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?
577 views
Replies
I do a lot of work in NLP (Natural language processing) and finding Ai text is actually pretty easy once you look at the structure of the output string. Ai text isn't "perfect" it just follows a certain order. It is the same in Images or Video. You will probably need some sort of algorithm to check the text, image pixels and Video frames. But it's entirely possible to check if something is made by Ai without using Ai.
minimalist phone: creating folders
@yash_shukla2 Do you know any tool that can detect AI the most precisely?
text is probably the easiest - there's typically a formulaic way to AI writing. for everything else i find it easiest to throw it back at AI to understand if it's AI-produced
minimalist phone: creating folders
@kohnigel IMO, when someone is good at writing texts, he/she is more likely to nail the AI version the way, that you can hardly recognise it was AI.
The funny part is that adding mistakes is becoming another AI pattern too. For me, the biggest giveaway is not grammar, but lack of lived specificity. Human content usually has weird details, uneven flow, and context that feels earned.
minimalist phone: creating folders
@alpertayfurr It is difficult to say what is true and what not, actually one guy shared an image of bot comments and I didn't recognize them.
We're building Marketing Stack and found that when AI generates content from an actual brand system ( your real tone, selling points, target audience etc..) the output stops feeling generic. The tells people notice (same sentence structures, vague hooks) mostly come from prompting with nothing. Feed it something specific and it reads differently. Still AI, but good.
minimalist phone: creating folders
@wissem_ksantini33 yeah, but I am not so sure whether I would like to feed AI with my data to such an extent to sound like me. I would find it dangerous :D
I think its less about more AI content but more about loads of people just outsourcing their thinking. The content that I post is 80% myself and 20% claude, making some adjustments, but it never loses the crux of it.
Anyone who regularly interact with LLMs can smell the AI Slop from a mile away but then you have people who believe in quantity over quality so they'd rather have AI write everything than to put in the effort of thinking.
minimalist phone: creating folders
@varun_dhamija1 yeah, and some people at least could try to change some parts in the texts.
@busmark_w_nika the bare minimum, honestly but then changing a few words doesn't fix the underlying problem which is that the thinking isn't there in the first place.
You can tell immediately when someone's edited an AI draft versus when someone actually had a thought and used AI to sharpen it. completely different texture.
One thing I notice is when content feels too polished but lacks personal examples or real-world detail.
Human content usually has small imperfections and specific experiences.
minimalist phone: creating folders
@stareena_patrick I think that when someone adds to the prompt "add example", it can create something, but will still feel like a template .D
@busmark_w_nika True, it can add examples, but they still feel kind of templated.
The difference is usually in the small, imperfect details real experiences aren’t that polished or structured.
Build Check
Every day it's becoming harder and harder! I'm a former graphic designer and currently it's too hard to differentiate real pictures from AI made. Challenging time!
minimalist phone: creating folders
@german_merlo1 "Former" graphic designer because of AI?
honestly the em-dash + "it's not X, it's Y" stuff is going to be a temporary tell — once people start prompting models to drop those tics, the surface signals go away. what's stuck for me is structural: AI-written stuff often has perfectly balanced paragraphs and zero loose ends. real writing usually has a weird tangent the writer didn't bother to cut, or a sentence that starts strong and trails off. for video it's even messier — watermarks help short term but the longer play is provenance metadata at capture time (C2PA etc). curious what others are seeing.
minimalist phone: creating folders
@tijogaucher Do you know what? I am really curious when people will start relying on AI at such extent that AI will train on AI and not on people actually :D Because there will be nothing left from people to train on :D
Even when using tools such as "https://gptzero.me", it shows that the text is 100% written by a human, although it was fully AI-generated. AI is getting better every single day.
minimalist phone: creating folders
@sorinify IMO, those "analysers" don't help a lot :D
Asa.team
Honestly I've stopped trying to detect it and started caring more about whether the point being made is real. AI content that adds something specific, an actual data point, a concrete tradeoff, a real opinion, reads totally different from AI content that just sounds polished. The uniformity problem is that most people are prompting for safe, agreeable output.
minimalist phone: creating folders
@ng_junsheng But how can AI add its own perspective? Isn't. Is it about giving AI your own idea?