0:00 / 0:00

Instagram vs Reality 😂🤥 Comment “AI” to do this for yourself on Higgsfield Soul 2 is Higgsfield’s new image model designed for high-fidelity, realistic character generation with preset styles that make consistency effortless. When combined with Kling Motion, you can recreate entire scenes with cinematic movement, giving creators a powerful workflow for AI filmmaking, ads, and storytelling. This is a paid promotion for Higgsfield #HiggsfieldAI #Soul2 #KlingMotion #AIVideoCreation #GenerativeAI

Rourke Sefton-Minns

@rourke · creator

INSTAGRAM · 2026-02-20Source

2.1Klikes

529comments

Remix This

Recreate with AI Meme Generator

Make your own AI viral video

Prompt

GLOBAL LOCK: 
Subject: A Caucasian male in his late 20s with a short brown beard and mustache. He wears a variety of casual headwear (green trucker hat, blue baseball cap, tan cap) and hoodies (brown, grey). 
Visual Style: Split-screen composition. Top half is cinematic, high-fidelity AI video with vibrant colors and professional grading. Bottom half is UGC-style, handheld or static phone footage in a domestic indoor setting with natural/practical lighting.
Consistency: The subject's facial features and beard must remain identical across all AI-generated scenes, matching the real person in the bottom half.
Speech: Energetic, fast-paced narration with clear enunciation. Mic is close-up, dry, and professional.

[00:00–00:06]
Top: A cinematic wide shot of the Leaning Tower of Pisa under a bright blue sky. The subject, wearing a white t-shirt and green trucker hat, stands in the foreground, smiling and holding his hands up as if leaning against the tower. High saturation, sharp details.
Bottom: The real subject in a home office, wearing a brown hoodie and blue cap, mimics the same pose against a black metal bookshelf.
Speech: "Hey, get this picture. No one's ever thought of this, but it's gonna look like I'm pushing the tower."

[00:07–00:09]
Top: A close-up of the subject in a snowy arctic environment, wearing a "Vans" t-shirt and green hat. He is playfully interacting with a large, realistic polar bear that is nuzzling his head. Cold blue color grade.
Bottom: The real subject in a hallway, wearing the brown hoodie, mimics the nuzzling motion against thin air.
Speech: "Okay guys, I'm here with a [unclear] polar bear."

[00:10–00:16]
Top: A wide shot at night in Giza, Egypt. The subject sits atop a camel with the Great Pyramid in the background. He is wearing a white t-shirt and shorts, gesturing with a "call me" sign. Warm, golden-hour lighting.
Bottom: The real subject sits on a white kitchen counter, mimicking the camel-riding posture and hand gesture.
Speech: "It's a Tuesday and I'm on a [unclear] camel. What do you mean you're at work? Just have your mom and dad pay for it."

[00:17–00:23]
Top: A low-angle medium shot in a sunny LA suburb. The subject sits on the ground in front of a bright red Ferrari. He wears a purple graphic tee and a tan hat, holding up a red car key fob. High-contrast, commercial aesthetic.
Bottom: The real subject sits on a wooden chair in his living room, holding up a small white object (a piece of cheese or soap) as if it were the key fob.
Speech: "Just bought my first car, age 23 by the way. Even got the keys. Whew!"

[00:24–00:48]
The video transitions to a full-screen UI walkthrough of the Higgsfield website. The subject appears in a circular talking-head overlay at the bottom.
Visuals: Cursor navigates through "Create Image," "Higgsfield Soul," "Character Upload," and "Motion Control" menus.
Final Shot (00:43-00:46): A split screen showing the subject in a high-end casino wearing a white tuxedo (AI) vs. the subject in his home office (Real), both performing a "come here" hand gesture.
Speech: "You can do this for yourself by going to Higgsfield. Select image, then go to Higgsfield Soul... upload a bunch of images of yourself... choose the style you want... then go to Kling Motion Control... upload your driving video and the image... and it will create this effect. Comment AI and I'll send you the link."

NEGATIVE PROMPT: Visual artifacts, flickering, face swapping glitches, inconsistent beard shape, blurry textures in the AI half, robotic lip-sync, muffled audio, background noise, distorted limbs, unnatural camel movement.

SPEECH PACK:
[00:00-00:06] "Hey, get this picture. No one's ever thought of this, but it's gonna look like I'm pushing the tower."
TAKE_A: (Excited, fast) "Hey, get this shot! Nobody's done this, it'll look like I'm holding up the tower!"
TAKE_B: (Sarcastic, deadpan) "Check this out. Totally original. I'm pushing the Leaning Tower."

[00:10-00:16] "It's a Tuesday and I'm on a camel. What do you mean you're at work? Just have your mom and dad pay for it."
TAKE_A: (Arrogant influencer tone) "Tuesday vibes on a camel. Why are you working? Just get your parents to fund it."
TAKE_B: (Playful) "Just riding a camel on a weekday. Work? Never heard of her. Ask your parents for the cash."

How rourke Made This Higgsfield Instagram Vs Reality AI Video — and How to Recreate It

This viral case study features creator @rourke demonstrating the power of Higgsfield Soul 2 through the classic "Instagram vs. Reality" trope. By splitting the screen, the video contrasts high-fidelity, cinematic AI-generated travel and luxury scenes with the mundane, low-budget "reality" of a home office. The video effectively hooks viewers with humor and "impossible" shots (like hugging a polar bear) before pivoting into a high-value tutorial on character consistency and motion control. It’s a masterclass in using AI character generation to create "scroll-stopping" content that feels both aspirational and relatable.

