Regarding AI responses in BHW threads

[playpen]

Suggestion:

Put a checkbox next to post submit that should be checked if any part of your response was generated by AI.

Then allow users to mute AI-generated answers. Either that, or clearly delineate them as containing AI so that you can distinguish if you're reading someone's genuine experience vs AI output.

It is frustrating reading a thread that asks a question and then weeding through a dozen different ChatGPT-generated responses before finding real posts.

The amount of time I'm spending trying to discern whether I'm reading actual advice from experience vs generic ChatGPT responses is concerning. If anyone wanted to read ChatGPT responses they could just go to ChatGPT?

I also think requiring it will cut down on the noise of those who are just making responses to inflate their post count with minimal effort, faced with having a majority of their responses labeled as AI generated.

I think would improve user experience. Just a suggestion.

 

[Rufai]

Plus for this. I think if people need AI answers, they will go straight to ChatGPT, but they come here to hear other peoples opinions. And while some are using it for translation, I think the majority are using it to spam and increase their post counts.

 

[playpen]

My bad, apparently this has been discussed before (sorry can't post a link to the discussion yet as I'm still in noob mode).

Apparently the solution is to report those who are abusing AI. I'll take that advice.

Sorry about the thread in that case, any mod feel free to delete it.

 

[JessiPinkme]

But how will you clearly distinguish something is generated with ai and something isn't ? Coz with a good enough prompt you can avoid all type of filters.

 

[Geasy]

But how will you clearly distinguish something is generated with ai and something isn't ? Coz with a good enough prompt you can avoid all type of filters.

If they put the good enough prompt then they deserve to reply in posts. It is really hard to see what the forum has come to. It is much better to have 2-3 genuine replies that actually help everyone than having 30 replies that nobody reads.

 

[playpen]

It is really hard to see what the forum has come to.

I used to go by phpbuilt, until my account was stolen/compromised in December. I haven't been active since the advent of AI generated responses so I'm only now experiencing what BHW has become vs what it used to be.

One of the things I used to love to do when I was active was to find people who violated the rules and get them reported and banned or whooped or whatever the mods like to do to them.

I feel like I might be able to help locate AI junkies and notify mods. They say to report them, so I'm cool with that.

 

[Venusaur]

I used to go by phpbuilt, until my account was stolen/compromised in December. I haven't been active since the advent of AI generated responses so I'm only now experiencing what BHW has become vs what it used to be.

One of the things I used to love to do when I was active was to find people who violated the rules and get them reported and banned or whooped or whatever the mods like to do to them.

I feel like I might be able to help locate AI junkies and notify mods. They say to report them, so I'm cool with that.

Sorry for it. I reported this phpbuilt account a few months ago, in December, they were selling it on some forums and random telegram groups, glad to know that after a while, mods finally banned this compromised account.

About this thread as a whole, it would be wonderful if some users are tagged as AI generated spammers and their answers are automatically collapsed (like happens on Reddit in some cases), that would be a shadowban instead of a full ban, if anyone wants to read their AI garbage, they can click and open the answer, but they are collapsed by default - after several verified reports, so we report those AI comments, after repeated reports, mods label that member as an AI poster and all their comments are collapsed.

Or just ban them all, I wouldn't mind xD

 

[playpen]

I reported this phpbuilt account a few months ago

Thank you. Really my fault for not being vigilant with keeping my password secure. Probably someone looping through data breach accounts on BHW until they got a hit. Really sucks.

Fortunately, I signed up this other account about the same time (playpen) but didn't use is because I don't like the name playpen lol. I'll have to change it someday.

The mods know this account is the real me, my other account was banned for being compromised, but the only thing I did wrong was not keep my password secure (really dumb of me to use the same password on multiple sites).

As I'm refamiliarizing myself with a site I loved long ago, I'm noticing alot of AI spam. It's not cool -- there has to be a solution to it. If that solution is currently reporting spammy accounts, so be it I'll figure out an elegant way of doing that.

 

[TheSplinter]

It should be banned, but i haven reported a response that was AI to one mod and he said the response was relevant to the topic or whatever and did nothing about it. If you're using AI to respond you're a bot by default. Why would we want bots here? You might as well just shut down the forum and tell people to go ask AI. Or create an extensive wiki that contains only AI articles and updates itself. We're getting to a point where people simply don't want to participate. The ones who do seem to be bots

 

[playpen]

and he said the response was relevant to the topic or whatever and did nothing about it

Perhaps there needs to be a pattern in multiple posts. I don't think I'd report someone for one AI response, but if 99% of their responses are AI generated you could perhaps get them warned.

There was a thread not long ago (sorry I can't link to it yet) on homepage, someone asked "is BHW full of bots?" and then stated they asked a question about chrome selenium, but all they got were AI generated responses. A moderator apologized and apparently cleaned it up.

And then there's the original thread I found talking about AI responses (sorry again, can't link yet) back in 2025 where they said they will address abuse.

