0:00 / 0:00

Tea with Lucifer and Woody Made by @klingai_official 3.0 — the best commercial-grade AI video model. Native 1080p output, built for brand storytelling, and professional film production. #klingai #Kling3 #WoodyAllen #Lucifer #scandal

Karol Życzkowski

@dreamweaver_ai_pl · creator

INSTAGRAM · 2026-02-24Source

235likes

9comments

Remix This

Recreate with AI Meme Generator

Make your own AI viral video

Prompt

GLOBAL LOCK: Two consistent characters sitting on a yellow sofa in a warm, cluttered study filled with books and lamps. Character A: Elderly Caucasian male, thin, white hair, thick black-rimmed glasses, wearing a white knit sweater with a brown pattern over a light blue collared shirt. Character B: Red-skinned humanoid Lucifer, large curved ram-like horns with realistic texture, pointed ears, small goatee, wearing a dark grey textured V-neck sweater. Cinematic lighting, photorealistic skin textures, 1080p, 24fps.

[00:00–00:01] Wide shot of both characters on the sofa. They are both laughing heartily, bodies shaking slightly. Both are holding white ceramic mugs. The background is a warm, out-of-focus study.

[00:01–00:04] Medium close-up on Character A (Woody lookalike). He is smiling and talking with an animated, neurotic expression. He gestures slightly with his head. "Lucifer, are you close with your family?" Lips are clearly visible and must sync to the dialogue.

[00:04–00:06] Medium close-up on Character B (Lucifer). He looks down slightly, his expression turning somber and reflective. He shakes his head slowly. "Not really lately." High detail on the red skin and horn ridges.

[00:06–00:10] Medium close-up back to Character A. He has a concerned, "advice-giving" expression. He blinks and leans in slightly. "Oh, that's not good, man. Being close with family is really important." Soft lighting on his face, emphasizing wrinkles and glasses.

[00:10–00:13] Close-up on Character B (Lucifer). He lets out a loud, boisterous laugh, head tilting back, mouth wide open showing teeth. 

[00:13–00:17] Character B stops laughing and leans forward with a sharp, mischievous, and slightly menacing grin. He delivers the punchline: "Oh my god... I don't want to be anywhere near as close to my dad as you are with your daughter." Tight framing on his face. The cut lands on the word "daughter" for maximum impact.

NEGATIVE PROMPT: robotic movement, plastic skin, flickering lights, inconsistent horn shape, blurry glasses, floating mugs, unnatural lip-sync, distorted fingers, morphing background, low resolution, watermark, text overlays.

SPEECH PACK:
[00:01–00:04] "Lucifer, are you close with your family?"
TAKE_A: (Inquisitive, high energy, slight stutter) "Lucifer... are you, uh, close with your family?"
TAKE_B: (Warm, friendly) "Lucifer, are you close with your family?"
TAKE_C: (Fast-paced, neurotic) "Lucifer-are-you-close-with-your-family?"

[00:04–00:06] "Not really lately."
TAKE_A: (Deep, gravelly, sad) "Not... really... lately."
TAKE_B: (Short, dismissive) "Not really lately."

[00:06–00:10] "Oh, that's not good, man. Being close with family is really important."
TAKE_A: (Preachy, sincere) "Oh, that's not good, man. Being close with family is **really** important."

[00:10–00:17] "Oh my god... I don't want to be anywhere near as close to my dad as you are with your daughter."
TAKE_A: (Laughing first, then sharp/sarcastic) "[Laugh] Oh my god... I don't want to be anywhere near as close to my dad as **you** are with your daughter."

How dreamweaver_ai_pl Made This Tea with Lucifer and Woody AI Video — and How to Recreate It

This case study analyzes a high-fidelity AI-generated skit featuring a satirical conversation between a Woody Allen lookalike and a cinematic Lucifer. Utilizing the Kling 3.0 model, the video achieves professional-grade character consistency and lip-syncing. The aesthetic is a "warm, cluttered intellectual's study" paired with high-contrast, prosthetic-level creature design. By blending celebrity controversy with supernatural absurdity, the creator taps into "dark humor" and "pop culture scandal" niches, resulting in high engagement through shock value and technical curiosity.

