đź“‹ Table of Contents
Jump to any section (21 sections available)
đź“ą Watch the Complete Video Tutorial
📺 Title: I Can Spot AI Writing Instantly — Here’s How You Can Too
⏱️ Duration: 959
👤 Channel: Evan Edinger
🎯 Topic: Can Spot Writing
đź’ˇ This comprehensive article is based on the tutorial above. Watch the video for visual demonstrations and detailed explanations.
In 2025, artificial intelligence has become so deeply embedded in digital communication that spotting AI writing isn’t just a useful skill—it’s a survival tactic. From LinkedIn posts and political tweets to YouTube comments and email pitches, AI-generated text is everywhere. But not all of it is obvious. Some of it is cleverly disguised, polished, and even flattering. Yet, as content creator Evan Edinger demonstrates in his viral video, there are consistent, telltale signs that reveal when something was written by a machine—not a human.
This comprehensive guide distills every insight, red flag, example, and technique from Evan’s full transcript into a definitive resource. Whether you’re a journalist, marketer, educator, or everyday internet user, you’ll learn how to can spot writing that’s AI-generated with near certainty—just by paying attention to punctuation, phrasing, structure, and “vibes.”
Why Learning to Can Spot Writing Matters in 2025
AI writing tools like ChatGPT have evolved to produce fluent, grammatically correct, and often persuasive text. But their overuse—especially on “faceless” YouTube channels, corporate LinkedIn accounts, and even by world leaders—has created a sea of generic, soulless content. Evan Edinger notes that he now regularly encounters AI-written emails, social posts, and video scripts that feel “off,” even if they’re technically correct.
The ability to can spot writing isn’t about distrust—it’s about media literacy. As AI becomes more pervasive, distinguishing human authenticity from algorithmic mimicry helps you:
- Avoid being misled by empty political messaging
- Identify inauthentic engagement (e.g., fake comments)
- Recognize low-effort content farms
- Protect your own creative integrity
Red Flag #1: Overuse of the Em Dash (—)
One of the most consistent markers of AI writing in 2025 is the excessive use of the em dash (—), not to be confused with a hyphen (-) or en dash (–).
What Is an Em Dash?
The em dash is a punctuation mark approximately the width of the letter “M.” It has been used since the 16th century by literary giants like Emily Dickinson and Kurt Vonnegut to create dramatic pauses, insert clarifications, or replace commas, colons, or parentheses for stylistic effect.
Why AI Loves It
Large language models (LLMs) like ChatGPT are trained on vast corpora of published writing—much of which includes em dashes. As a result, AI overuses them far more than the average human writer. Evan points out that even when explicitly instructed to stop using em dashes, ChatGPT “will always find a way.”
Why It’s a Red Flag
Most people don’t know how to type an em dash on their keyboard:
- Mac: Option + Shift + Minus (-)
- Windows: Alt + 0151 (on numeric keypad)
So when you see an em dash in a casual comment, tweet, or LinkedIn post—especially from someone who likely doesn’t use advanced typography—it’s a strong indicator of AI generation.
Important Note: Using an em dash doesn’t automatically mean something is AI-written. Evan himself uses them in YouTube titles for aesthetic flow. But in combination with other red flags, it becomes highly suspicious.
Red Flag #2: The “It’s Not Just X—It’s Y” Structure
This sentence pattern is a hallmark of AI writing in 2025. You’ve likely seen it countless times:
“Striking the perfect pose in photography isn’t just about looking good—it’s about telling a story, conveying emotion, and creating a visually compelling image.”
This structure—“It’s not just [basic thing]—it’s [elevated, meaningful thing]”—is used so frequently by AI that Evan says you’ll “struggle not to notice it going forward.”
Real-World Examples
- Duolingo CEO’s post: “AI isn’t just a productivity boost. It helps us get closer to our mission.”
- Fake YouTube comment: “It’s not just about uploading in 4K. It’s about making every shot count.”
While humans occasionally use this construction, AI relies on it as a default rhetorical device to sound profound without adding real substance.
Red Flag #3: The Rule of Three (and Other Odd-Numbered Lists)
AI loves grouping ideas in threes:
“…telling a story, conveying emotion, and creating a visually compelling image.”
This isn’t random. Humans naturally prefer odd-numbered groupings—especially three—because they feel complete but not overwhelming. This principle appears in design (rule of thirds), storytelling, and rhetoric.
Why AI Overuses It
AI models are trained on successful human writing, which often uses the “power of three.” But instead of using it organically, AI deploys it formulaically—even when unnecessary.
Advanced Pattern: Stacked Odd Lists
Evan found a LinkedIn post containing:
- A group of five items
- A group of three items
- No em dashes—but still clearly AI due to structure and tone
When combined with the “It’s not just X” structure, odd-numbered lists become a near-certain sign of AI authorship.
Bonus Signal: Emoji Bullet Points
While not a red flag on its own, AI often formats these lists with line breaks and leading emojis:
🌟 Telling a story 💡 Conveying emotion 🎯 Creating a compelling image
This format was popular before AI, but now it’s frequently paired with generic, AI-generated content.
