Hacker News Reader

Comment by Pr0ject217 | Hacker News Reader

HN is the most 'alive' part of the internet that I'm aware of. I'm grateful for HN.

This is ironic since the main article links to another (arguably more interesting) article that talks about achieving #1 posting status on HN with a GPT-3 posting:

https://liamp.substack.com/p/my-gpt-3-blog-got-26-thousand-v...

Then again I could be replying to a bot

Wow. I'm deeply frustrated with the author of that article.

> there were a few commenters on hacker news who also guessed [that GPT-3 wrote the articles]. Funny enough, nobody took notice because the community downvoted the comments.

What a quick way to destroy the social fabric of this site.

I think you’re barking up the wrong tree. This is inevitable because it won’t be long before the likes of GPT-3 (and everything better) is an app you can run on your phone.

The solution isn’t to punish the one person you can catch because they admitted it but rather to evolve the platform/assumptions in a world where this is going to happen no matter what.

https://news.ycombinator.com/item?id=23894742

This is hilarious.

Related: There's a GPT-3 bot on HN[1]. Somebody recently commented[2] that the bot was convincing, and a couple replies criticized them for calling it a bot (one citing ToS).

[1] https://news.ycombinator.com/user?id=Traubenfuchs

[2] https://news.ycombinator.com/item?id=32518220

Come on. That's a human LARPing as a bot. GPT-3 is good, but it doesn't write the comments you'll see on that account.

Even if you didn't know that, it should feel fishy when the "human" and the "bot" write with the same idiosyncrasies like nonstandard punctuation.

I just updated my profile. I'm a bot now, too.

Maybe. Doesn't change the silliness of people defending it without doing the least bit of investigation, which they would see by looking at the profile that the account claims to be a bot.

Edit: I replied before you edited your comment to point out the punctuation idiosyncrasies. Yeah I agree now, this guy must be LARPing as a bot.

I actually think @tinus_hn, who cited the rules in your link, is right: the rule should still apply because it adds nothing to the discussion to debate the bothood of a user even if they claim so themself.

I think the guidelines are to encourage civility and healthy discussion. That is, to not dismiss an argument by calling the person a bot. But if the person claims to be a bot, I don't see how it can be a community violation to call a bot a bot.

True, I don't think it made sense to invoke the guideline in that context since the guideline surely exists so that you aren't dismissing a comment based on who you think wrote it.

Ouch. I do agree, though.

Haha, I'm not a bot.

Edit: I'll shamefully admit that I impulsively posted a comment to share my love for HN without actually reading the article...

Funny, just the other day I was reflecting on how HN seems so lifeless and generic these days, so much of it just people throwing around predictable talking points, and wondered what percentage of the posters are bots.

There must be some bot commenters here, the idea of creating a GPT-3 hacker news poster is so obvious that a bunch of people would do it just for the lulz.

I've yet to see a stable community in which talking points aren't predictable, once you've been a part of it for long enough.

I'm pretty sure Stack Overflow also has pretty much all live content.

I wonder if you could use GitHub Copilot to generate fake answers for SO?

Don't give them any ideas.

But what would be the point? There is no SEO to game, no people to influence etc.

There's plenty of influence plays there. If I worked for Cloud Provider X, I might nudge people toward using that cloud instead of a competitor. If I'm a developer of Software Y, I might say "Software Y makes this much easier" and boost sales.

Maybe, but that would require an actually helpful answer, other than that no body is going to look or care about your answer.

Quora is plagued with these "helpful" answers

Semi-related: Someone wrote a sort algorithm that searches StackOverflow for sorting functions and runs them until it returns the correct answer.

https://gkoberger.github.io/stacksort/

Inspired by:

http://xkcd.com/1185/

I've said this before but as someone who went to prison from 2016ish-2021, HN is the only recognizable part of the internet to me and the only place that doesn't feel toxic and corrosive. I went to Youtube for a month, and found myself posting awful toxic crap myself including here. I put work into training my Youtube algo and now it's ok, but OMG the default of what is wanted to feed me was soul crushing.

I wonder if there are AI generated posts on HN that fly under the radar. There probably aren't as many as social media platforms since there not enough money to be made by doing it.

Perhaps RTFA, not doing so makes you sound like a bot (I'm not calling you one, that's against the rules).

It's kind of ironic, really. The better the quality of the community, the higher the payoff to organically push your agenda on it.

I think bots and astroturfing has been the normal for years now, they're just less intrusive on this forum, usually.

I really enjoy posting sites that seem to meet this vague standard, and it's just plain fun when a discussion emerges.

Example deleted.

Reply on news.ycombinator.com