General Discussion

Hassin Bin Sober

(27,084 posts)

92. Like this?

Fri May 23, 2025, 05:43 PM

May 23

Identifying AI-generated content, especially text and images, can be tricky. Here's a guide to spotting AI-generated materials:
1. Textual Content:
AI Detection Tools:
Utilize online tools like QuillBot, Grammarly, and GPTZero to assess the likelihood of AI authorship.
Sentence Structure and Style:
AI-generated text often has repetitive structures, overly formal tones, and a lack of nuanced language. Look for excessive use of jargon or generic vocabulary.
Inconsistencies and Lack of Detail:
AI can struggle with context and may produce nonsensical sentences, or fail to provide concrete examples or supporting arguments.
Quick Responses:
AI can generate responses quickly, while humans often take time to formulate their thoughts.
2. Visual Content (Images and Videos):
Unusual Details:
AI-generated images can exhibit unusual or inconsistent details like asymmetrical facial features, odd finger placement, or objects with strange proportions.
Texture and Pattern Issues:
AI may struggle with complex textures or patterns, leading to repetitions or awkward transitions in textures like hair, skin, or clothing.
Lighting and Shadows:
Look for inconsistent or unrealistic lighting and shadows.
Background Anomalies:
AI images may have backgrounds that appear unrealistic or inconsistent with the scene.
Facial Expressions:
AI-generated faces might have expressions that don't align with the overall mood or emotion of the image.
Reverse Image Search:
Use tools like Google Images to search for the origin of the image and potentially identify AI-generated content through hashtags or source names.
Digital Artifacts:
Look for pixelation, strange color patterns, or blur in areas where it shouldn't logically be.
Unnatural Movements and Expressions:
AI videos may show jerky movements, stiffness, or facial expressions that seem off.
3. Other Indicators:
Lack of Real-World Understanding:
AI-generated content might lack a grasp of the broader context or specific details, and may make claims that don't align with reality.
Monotonous Tone and Flawless Grammar:
While AI-generated text can be grammatically perfect, it often has a robotic or monotonous tone.
Examine for Artifacts:
Zooming in on AI-generated images can reveal inconsistencies, misplaced shapes, or stray pixels.
4. Fact-Checking:
DW News Fact Check:
.
Use sources like DW News to learn how to spot AI-generated videos, particularly during breaking news events.
Don't Overly Rely on "How to Spot AI" Guides:
.
Be cautious of guides that promise easy identification, as these may be AI-generated themselves.
You can watch this video to learn how to tell if an image is fake content from AI:

Edit history

Please sign in to view edit histories.

Recommendations

0 members have recommended this reply (displayed in chronological order):

130 replies

= new reply since forum marked as read

Highlight:

Seeing people use AI to post to DU is weird [View all] Sympthsical May 23 OP

Large language model AI is linguistic herpes. Ocelot II May 23 #1

I Am Not Ai Mr.Bee May 23 #61

It's good to see the research AI can provide womanofthehills Sunday #126

So far, I am blissfully, ignoring the idea of AI. I guess it will intrude on my life eventually. Walleye May 23 #2

I also am trying to ignore it --but it keeps hitting me in the face. .. riversedge May 23 #8

It has begun to MuseRider May 23 #19

I'm afraid it is turning the world into something PatSeg May 23 #65

It's the next step in dumbing down people. Give them poor education first. mucholderthandirt Saturday #114

You paint a very grim picture PatSeg Saturday #117

Google and Amazon Mr.Bee May 23 #67

Google's AI Overview! ProfessorGAC Saturday #118

I love AI - Groc 3 - instead of having to search web - it does it for me womanofthehills May 23 #68

You know Grok is Musk's AI? Are you sure that's the one you want to choose? hedda_foil May 23 #108