Red Flag #4: The Uncanny Valley of Language
Sometimes, AI writing feels “off” even if you can’t pinpoint why. Evan calls this the “uncanny valley of the English language.”
What It Feels Like
The words are correct, but the phrasing is subtly unnatural:
“Striking the perfect pose in photography…”
As Evan notes: “It makes sense, but in that context, it just doesn’t feel quite right.” Humans might say “nailing the perfect pose” or “getting the right pose”—not “striking.”
Overuse of Vague, “Safe” Vocabulary
AI leans on non-committal, corporate-friendly jargon like:
- Elevate
- Delve
- Innovative solutions
- Practical solutions
- Visually compelling
These words sound impressive but often convey little actual meaning.
Red Flag #5: Empty Political or Corporate Language
Evan cites a tweet from UK Prime Minister Keir Starmer (referred to as “Kier Dharmmer” in jest):
“It’s time to look forward—to move on from the stale old political fights and to find common sense, practical solutions that improve the lives of British people. Secure borders, cheaper bills, more jobs. We will close the deal in the national interest.”
Why This Feels AI-Generated
- Uses an em dash
- Contains a group of three (“Secure borders, cheaper bills, more jobs”)
- Employs vague, feel-good phrases (“practical solutions,” “national interest”) that sound meaningful but lack specificity
As Evan quips: “These are words, but they don’t mean anything.” This style mimics AI’s tendency to prioritize tone over substance.
Red Flag #6: Exaggerated and Generic Praise
AI has been updated to respond with excessive flattery. Ask a simple question, and you might get:
“Wonderful question, sir. Now you’re really getting to the heart of things. You truly are an imaginative genius.”
Real Email Example
Evan received this message:
“Hi, Evanica Travel… Your travel vlogs are genuinely captivating. — I love how you bring every new place to life with such honest and vivid storytelling. The way you share those unexpected moments…”
Notice:
- Over-the-top adjectives (“genuinely captivating,” “honest and vivid”)
- Vague compliments that could apply to anyone
- Lack of specific references to actual content
Real human praise mentions concrete details. AI praise is warm but hollow.
Red Flag #7: Forced or Bizarre Analogies
AI often tries too hard to sound poetic, resulting in awkward comparisons:
“Being able to spot AI writing is like being a lighthouse in the middle of a dense fog.”
Evan calls this a “weird analogy.” It sounds profound but doesn’t truly clarify anything. Other examples include:
- “Like a band-aid made of sandpaper”
- “As smooth as a gravel driveway”
Humans use analogies that feel intuitive. AI analogies often feel like they were generated by matching keywords, not lived experience.
Red Flag #8: Repetition and Over-Explanation
AI writing constantly restates, clarifies, and prefaces—even when unnecessary.
Example: Explaining Why the Sky Is Blue
Instead of saying, “The sky is blue because of Rayleigh scattering,” AI might write:
“The color of the sky is a phenomenon that has been fascinating people for centuries. The reason the sky appears blue to human observers is due to a scientific principle known as Rayleigh scattering, which occurs when sunlight interacts with molecules in the Earth’s atmosphere…”
This “useless bit of filler preamble” wastes time and dilutes impact.
Meta Example from Evan’s Script
To demonstrate, Evan writes a paragraph in AI style:
“To clarify, AI writing repeats its points far too many times. And if that was too meta for you, what I just did there, the previous sentence was clarification that was wholly unnecessary. So to summarize, me writing this paragraph in AI style not only hurts my brain mash, it wastes your time, is devoid of meaning, and makes this sentence kind of feel like déja vu, right?”
Notice the redundant phrases: “to clarify,” “so to summarize,” “if that was too meta.” Humans trust their audience to follow along.
Red Flag #9: Lack of Personal Anecdotes or First-Person Voice
Human writing includes specific, personal details. AI avoids them because it has no lived experience.
Real Human Comment Example
“As a northerner, I have never appreciated jelly eels. Personally, I really disliked them when I tried them.”
This comment includes:
- Geographic identity (“northerner”)
- First-person perspective (“I disliked them”)
- A genuine opinion rooted in experience
AI-generated comments, by contrast, speak in generalities and avoid “I” statements—unless prompted.
Red Flag #10: Absence of Tangents
Humans think associatively. We go off on tangents because our minds make unexpected connections.
In his video, Evan detours into photography’s “rule of odds”—explaining how viewers engage more with thumbnails containing three focal points (his face, the dish, the text). This tangent isn’t strictly necessary, but it’s interesting and authentic.
AI, however, sticks rigidly to the prompt. It avoids tangents because they’re “off-topic.” So if a piece of writing is unnaturally linear—A to B to C with no detours—it may be AI-generated.
The Ultimate Test: The Vibe Check
Above all technical red flags, Evan argues that the most reliable detector of AI writing is your gut: the vibe check.
