0:00 / 0:00

Comment “AI” and get free AI workflow👇 Have you also seen the AI twin courses? You don’t need them. Just 1️⃣ Follow @shedoesai 2️⃣ Grab my new @weavy_ai workflow #aiworkflow #weavyai #aitwin

Gunhild Johanne Reumert

@shedoesai · Digital creator

INSTAGRAM · 2025-12-09Source

323likes

1.2Kcomments

Remix This

Recreate with Runway Gen 4.5

Make your own AI viral video

Prompt

MASTER PROMPT

GLOBAL LOCK: A vertical split-screen composition. The left half features a 30-year-old Caucasian woman with blonde hair tied back, wearing a chunky pink knit sweater and blue jeans, sitting by a window in a cozy cafe with warm, natural daylight. The right half features the exact same woman, wearing a plain white t-shirt, sitting at a modern desk with a silver laptop, illuminated by cool, soft office lighting. Both sides maintain photorealistic cinematic quality, 35mm lens feel, soft depth of field, and identical facial features.

[00:00–00:05] Left side: The woman sits relaxed, holding a white ceramic mug with both hands near her chest. She looks out the window to her left with a gentle, serene expression. Subtle movement in her hair and slight breathing motion. Right side: The woman is focused, looking down at the silver laptop screen, typing on the keyboard. At 00:02, she pauses typing, reaches with her right hand to pick up a grey ceramic mug on the desk, brings it to her lips to take a sip, places it back down, and resumes typing. The camera remains completely static on both sides throughout the duration. No speech.

NEGATIVE PROMPT
text, watermarks, logos, split-screen bleeding, morphing faces, inconsistent identity between left and right, unnatural hand anatomy, extra fingers, flickering lighting, temporal jitter, robotic movements, exaggerated expressions, harsh shadows, low resolution.

SHOT PROMPTS
Shot 1 (Left Side Base): A 30-year-old Caucasian woman with blonde hair tied back, wearing a chunky pink knit sweater, sitting by a window in a cozy cafe, warm natural daylight, holding a white ceramic mug, looking out the window, serene expression, cinematic, 35mm lens, soft depth of field.
Shot 2 (Right Side Base): A 30-year-old Caucasian woman with blonde hair tied back, wearing a plain white t-shirt, sitting at a modern desk typing on a silver laptop, cool soft office lighting, focused expression, cinematic, 35mm lens, soft depth of field.

SPEECH PACK
speech_present: false
transcript_segments: []
delivery_direction: N/A
mic_room_signature: N/A
sync_requirements: N/A
mix_notes: Background music only, no dialogue or voiceover.

How shedoesai Made This AI Twin Split-Screen Workflow and How to Recreate It

This 5-second Instagram Reel features a cinematic AI twin split-screen, contrasting a warm, cozy cafe aesthetic (pink sweater, natural light) with a cool, focused workspace environment (white t-shirt, laptop), effectively demonstrating the power of AI video generation tools like Google Veo and Kling to create seamless, consistent B-roll for creators.

2. What You're Seeing

Visual Breakdown

The video utilizes a strict 50/50 vertical split-screen layout. On the left, the subject—a blonde woman—is dressed in a chunky pink knit sweater and blue jeans, sitting in a cafe setting. The lighting here is warm and directional, mimicking natural sunlight coming from the left window. Her actions are passive: holding a white mug and looking away. On the right, the exact same subject appears in a plain white t-shirt at a desk. The lighting shifts to a cooler, diffused office tone. Her actions are active: typing on a silver laptop, reaching for a grey mug, taking a sip, and returning to work. Bold, sans-serif text overlays dominate the top ("STEP 6: ANIMATE...") and bottom, framing the visual content.

Shot-by-Shot Breakdown

Time Range	Visual Content	Shot Language	Lighting & Color Tone	Viewer Intent
00:00 - 00:02	Left: Woman holding mug, looking out. Right: Woman typing on laptop. Top/bottom text visible.	Static medium shots on both sides. 35mm focal length feel.	Left: Warm, high-key natural light. Right: Cool, soft office light.	Establish the "AI Twin" concept immediately; hook the viewer with the visual contrast.
00:02 - 00:04	Left: Subtle breathing/hair movement. Right: Woman stops typing, picks up grey mug, takes a sip.	Static framing maintained. Soft depth of field blurs the backgrounds.	Consistent with the opening. No lighting shifts.	Demonstrate the AI's ability to handle complex object interaction (drinking from a cup) without morphing.
00:04 - 00:05	Left: Continues looking out window. Right: Places mug down, resumes typing.	Static framing.	Consistent.	Complete the action loop, encouraging the viewer to rewatch the short 5-second clip.

3. Why It Went Viral

The Psychology of the "AI Twin"

This topic strikes a chord because it addresses a massive pain point for indie creators: burnout. The idea of an "AI Twin" taps into the biological instinct to conserve energy while maximizing output. By positioning this short clip as an alternative to expensive "AI twin courses" (as explicitly stated in the caption: "Have you also seen the AI twin courses? You don't need them"), the creator leverages a Robin Hood dynamic—giving away premium secrets for free. The visual proof of the twin working in two distinct environments (cafe vs. office) provides immediate credibility, satisfying the viewer's skepticism about AI consistency.

Platform Signals and Watch Time

From an algorithmic perspective, this video is engineered for retention. The 5-second duration is incredibly short, meaning it loops before the viewer has finished reading the top and bottom text overlays. The split-screen format forces the eye to dart back and forth between the pink sweater scene and the white t-shirt scene, artificially inflating watch time. Furthermore, the caption's call-to-action ("Comment 'AI'") triggers an automated DM workflow, flooding the post with comments and signaling to Instagram that the content is highly engaging.

