Nika

How do you distinguish AI content from real, human-made content?

AI is incredibly good, I’d even say almost perfect.

And for many people, that uniformity of perfect templates is starting to feel annoying.

  • For example, a few days ago, someone publicly showed that they built Anti-Grammarly – a tool that intentionally adds mistakes to text instead of removing them (to make it feel more human). But the tool itself is AI, so it’s a bit contradictory.

1) When we’re so flooded with AI-generated content, do you have any methods to recognise it?

For example, I keep noticing the same patterns:

– long dashes,
– phrases like “It’s not X, it’s Y,” and similar structures.

2) But what about beyond text, like images or video?

489 views

Add a comment

Replies

Best
Saad El Gueddari

honestly with images and video i don't think we're at the point where u can reliably tell anymore. for me it comes down to effort more than anything.

low effort ai stuff, midjourney prompt + canva slapped together, ur gonna catch it every time. but ai ugc made by a decent marketing firm running a real video pipeline, u basically never clock it as ai. and at that point whether it's "ai" or not stops mattering, the content lands, the point gets across.

so detection's gonna stop being about ai vs not ai and start being about how much effort someone put in. that gap only shrinks as the tools get better. give it 18 months, even mid effort pipelines will be undetectable.

Bengeekly
@saad_el_gueddari Right now for Video I believe it’s still obvious for long format videos as character consistency… breaks fast.
Amrani Yasser

@saad_el_gueddari Yeah exactly, it’s less about AI vs not AI, and more about the level of effort behind it !

Nika

@saad_el_gueddari I can say that short-form videos are not detectable now. But agree with @bengeekly – long-form videos are now better recognisable.

Mikita Aliaksandrovich

I’d say repetition is usually the biggest giveaway.

When too many posts sound a little too polished, too structured, and too “perfect,” people start to notice it, even if they can’t explain why.

Funny enough, the more AI content we see, the more small human imperfections start to feel real and trustworthy.

Nika

@mikita_aliaksandrovich yeah, but when everyone is using anti-grammarly to "sound human", we will be confused about what AI is, what is not AI, but most likely, everything will be AI. :D

Stoyan Minchev

I don't. search for methods. As long as the content can genuinely make me better, smarter, does not really matter how it is created.

Sometimes people use AI just to fine-tune the text, search for error, suggest better style. This does not mean that the thoughts written are of the AI itself.

Some people have nice ideas and thoughts, but can't express themselves in English, should we really not learn from them?

Nika

@stoyan_minchev This is good approach, but not many people think like this tho.

Stoyan Minchev

@busmark_w_nika I don't know why it is like that. In other places, the moment they smell AI content and a big hate comes out of everywhere and nowhere and it is really painful.

I had chats with other people in other forums. The whole situation creates a fear of sharing and transparency, instead of free will and open discussions.

Amrani Yasser

@busmark_w_nika With images and video, it’s getting harder and harder to tell now. Tools are improving so fast that even "average stuff" can look pretty good !!

Nika

@amraniyasser Imagine where we will be in 2 years. We may no longer even be able to verify things that are in reality. 😅

Amrani Yasser

@busmark_w_nika in 2 years 😂 :

Nika

@amraniyasser This is painful, it hurts lol :D

Ajin Sunny

For videos, its really the physical motion and lip syncs. Don't get me wrong they are so good that its hard to tell apart. However, if you have sharp eyes you can tell something is off based on the physics of the video render.

Nika

@ajinsunny I noticed that their walking is not correct, it is like a glitch in some parts.

John Baek

@ajinsunny  @busmark_w_nika Give it a year or less when this too is corrected.

When I see 50k likes on an AI video, I assume casual social media users simply roll with the new norm which is that AI is here to stay.

Je Yue Yip

Long dashes are one of the easiest at ways to spot AI gen content!!

If someone paraphrases and edits AI content, would that count as AI content though?

Nika

@je_yue_yip1 Yep, but some people started being careful with that, and just with one prompt will swap long dashes for normal ones. I cannot rely on that anymore.

The second question – difficult to say, it depends on how the paraphrase looks.

Je Yue Yip

@busmark_w_nika That’s true. The lines are getting blurred, and original content is becoming harder and harder to find.

Tim Joerg

@je_yue_yip1 I really like writing with dashes. It fits the rhythm of how I speak and think quiet well. But people see just normal dashes now and start to assume you're an AI...

Stan Kolotinskiy

I like to re-read my posts before sending them in order to make sure that there are no typos/obvious mistakes :) Sometimes, when I'm writing a bigger post, I'm just running it through a LLM to check if it has any advice to make. Answering the question: it's indeed a gut feeling, however I'm pretty confident that sometimes I'm mistaking human posts for LLM ones :)

Nika

@sk_uxpin Not gonna lie, but lately I suspect everyone is using AI for generating outputs :D I am just paranoid :D

Stan Kolotinskiy

@busmark_w_nika hahaha, good one :D Well, English is not my native language and while I'm trying to learn something new by reading books and watching movies in English, the more I do that, the more I realize that English is not even half as simple as I thought it was :D

Nika

@sk_uxpin It will be okay (even with our EN) :D

Stan Kolotinskiy

@busmark_w_nika that's reassuring, thanks :)

Maliik

Comments are where I notice it most. Someone replies to you and their first move is to paraphrase your point back with "that's so underrated" attached. Or they hit you with an instant-alignment opener before they've actually said anything. People just say what they think.

For longer content, I check for specificity. AI sounds smart about nothing in particular.

Nika

@maliikb my competitive advantage is that AI rarely write such things like :) :D :P ... so I am safe, a bit outdated, but real :D

Maliik

@busmark_w_nika haha that's actually a solid test. the emoticons are like a digital fingerprint at this point XD

Lara Bishop

i do not rely on a single signal anymore. AI content is good enough that any one pattern can be misleading, so I usually look for a combination of small cues.

Nika

@lara_bishop What for example? Like people who post it etc?

Joshua Herrera

On a similar note, I now have an irrational insecurity when commenting or posting online with correct grammar and punctuation. I assume people will think it was written with AI when I'm really just doing my best to write coherently. I try not to use em dashes even when they're appropriate.

Nika

@joshua_herrera yeah, I started also thinking that I shouldn't use those long dashes too, or any dashes at all :D

1234
Next
Last