Ask HN: Is it time for HN to implement a form of captcha?

This wouldn't solve anything.

To see what I mean, take a screenshot of a random captcha that needs solving and ask an LLM to solve it for you. It will do it accurately.

I've been thinking the same. I'm actually building a little site that presents a textarea that you can type your comment into and it will track its changes over time (typing, editing, pasting, etc) and provide a little playback widget so someone can see the composition of the comment. The idea being you can include a link to the playback in your comment that you post here and someone can eyeball it and see if it looks like you really spent some time writing it, vs just pasting in LLM slop. Of course, a sophisticated agent could _simulate_ writing the comment, but I think it could still help in general.

I've always wished there was a "block comments from this user" feature that didn't rely on vibe-coding my own Chrome extension (and thus not work on Safari where I spent at least 50% of my HN time). I imagine it could even work like Sponsorblock does, and we could crowdsource people who's comments are inflammatory.

I've also noticed that very obviously LLM-generated comments are called out, and the community tends to agree, but those that have any plausible deniability are given far too much leniency, and people will over-index on the guidelines to give them the benefit of the doubt.

I don't think a captcha is the solution, as it'll degrade conversation by an OOM though.

[flagged]

[flagged]

HN already enforces ReCAPTCHA for registrations. More CAPTCHAs will not do much if anything to improve.

Posts like this should just implement the poll feature. You'll have your answer in 24h. Then go and quibble with your data.

https://news.ycombinator.com/newpoll

[flagged]

> "there is a tremendous amount of attempts by bots, and specifically AI agents, to inject slop into HN threads"

Do you have some examples of this? I am on HN almost every day, and I read a lot of comments, and I haven't noticed this

NO

"Me furiously trying to decide what a EURO thinks a motorcycle is" for 60s

Captcha is only effective at annoying legitimate users. If there is any incentive to do so, bots have no problem bypassing/solving them.

That is a losing battle.

Even if you manage to make bot usage more expensive, which is all a captcha can do, the content posted by humans in discussions and shared links is increasingly generated by machines.

It's ironic having a community of people object to the same technology they helped build. Enjoy the show, and learn to live with it. It's going to get much worse before it gets any better, if at all.

This won't work, HN is a high enough value target (Not a random site where bots try to spam some guestbook) that people would adapt to that quickly. Headless browsers, browser extensions, outsourcing captcha solving etc. - there's too many ways to do that if you are determined unless you want to also throw captchas at regular users for every action.

Do you have some examples of ai slop posts?

There are quite a few third party apps for Hacker News, such as Hacki (ios/android). [1]

Something like using a third party app that includes forms of spam filtering like checking when the user joined, how many posts they have, amount of 'karma' (or whatever it's called here). You could implement blocking individual users & etc etc. This app does not have that but it could be forked and modified or talk to the dev...

That might be a better solution than trying to implement all types of annoying captchas & other extremely annoying checks on HN's side.

[1] https://github.com/Livinglist/Hacki

The value of an HN post or comment that people actually see is so much higher than the value of a CAPTCHA-solve that there's no point in even talking about this.

I just realised that one day, an AI moderated board will receive such a post from an AI, not a human. And then a captcha only an AI can solve will appear, and the board will be rid of all "human slop"

Seeing "is this a bike?" captchas on a forum like this one would mean that the web is well and truly dead. Bring it on, for all intents and purposes, and 2fa also, while we're at it.

This is evolution in action. An ecosystem is generating with different things populating it. Is there a better method than captcha out there? For instance, hide things in html comments that only bots would see and if they are reacted to then flag that as a bot account and silently hide their comments (so that another account isn't created)? Do this randomly so that it is hard to find but bot code would catch it. Or other things like text with the same background color so only a bot could have seen it. Basically, instead of staying defensive, go on the attack?

What's wrong with simple categories for comments... Informative, Funny, Flamebait, etc.

Is complaining about the rise of AI Slop itself a sub-category of AI Slop?

We've always wanted to build a service which provides authentication through credit score verification. Whether its applied to dating apps, product review sites, HN. I'd sure love to filter by 650+ only. I'm certain it's illegal but it sure would help.

[dead]

I think you should first ask if captchas are at all effective at stopping bots.

(They are not and haven't been for a long, long time)

Captchas are not effective. You can pay 2captcha less than a penny per captcha and humans solve them for you.

Its too bad Team Blind doesn't support a dev api to their auth service. Work emails are a good candidate for a simple "blue check mark" system for the HN crowd, but with a layer preserving anonymity. Ex. Generate a token, add to profile, browser extension performs verification.

Otherwise agreed with the sentiment.