5 Testable Viral Hypotheses

The Split-Screen Loop Effect: Evidence: Two simultaneous actions (typing vs. resting). Mechanism: Cognitive overload prevents the viewer from processing the whole video in one 5-second pass, leading to automatic rewatches. Replication: Place two contrasting AI-generated clips side-by-side to force multiple views.
The "Step 6" Information Gap: Evidence: The top text reads "STEP 6: ANIMATE YOUR VIDEOS". Mechanism: Starting in the middle of a list creates an information gap. Viewers wonder what steps 1-5 were, driving them to the creator's profile. Replication: Label your standalone tips as part of a larger, numbered sequence.
Authority via Specificity: Evidence: The bottom text explicitly names "Google Veo 3.1" and "Kling 2.6 Pro". Mechanism: Naming specific, advanced tools elevates the perceived value of the video, making it a highly "saveable" resource. Replication: Don't just say "use AI"; name the exact software version you are using.
The "Comment to DM" Funnel: Evidence: Caption says "Comment 'AI' and get free AI workflow". Mechanism: Low-friction engagement. Users type two letters to receive a high-value reward, spiking the algorithm's engagement metrics. Replication: Set up a ManyChat automation and use a single-word CTA in your first line.
Visual Contrast Retention: Evidence: Warm lighting/pink clothing on the left vs. cool lighting/white clothing on the right. Mechanism: The stark color contrast keeps the visual cortex stimulated, preventing the scroll reflex. Replication: When generating AI clips, intentionally prompt for opposing color palettes (warm vs. cool).

4. How to Recreate

Phase 1: Topic Selection & Character Consistency

Topic Selection: Choose a "day in the life" or "workflow" theme. This split-screen format works best for educational accounts showing the balance between work (typing) and lifestyle (coffee).
Character Consistency: Use an image generator like Midjourney. Create a base character and use the --cref (character reference) parameter. Generate two images: one of the character in a pink sweater by a window, and another in a white t-shirt at a desk.

Phase 2: Keyframe Generation & Video Animation

Keyframe Prep: Upscale your two chosen images. Ensure the facial features match perfectly before animating.
Video Generation (Scene 1): Import the cafe image into Kling 2.6 Pro or Runway Gen-3. Prompt for subtle motion: "Gentle breathing, looking out the window, slight hair movement in the breeze."
Video Generation (Scene 2): Import the office image. Prompt for specific action: "Typing on a silver laptop, reaches for a mug, takes a sip, resumes typing." Keep the camera static in the prompt.

Phase 3: Editing, Cover & Publishing

Compositing: Open CapCut. Place Scene 1 on the main track and Scene 2 as an overlay. Use the mask tool to create a perfect 50/50 vertical split.
Text Overlays: Add bold, white sans-serif text at the top and bottom. Add a black background to the text for readability. Do not rely on AI to generate this text; do it in post-production.
Publishing Strategy: Export the video at exactly 5 seconds. Write a caption that challenges a common paid solution (e.g., "Don't buy courses") and offer your workflow for free via a comment CTA.

5. Growth Playbook

Ready-to-Use Opening Hooks

"Stop paying thousands for AI twin courses. Do this instead."
"How I cloned myself for infinite B-roll (Step-by-step)."
"The secret to posting 3 times a day without ever turning on a camera."

Caption Templates

The Workflow Reveal: [Hook] Want to clone yourself for free? [Value] I figured out the exact workflow to create consistent AI B-roll using just two tools. [Question] Have you tried making an AI twin yet? [CTA] Comment "TWIN" and I'll DM you the exact prompts I used.
The Time Saver: [Hook] Stop wasting time filming B-roll. [Value] By generating a digital twin, you can create cinematic clips while you sleep. [Question] Which vibe do you prefer, cozy cafe or focused office? [CTA] Save this post for your next content creation day.
The Tool Drop: [Hook] The AI video tools are getting too good. [Value] You can now get perfect lip sync and smooth cinematic clips without a camera using Kling and Veo. [Question] What would you use an AI clone for? [CTA] Follow for Step 7 of this workflow series.
The Anti-Guru: [Hook] Don't buy that AI course. [Value] Everything you need to know is in this simple 5-second workflow. [Question] Are you team Kling or team Veo? [CTA] Comment "WORKFLOW" to get the full PDF guide sent to your inbox.

Hashtag Strategy

Broad (High Volume): #AIVideo, #ContentCreation, #VideoEditing. These signal the general category to the algorithm, ensuring you are placed in the right content graph.
Mid-tier (Targeted): #AITwin, #DigitalClone, #CreatorTips. These target creators specifically looking for solutions to scale their presence without burning out.
Niche Long-tail (High Intent): #AIWorkflowTutorial, #KlingAI, #CapCutHacks. These capture users actively searching for specific software tutorials and step-by-step guides.

6. FAQ

What tools make it look the most similar to this video?

Midjourney for the base image consistency, and Kling AI or Google Veo for the smooth, realistic motion.

What are the 3 most important words in the prompt?

"Cinematic," "Photorealistic," and "Consistent character."

Why does the generated face look inconsistent between the two sides?

You likely didn't use a strong character reference image (--cref) or the lighting differences were too extreme in the base prompt.

How can I avoid making it look like AI?

Keep the motion prompts subtle (e.g., "breathing," "typing") rather than complex full-body movements.

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

Instagram Reels currently favors this aesthetic, high-value, "comment-to-DM" workflow format.

How should I properly disclose AI use for this type of content?

Use the platform's built-in "AI Generated" label and mention it clearly in your hashtags and caption.