There's always been an issue of low quality posts from people inflating their post count for the reward of credibility it brings (allowing you to become Jr. VIP or just avoiding "Newbie" under your name). Those accounts were easier to spot and disregard many years ago.

We'll see. I'm a creative guy I like to come up with creative solutions to problems, hehe.

 

[Diamond Damien]

We are currently trialing an update that identifies AI responses and marks them as such.
Interested to know what else we’d like to see in there as part of this package?
Especially considering what to do once they are identified as AI.

 

[SettingSun77]

If you're using AI to respond you're a bot by default.

This isn't really true. As has been stated numerous times, many members use various AI programs (especially ChatGPT) as a translation tool. Doesn't make them a bot.

 

[playpen]

Especially considering what to do once they are identified as AI.

AI is a fantastic thing. Comments from experience and first-hand knowledge is also a fantastic thing.

When someone wants AI they go to ChatGPT and when they want expert first hand advice they come to BHW.

I think the AI here is detrimental, because it pretends to be something it isn't -- which is an authoritative account of personal expertise.

The problem isn't necessarily AI, but that they are using AI poorly in a way to inflate their post count or misrepresenting AI responses as their own expertise. It seems logical then that the problem would be resolved if warnings are issued to people who are using AI too often, which could be programmed as a percent of posts flagged. Give people the option to mute users who are using AI over 70% of the time, for instance.

Not sure the perfect solution, but much thanks for working on it!

 

[TheSplinter]

AI is a fantastic thing. Comments from experience and first-hand knowledge is also a fantastic thing.

When someone wants AI they go to ChatGPT and when they want expert first hand advice they come to BHW.

I think the AI here is detrimental, because it pretends to be something it isn't -- which is an authoritative account of personal expertise.

The problem isn't necessarily AI, but that they are using AI poorly in a way to inflate their post count or misrepresenting AI responses as their own expertise. It seems logical then that the problem would be resolved if warnings are issued to people who are using AI too often, which could be programmed as a percent of posts flagged. Give people the option to mute users who are using AI over 70% of the time, for instance.

Not sure the perfect solution, but much thanks for working on it!

Why would anyone come here to read AI responses? That's my point. I don't need BHW to talk to AI. The only reason to be here is first-hand experiences.

As far as the "translation" argument, that's valid, but what i usually see here is comments totally constructed by AI. Translations i probably don't even notice, cause if you ask for a translation Chatgpt doesn't create a typical Chatgpt response.

Just saying. It's kind of killing the forum.

 

[bilbo09]

Suggestion:

Put a checkbox next to post submit that should be checked if any part of your response was generated by AI.

Then allow users to mute AI-generated answers. Either that, or clearly delineate them as containing AI so that you can distinguish if you're reading someone's genuine experience vs AI output.

It is frustrating reading a thread that asks a question and then weeding through a dozen different ChatGPT-generated responses before finding real posts.

The amount of time I'm spending trying to discern whether I'm reading actual advice from experience vs generic ChatGPT responses is concerning. If anyone wanted to read ChatGPT responses they could just go to ChatGPT?

I also think requiring it will cut down on the noise of those who are just making responses to inflate their post count with minimal effort, faced with having a majority of their responses labeled as AI generated.

I think would improve user experience. Just a suggestion.

Yes, it really is very unpleasant. Because as a user, I expect to read a response from another person, because that's precisely why we're posting.

However, it is difficult to reliably detect whether a text has been generated by AI or not. Therefore, implementing an AI filter could be counterproductive, as it is difficult to distinguish between them for very short answers.

This is the bad thing about ChatGPT, Claude, and more...

 

[Utakata]

I swear AI answers are one of the worst things that happened on the internet xD

 

[Raimundo]

If you’re talking about AI responses in BlackHatWorld threads, the real issue isn’t whether an answer *came from AI*, it’s whether it adds value or looks like a generic bot dump. Most forums (especially BHW) reward **practical experience, real examples, and clear reasoning** AI can help draft answers, but if it’s obvious, repetitive, or too polished, people sniff it out fast and ignore it. The best use of AI there is as a *drafting assistant*, then you sprinkle in your own wins, data, and personality before posting so it doesn’t read like a robot.

 

[thehermeticdev]

Question: i'm not so good at English, I use to make my thread hand-made in spanish and then improve it on English with AI. Is that so bad? I mean, not allways using AI for replying, but surely when I use AI is NOT to gen answers, but only for improving the structure and the language barrier. Is that so disliked anyway?

 

[SettingSun77]

Question: i'm not so good at English, I use to make my thread hand-made in spanish and then improve it on English with AI. Is that so bad? I mean, not allways using AI for replying, but surely when I use AI is NOT to gen answers, but only for improving the structure and the language barrier. Is that so disliked anyway?

The way you're using it is absolutely fine and we don't want you to stop doing so.

 

[navyraiser]

In general a good idea! The problem I see is that most "hardcore" AI responses are 90% from users who want to warm up their account. Meaning...they would never check the "AI checkmark box"....unfortunately