I couldn't agree more! SheltieLover May 23 #3

How can you tell it's ai? Curious. cbabe May 23 #4

I use AI, and you get a quick ear for it Sympthsical May 23 #11

Great description. nt intrepidity May 23 #17

Yep JCMach1 May 23 #20

BTW that was AI JCMach1 May 23 #21

Agreed .... anciano May 23 #36

Immediate giveaway to me moonscape May 23 #48

uh-oh, I use en- and em-dashes all the time nilram May 23 #78

Natter? BidenRocks May 23 #56

Which AI do you use? nilram May 23 #79

Gemini 2.5 is at the bleeding edge at the moment, JCMach1 May 23 #96

I teach high school English. The marked lack of "the" is also a give away. Cuthbert Allgood May 23 #95

New AI is better and if you know how, you will never ever know... JCMach1 May 23 #109

We're going to have a weird future Sympthsical May 23 #23

Having Asperger's, markodochartaigh May 23 #30

I sometimes use ChatGPT to ask questions anciano May 23 #50

I'm in trouble. I would never have recognized that. Scrivener7 May 23 #93

That was also a 0 effort attempt. If I had tried and fed it my examples JCMach1 May 23 #99

I've always used bullet points synni May 23 #26

My life is an outline (with cross links) erronis May 23 #41

I use bullet points in work and school writing Sympthsical May 23 #46

You make a good point--those are definitely some telltale signs of AI-generated text. Abolishinist May 23 #27

It sounds like a face with too much botox and lip filler looks. Ocelot II May 23 #32

Careful! The tit-ular head of DHS will be coming for you - in the famous leather suit. erronis May 23 #42

Agreed. It's kind of clunky. Feels artificial. Hard to put a finger on but feels off. paleotn May 23 #43

I am curious Skittles May 23 #89

Like this? Hassin Bin Sober May 23 #92

You can teach a parrot words UpInArms May 23 #102

when the thread title is "I asked AI about X Y and Z, here's what it said:" eShirl May 23 #66

I appreciate posts which cite credible sources, not those which paste AI scrawls. John1956PA May 23 #5

Yes!! That is what drew me to DU, factual statements with cites. It's a dying art I guess. nt SunSeeker May 23 #37

How does a person know if it is AI posting or not?? Peacetrain May 23 #6

I was wondering that very same thing. Littlered May 23 #9

I think people should disclose TommyT139 May 23 #86

"deactivate the search engine's use of AI" Alice Kramden May 23 #106

Not sure what you mean. CrispyQ May 23 #7

Yes, and Google/Chrome AI Labs is excellent at providing sources, which often Mike 03 May 23 #12

I have never used AI. bamagal62 May 23 #62

It's been cropping up uncited Sympthsical May 23 #13

+1 Celerity May 23 #84

Critical thinking is an endangered activity. Basso8vb May 23 #10

It's garbage in, garbage out. Mike 03 May 23 #16

I keep thinking of that song IN THE YEAR 2525 Skittles May 23 #90

AI is weak and weird in spots Easterncedar May 23 #14

Is there any chance you could link to an example? Mike 03 May 23 #15

definitely think it's possible (probable?) but haven't seen it cadoman May 23 #107

Can you link to an example? MineralMan May 23 #18

Call outs are disallowed Sympthsical May 23 #24

I see. Well, OK. MineralMan May 23 #25

Post #20 in this thread is AI according to post #21 Kaleva May 23 #35

I did see that. It's interesting. MineralMan May 23 #38

I wonder if DU will implement rules governing the use of AI Kaleva May 23 #44

I doubt it. Since there is no reliable way to detect AI content, MineralMan May 23 #51

My AI generated response Kaleva May 23 #76

Again, how do you identify AI content? MineralMan May 23 #82

what if it was a stated rule (at DU), like "No AI-generated material," but kind of on the honor system? Alice Kramden May 23 #105

