How to spot AI posting to this forum?

Just wondering if anyone else has experience with this and if there are any signs to look for?

Good question, but a little vague.
First of all, there's no actual AI - Artificial Intelligence is just a buzzword, what we really have nowadays is electronic mimikry that is so good that humans are gladly fooled by it.
But for the sake of this discussion we can accept that that's what AI stands for.

Secondly, and specific to this situation, one needs to differentiate between bots and AI-like content creation.
Bots are interested in automating the process of posting. Often they use a few very simple, sometimes even nonsensical or out-of-context sentences, just to spread search terms & links.
Very few bots use AI to produce that spam.
Things might be different on "social" media platforms, IDK.

So, be precise.

I recently read an article written by an AI and it sounded a lot like a student who is trying to gloss over the fact that they don't know enough about the topic at hand, filling space with empty, noncommittal phrases.
That in itself is spooky enough, though...
So, if it feels like someone is trying to fill your head with wool, that might be an indicator.

But you will soon make sure when you try to engage in a conversation.
Ask direct questions, and make it clear that you want an answer.

Edit:
the additional powers that moderators hold could go a long way in detecting bots. IP addresses, duration of session, how quickly does one post follow the other etc. But in my 7 years on LQ I have not once seen a mod making clear statements about it.

Originally Posted by JockVSJock Just wondering if anyone else has experience with this and if there are any signs to look for?

Nothing I can add above Ondoho's thoughts to agree that people "mention" AI like it's a thing, but I fully agree that it usually seems full of guff.

Originally Posted by ondoho the additional powers that moderators hold could go a long way in detecting bots. IP addresses, duration of session, how quickly does one post follow the other etc. But in my 7 years on LQ I have not once seen a mod making clear statements about it.

Not as much as you surmise. We can see the IP address which someone signed up from and a list of IP addresses which they appear under, but that information is not always helpful with the way VPNs and internet access work. I've seen situations where people complain about a person with multiple accounts where it is very obvious, posting style, misspellings, agenda, etc. But the multiple usernames might have entirely different IP addresses, not even close subnets.

Suggest you ask using your own thread in the Suggestions and Feedback forum. We certainly will answer questions, you need but ask.

There was an odd post that I responded to the other day, here on Linux Questions. Odd in the way the language was structured.

Then I thought about it and one of the sub reddits that I visit frequently is overrun by bots that post.

Just wondering if there was a way to tell.

People from all over the world post here so it may be a translation issue as well.

^ or a google translate issue.
I often have the feeling it makes communication harder, not easier.
I prefer genuine broken English any day.

I’d be surprised if spambots aren’t already using GPT 3.

Here’s something I look at. If the post looks like it’s relevant only to the forum’s META tags and not to the actual contents (or age) of the thread, or if the OP never responds in any way that actually shows understanding other than spotting a few keywords, it’s a bot.

EDIT: If it's only responding to the subject and not to any of the posts that were already in the thread, that's a big hint too.