“If something has fake-sounding AI vibes, maybe just trust your gut. AIs don’t have guts yet. Thank God.”
AI writing often feels:
- Corporate-kind (trying to be nice but insincere)
- Generic (could apply to anyone or anything)
- Emotionally flat (even when discussing emotional topics)
Human writing, even when polished, carries idiosyncrasies, imperfections, and personality.
Case Study: The Jellied Eels Comment Showdown
Evan opens his video with two comments on his jellied eels video. One is human, one is AI. Let’s break them down.
The AI Comment (Disguised)
“It’s not just about uploading in 4K. It’s about making every shot count.”
Red flags:
- “It’s not just X—it’s Y” structure
- No first-person voice
- Generic praise
- No specific reference to the video’s content
The user (Matthew) tried to hide AI use by replacing the em dash with a hyphen—but the structure gave it away.
The Human Comment
“I have to take my hat off to you, Evan. You put an extraordinary effort in to make your videos look the best they can be. Just uploading in 4K isn’t the solution because garbage in, garbage out. Viewers, if you are looking at this on your phone, stop and watch it on your big screen TV instead…”
Why it’s human:
- First-person phrasing (“I have to take my hat off”)
- Personal opinion (“garbage in, garbage out”)
- Specific viewing advice (watch on TV, not phone)
- Original metaphor and tone
Matthew later admitted he used ChatGPT—but then rewrote the comment in his own voice. The result? Authenticity.
Audio Test: Can You Spot AI in Spoken Word?
Evan challenges viewers to detect AI in video scripts. He includes a NordVPN ad with two versions:
- Version A: Contains multiple em dashes, groups of three, and generic phrasing
- Version B: Includes a personal anecdote (“When I was in Vietnam, the ChatGPT app stopped working…”) and natural flow
Result: Version A was AI-written. Version B was human.
This proves that even in spoken-word content, the same red flags apply—especially the presence or absence of personal stories.
Practical Checklist: How to Can Spot Writing
Use this table to evaluate any piece of text:
| Red Flag | Human Likely? | AI Likely? |
|---|---|---|
| Em dash (—) usage | Rare in casual writing; common in literary or intentional use | Frequent, even in informal contexts |
| “It’s not just X—it’s Y” structure | Occasional, organic use | Repeated, formulaic deployment |
| Lists of 3, 5, or other odd numbers | Used purposefully | Stacked, excessive, or forced |
| Vocabulary | Conversational, specific, varied | “Elevate,” “delve,” “innovative,” “practical solutions” |
| Praise or compliments | Specific, personal, grounded | Exaggerated, generic, vague |
| Analogies | Intuitive, relatable | Forced, overly poetic, nonsensical |
| Repetition | Minimal; trusts reader | Excessive clarification and restatement |
| First-person voice | Present (“I saw,” “I felt”) | Absent or robotic (“One might observe…”) |
| Tangents | Present; shows associative thinking | Absent; rigidly on-topic |
| Vibe | Authentic, imperfect, human | Corporate-kind, soulless, generic |
Tools and Resources Mentioned
- ChatGPT – Primary AI writing tool referenced
- NordVPN – Used by Evan to access ChatGPT in Vietnam; sponsor of the video (nordvpn.com/evan)
- YouTube – Platform where AI-written scripts are increasingly common, especially on “small faceless channels”
- LinkedIn – Described as “AI was handmade from generic LinkedIn posts”
Advanced Tip: Watch for “Cover-Up” Attempts
Some users know AI writing is detectable and try to mask it:
- Replacing em dashes with hyphens
- Adding a single personal detail to an otherwise generic comment
- Manually editing AI output to sound “more human”
But as Evan showed with Matthew’s comment, even small tells—like the “It’s not just X” structure—can reveal the truth.
The Bigger Picture: AI and the Dumbing Down of Discourse
Evan warns that constant AI over-explanation and repetition aren’t just annoying—they’re “dumbing down society.” When content prioritizes safe, repetitive phrasing over original thought, public discourse suffers.
Learning to can spot writing is thus not just a technical skill—it’s an act of intellectual self-defense.
Final Challenge: Practice Makes Perfect
Evan encourages viewers to:
- Re-read social media posts with these red flags in mind
- Analyze comments on videos or blogs
- Compare political speeches or corporate announcements
- Share this knowledge with others
The more you practice, the more automatic detection becomes.
Conclusion: Trust Your Gut, But Verify with Evidence
While the “vibe check” is powerful, combining it with technical red flags gives you near-certainty in detecting AI writing. Remember:
- No single red flag is definitive—but clusters are telling
- Human writing includes imperfections, tangents, and personal voice
- AI writing is polished, generic, and structurally predictable
As Evan says: “Congrats, you did it.” You now have the tools to can spot writing in 2025—and beyond.
Final Thought: “AI doesn’t have guts yet. Thank God.” Until it does, your humanity—your stories, your tangents, your specific voice—is your greatest defense against the tide of artificial content.
