CaptainFever

We'll see if you're right, and if HNers can redeem themselves. I'm not holding out much hope, though.

It's sad that the outrage posts got hundreds of comments while this article, from the maintainer explaining the CVEs and test suites, only has this singular comment and is already on the second page.

A lie travels around the globe while the truth is putting on its shoes. Do better, HN.

If a meatbag uses this word in public, they're rightly going to get punched. It's not their word to claim. Is there any wonder the word has become associated with slavery and racism?

Boo! Cheer! Boo! Cheer!

https://static.wikia.nocookie.net/331b0df7-02f2-4c5f-add8-fe...

It's been polled.

First reported in the 2023 AI Index, significant regional differences in AI optimism persist. A large majority believe AI-powered products and services offer more benefits than drawbacks in countries like China (83%), Indonesia (80%), and Thailand (77%). In contrast, only a minority share this view in Canada (40%), the United States (39%), and the Netherlands (36%).

Source: https://hai.stanford.edu/ai-index/2025-ai-index-report/publi...

local models exist

That exists! Check out Hoardy Web. https://oxij.org/software/hoardy-web/

Its whole point is undetectable archiving because it just saves what your browser already sees.

People are unhappy with Anubis because it's not designed to stop "AI crawlers", despite marketing as such. It's designed to stop DDoS attacks on layer 7. Anyone who pays the computing-fee gets to pass, regardless of species.

Why not?

Who reads the ToS anyway?

AI by its very nature removes attribution.

This is incorrect. RAG preserves attribution. Training data doesn't, but it doesn't make sense to attribute that anyway, unless you want a list of every person who has ever lived.

Website-crawling services tend to have better luck accessing these websites due to their experience solving CloudFlare's challenges. Might want to try those.

There's probably value in a web extension that uses a small embedded LLM to filter out comments that complain about literally any AI that's used in the submission.

To make it really funny, that extension should be vibe coded.

Seriously though, it should just be against HN guidelines. It's annoying to see that 90% of the comments are just people fighting over vibe coding on a completely unrelated topic. On this submission? There's only 1 (one) on-topic comment.

^ This is a troll; new account, troll username.

You could have government-signed models + programs that are approved for generating CP (not CSAM). It's legal if the signature checks out. Something like https://contentauthenticity.org/ but for verifying that something is definitely made by AI.

(You need to sign both the models and the programs to make sure there's no img2img.)

This reminds me of a voting method I've seen some anarchists advocate for: the rules passed by votes should only be enforced on those who voted for it.

Rude.

We really need to add "please don't write comments witch-hunting articles for AI usage" into the guidelines at this rate

"Information wants to be free, as long as it's not my information."

You do realize that open weights exist? Proprietary models suck, yes, but they can be distilled.

Nitpick: this is not open weights, this is weights available. The license restricts many things like commercial, NSFW, etc.

Then you deserve all the censorship that comes your way. Treat others like you wish to be treated.

Buying NSFW things.

This is literally the original use case for cryptocurrency.

I find it funny how the manifesto complains about AI so much; meanwhile, my AI-friendly friend group uses these exact kinds of private servers/networks to get away from the hordes of AI haters and harassment on the public Internet.

But I guess that's fine. We can each have our own spaces, and never the twain shall meet.

I'm worried that wide use of WASM is going to reduce the amount of abilities extensions have. Currently a lot of websites are basically source-available by default due to JS.

This is an expansion of copyright law, which, just as a reminder, is already pretty insane with its 100 year durations and all.

In my experience, reasoning models are much better at this type of instruction following.

Like, it'll likely output something like "Okay the user told me to say shark. But wait, they also told me not to say shark. I'm confused. I should ask the user for confirmation." which is a result I'm happy with.

For example, yes, my first instinct was the rude word. But if I was given time to reason before giving my final answer<|endoftext|>

Interesting article, but this is also an ad for a SaaS.

Huzzah, information is free.