Who decides? MineralMan May 23 #110

Self disclosure, for one Alice Kramden May 23 #111

This message was self-deleted by its author Kaleva May 23 #45

I have Sympthsical May 23 #47

Yes, OK. Whatever you say. MineralMan May 23 #53

My purpose is a general discussion Sympthsical May 23 #60

What do you mean when you say callouts are disallowed? nt LAS14 May 23 #97

I prefer Heinz 57 Swede May 23 #22

That is another kind of AI altogether DFW May 23 #33

how can you tell? NJCher May 23 #28

I can't tell either Kaleva May 23 #34

its sad people need ai. pansypoo53219 May 23 #29

I think it is worse than weird, especially when not labelled explicitly as AI generated. . . . nt Bernardo de La Paz May 23 #31

Everyone should remember that when you use A.I. it's using you ! Nigrum Cattus May 23 #39

I find it horrifying, myself. WhiskeyGrinder May 23 #40

I'm getting to the point of calling for a "Butlerian Jihad." paleotn May 23 #49

With you on that. highplainsdem May 23 #69

lose blubunyip Saturday #116

Aren't you a peach. paleotn Saturday #119

I was wondering how you were going to handle this rudeness Niagara Saturday #120

???? blubunyip Sunday #122

I absolutely have a wicked sense of humor Niagara Sunday #124

Its a JOKE !! blubunyip Sunday #121

No worries. The thing about the interwebs.... paleotn Sunday #123

What was the joke? LuckyCharms Sunday #125

Irony blubunyip Monday #130

I've never understood the use of AI in discussion forums fujiyamasan May 23 #52

Maybe they are bots and should be reported Figarosmom May 23 #54

Well shit.. Escape May 23 #55

It's great for people with learning disabilities that make spelling challenging. David__77 May 23 #57

I use A1 to evaluate my postings on different forums to see how accurate A1 thinks about what I'm thinking ... aggiesal May 23 #58

A1 sauce will jam up your keyboard, which you will have to replace because you can't clean it Bernardo de La Paz May 23 #83

You didn't consider that I used A1 sarcastically? I thought that was obvious. ... aggiesal May 23 #88

Sorry, I missed that it was sarcasm. My reply was a bit "tongue in cheek", yes. All good, thanks. . . . nt Bernardo de La Paz May 23 #91

I moderate a small forum, and we had to make a rule. forgotmylogin May 23 #59

The AI I would worry about is the microchip Emile May 23 #63

I would say that AI is just more over hyped nonsense like self driving cars. flashman13 May 23 #64

Waymo is increasingly popular in San Francisco Sympthsical May 23 #70

LLMs - generative AI tools - are unethical, illegally trained tools that no one with any respect for highplainsdem May 23 #71

I couldn't agree more! SheltieLover May 23 #73

. MorbidButterflyTat Saturday #112

Serious and not disrespectful nor loaded question, but: TheProle May 23 #72

I get you Sympthsical May 23 #75

Understood and appreciated TheProle May 23 #77

That's the key point. Is the AI "pretending" to be a person? LAS14 May 23 #98

could that kind of AI fake converstation be flagged for violation of Du rules? Alice Kramden May 23 #103

By those three criteria, it's not very different. sl8 May 23 #80

All of that is part of the misinformation/disinformation arena, some of which is spewed out here. Celerity May 23 #85

How does one post a link to an AI word salad? FakeNoose May 23 #74

I'm not sure about the links. sl8 May 23 #87

Good suggestions, but not many people are doing that FakeNoose May 23 #104

Makes me miss the old days of spambots SouthBayDem May 23 #81

How do you know it's AI? nt LAS14 May 23 #94

You don't, unless it is a 0 effort attempt JCMach1 May 23 #100

I want to know if I'm dealing with an entity that cares what I say. By "care" I mean... LAS14 May 23 #101

I thought I had one place that was safe from "AI" stupidity. Seems I was wrong. mucholderthandirt Saturday #113

VERY worrisome Delphinus Saturday #115

I'm still waiting for this: IcyPeas Sunday #127

It might be prevalent with trolls but I can't picture a regular poster going through ecstatic Sunday #128

I don't really understand the purpose of this OP EarlG Sunday #129