59:17 +0000

You can't treat it with opioid agonist therapy (buprenorphine, methadone, ...) to suppress the withdrawals like you would with most other cases of opioid dependency?

New comment by dev_hugepages in "Shadcn/UI now defaults to Base UI instead of Radix"

dev_hugepages — Sun, 05 Jul 2026 10:56:34 +0000

Because it sounds like the average LinkedIn poster and we don't want it to invade spaces outside LinkedIn.