What You’re Seeing: A Visual Breakdown

The video utilizes a vertical split-screen format. The top half displays "Instagram"—vibrant, color-graded, and perfectly composed AI videos. The bottom half shows "Reality"—the creator in a dimly lit room, wearing casual clothes, and using household items as props to mimic the AI's movements. This visual contrast is the core engine of the video's engagement.

Shot-by-Shot Breakdown

Time Range	Visual Content	Shot Language	Lighting & Tone	Viewer Intent
00:00–00:06	Leaning Tower of Pisa. AI leans against it; Real creator leans against a shelf.	Medium Shot (MS), static. Split-screen comparison.	Top: Bright daylight, high saturation. Bottom: Warm indoor practical light.	The Hook: Establish the "Instagram vs Reality" joke immediately.
00:07–00:09	Polar Bear encounter. AI hugs a bear; Real creator mimics the gesture in his hallway.	Close-up (CU). Fast-paced transition.	Top: Cold, blue "arctic" tones. Bottom: Flat, neutral hallway lighting.	Shock Value: Show something physically impossible to grab attention.
00:10–00:16	Riding a camel at the Pyramids. AI is on a camel; Real creator sits on a kitchen counter.	Wide Shot (WS) for AI; MS for Reality.	Top: Golden hour, cinematic desert glow. Bottom: Harsh kitchen overhead light.	Humor: The absurdity of the kitchen counter "camel" creates a laugh.
00:17–00:23	Ferrari in LA. AI sits by a red supercar; Real creator sits on a wooden chair.	Low-angle MCU.	Top: High-contrast, "commercial" look. Bottom: Soft, natural window light.	The "Flex": Satirizing luxury influencer culture.
00:24–00:48	Tutorial UI walkthrough. Creator explains the Higgsfield interface.	Screen recording + Talking head overlay.	Clean, digital UI; Creator in consistent home setting.	Retention/Value: Transition from entertainment to "how-to" education.

Why It Went Viral: The "Fake Flex" Psychology

The Power of the Trope

The "Instagram vs Reality" format is a perennial winner because it taps into social comparison theory. Viewers love seeing the "curtain pulled back" on perfect social media lives. By using AI to create the "Instagram" side, @rourke adds a layer of technological novelty. It’s not just a joke; it’s a demonstration of a "superpower" that viewers want for themselves.

The "Aha!" Moment

The video succeeds because it identifies a specific pain point for creators: character consistency. Most AI video tools struggle to keep the same face across different scenes. By showing himself as the consistent character in wildly different environments, the creator proves the tool's utility. This "proof of concept" drives saves and shares from other creators looking for similar solutions.

Platform Signals & Engagement

From a platform perspective, the loop effect is strong. The first 3 seconds are so visually dense (Pisa vs. Bookshelf) that viewers often re-watch to catch the details. The "Comment 'AI'" call-to-action (CTA) is a classic engagement hack that triggers the algorithm to see the post as "highly conversational," boosting its reach to the Explore page.

How to Recreate: From Reality to AI Royalty

Identify Your "Flex" Scenes: Choose 3-4 iconic, high-budget locations (Eiffel Tower, Private Jet, Underwater) that contrast sharply with your actual home environment.
Film the "Reality" Base: Record yourself in your home mimicking the actions you want your AI avatar to perform. Use props (a chair for a car, a shelf for a landmark). Keep your movements deliberate.
Character Training: Upload 20+ clear photos of yourself to Higgsfield Soul 2. Ensure different angles and lighting to help the model learn your features.
Generate the "Instagram" Stills: Use the "Presets" feature (like "Flash Editorial" or "Theatrical Light") to generate high-quality images of your character in those dream locations.
Apply Motion Control: Use the "Motion Control" feature. Upload your "Reality" video as the driving motion and your AI-generated character image as the subject.
Split-Screen Editing: Use CapCut or Premiere Pro to create a vertical split. Place the AI video on top and the "Reality" video on the bottom.
Sync the Audio: Use a trending, upbeat track or record an energetic voiceover explaining the "lie."
Add Text Overlays: Use a clear, bold font for the "Instagram VS Reality" header to provide immediate context.

Growth Playbook: Captions & Strategy

3 Ready-to-Use Hooks

"Stop traveling. Start prompting. ✈️🤖"
"My travel budget is $0, but my feed looks like a million bucks."
"The secret influencers don't want you to know... (it's AI) 😂"

Caption Templates

The "Value-First" Template:
Instagram vs Reality is getting out of hand! 😂 I created all these 'travel' videos without leaving my living room using [Tool Name]. Here’s how you can maintain character consistency for your own AI films. [Value Point: Mention Soul 2]. Want the link? Comment 'AI' below! 👇

Hashtag Strategy

Broad: #AI #ArtificialIntelligence #ContentCreator #DigitalArt
Mid-Tier: #AIVideo #Higgsfield #InstagramVsReality #CreativeTech
Niche: #CharacterConsistency #AIForCreators #IndieMaker

Frequently Asked Questions

What tools make it look the most similar?

Higgsfield Soul 2 is specifically designed for this type of character-driven motion control.

What are the 3 most important words in the prompt?

"Photorealistic," "Consistent Character," and "Cinematic Lighting."

Why does the generated face look inconsistent?

Usually, this is due to poor quality or too few training images of the original person.

How can I avoid making it look like AI?

Use "Motion Control" to ground the AI's movements in real human physics from your source video.

Is it easier to go viral on Instagram or TikTok with this?

Instagram Reels currently favors high-aesthetic "visual magic" like this split-screen effect.

How should I properly disclose AI use?

Use the platform's built-in "AI-generated" label and mention it in your caption to build trust.