Author Topic: AI and profile detection - risk and what can we do?  (Read 705 times)

Offline theonlywayisup

With the advent of AI and tracking etc is there any risk in our profiles here being uncovered? Ie detecting word type or styles , backsolving IP addresses etc. I'm not overly familiar with all this but sure others are more so. Have poeple thought about this /is there things we should do(ie not allow the site to be indexed by Google /search engines)? Cheers.

Offline RandomGuy99

Many areas of the site are already not accessible to non-members and thus search engines cannot index them and they also can't be used to train AI platforms either. .
« Last Edit: July 20, 2025, 11:43:39 am by RandomGuy99 »

Offline scutty brown

  • Age Check : 18+
  • Forum Helper
  • ****
  • Posts: 25,620
  • Likes: 529
  •  
  • Reviews: 125
Many areas of the site are already not accessible to non-members and thus search engines cannot index them and they also can't be used to train AI platforms either. .

I don't think that's now true

Offline scutty brown

  • Age Check : 18+
  • Forum Helper
  • ****
  • Posts: 25,620
  • Likes: 529
  •  
  • Reviews: 125
Detecting "word type or styles" is only a problem if you also post elsewhere in enough volume to be comparable.
For instance if you posted here, and then maybe on a shipping or aviation spotters forum. But the number of posts on both sites would need to be enough for the AI to do its work.
If this is the only location you post, then what have they got to compare you with? One word of caution: it's not impossible that your posting style could be compared with memos and e-mails you posted at work, that's when things could get dodgy. Much as I hate the idea of using AI to create text, there could be a case for running all office memos and mails through an AI filter before publication in a deliberate attempt at obfuscating your own individual style

Offline lostandfound

With the advent of AI and tracking etc is there any risk in our profiles here being uncovered? Ie detecting word type or styles , backsolving IP addresses etc. I'm not overly familiar with all this but sure others are more so. Have poeple thought about this /is there things we should do(ie not allow the site to be indexed by Google /search engines)? Cheers.

Probably there is a risk, and there's not much you can do about it.

AI can do things that are beyond the human mind, comparable to the way engines are more powerful than humans. It's kind of the whole point of AI, or certainly one of its main advantages; at least until a general AI is developed to which there is no clear path at the moment / it is several breakthroughs away.

But the fact that it's possible doesn't mean that it will happen. Why would anyone care enough to track you or me, especially if there is a cost?

Also who is entirely innocent? Anyone using such techniques should be aware they could just as easily be used against them.

An interesting area of speculation eg future PIs employed to prove adultery? And the ethicists who try to invent rules to protect privacy.
« Last Edit: July 20, 2025, 12:43:00 pm by lostandfound »

Online PepeMAGA

What is the risk exactly though? We aren't doing anything illegal, so more if friends or family used it to "hunt down" posts on the internet?
I would guess if it gets to the point where that's common place then appropriate measures would be added to site security

Offline theonlywayisup

What is the risk exactly though? We aren't doing anything illegal, so more if friends or family used it to "hunt down" posts on the internet?
I would guess if it gets to the point where that's common place then appropriate measures would be added to site security

The risk is partners I think for the main extent... I suppose is it like an Ashley Maddison thing - ie once it's out it's too late. Ie you ask the super AI is my partner cheating or has he ever and then in 2 mins you have the full online profile of your partner.

Offline lostandfound

What is the risk exactly though? We aren't doing anything illegal, so more if friends or family used it to "hunt down" posts on the internet?
I would guess if it gets to the point where that's common place then appropriate measures would be added to site security

The data is already being / has been harvested and stored so that future advances in AI may be used to analyse it.

China, the US, Cheltenham site in the UK ... ?

And of course Big Tech - this is, after all, how they train their AI.
« Last Edit: July 20, 2025, 04:37:42 pm by lostandfound »