What You’re Seeing

The video is a multi-shot dialogue sequence. It begins with a wide shot establishing the two characters on a yellow sofa, then moves into a series of medium close-ups (MCUs) to emphasize facial expressions and dialogue delivery. The lighting is "motivated" by warm indoor lamps, creating soft shadows that give the AI-generated skin a realistic, non-plastic texture.

Shot-by-Shot Breakdown

Time Range	Visual Content	Shot Language	Lighting & Tone	Viewer Intent
00:00–00:01	Woody and Lucifer laughing together on a sofa, holding mugs.	Wide Shot (WS), static.	Warm, cozy, high-key.	Establish the "absurd buddy" premise.
00:01–00:04	Woody speaking directly to Lucifer.	Medium Close-Up (MCU), 50mm feel.	Soft side-lighting, shallow DOF.	Hook the viewer with the celebrity voice/likeness.
00:04–00:06	Lucifer responding with a somber look.	Over-the-shoulder (OTS) to MCU.	Red skin tones, detailed horn texture.	Build character depth and "straight man" dynamic.
00:06–00:10	Woody giving "fatherly" advice.	MCU, slight camera push-in.	Warm, sincere aesthetic.	Lull the viewer into a sense of normalcy.
00:10–00:17	Lucifer laughs, then delivers a sharp, scandalous punchline.	Tight Close-Up (CU).	High contrast, emphasizing the devil's grin.	The "Payoff": Shocking satire that drives shares.

Why It Went Viral: The "Scandalous Satire" Hook

The Power of the "Forbidden" Punchline

The core of this video's success is its fearless use of satire. By referencing Woody Allen's real-life family controversies through the mouth of a literal demon, the creator creates a "cognitive dissonance" that forces a reaction. Viewers are either shocked, amused, or offended—all of which lead to high comment counts and shares. This isn't just "AI art"; it's AI-enabled social commentary.

The "Uncanny Valley" as a Feature, Not a Bug

While many creators try to avoid the "uncanny valley," this video leans into it. The hyper-realistic red skin of Lucifer and the slightly exaggerated "neurotic" movements of the Woody character make the scene feel like a high-budget prosthetic film from the 90s. This nostalgic, cinematic quality (Kling 3.0's 1080p output) signals "high effort" to the platform's algorithm, increasing its distribution.

Platform Perspective: Retention & Controversy

From an Instagram/TikTok perspective, the video succeeds because of Dialogue Retention. Users stay to hear the end of the conversation. The "scandal" hashtag and the specific mention of "daughter" in the punchline act as a "save-worthy" moment—users share it with friends to say, "Did you see what they made the AI say?" This "mild controversy" is a proven growth lever when handled within platform guidelines.

5 Testable Viral Hypotheses

The Absurd Pairing: Placing a historical/celebrity figure with a mythological creature (e.g., Napoleon and a Minotaur) creates instant visual curiosity.
The "Wait for the Burn": A 15-second setup leading to a 2-second "savage" insult maximizes watch time.
Hyper-Specific Textures: Using AI models that can handle complex textures (horns, wrinkles, knit sweaters) increases perceived production value.
Voice Mimicry: Using a voice that perfectly matches a famous person's cadence (the "Woody stutter") creates a stronger emotional connection than a generic AI voice.

The "Tea/Coffee" Trope:

How to Recreate: From 0 to 1

Step 1: Topic Selection & Scripting

Choose a "high-contrast" pair. One should be a recognizable human archetype (the intellectual, the jock, the grandma) and the other a fantasy creature. Write a script that starts friendly and ends with a "roast" or a "truth bomb."

Step 2: Character Consistency (The "Lock")

Use a reference image for your characters. In Kling or Midjourney, define the "Woody" character by his glasses, sweater, and age. Define the "Lucifer" by his skin color (hex code), horn shape, and clothing. Tip: Keep the clothing simple (sweaters) to avoid AI "hallucinations" during movement.

Step 3: Environment Setup

Generate a background image of a "cluttered study" or "cozy living room." Use this as a consistent backdrop for all shots to ensure the scene doesn't "jump" between cuts.

Step 4: Video Generation (Shot-by-Shot)

Don't try to generate the whole 17 seconds at once. Generate the wide shot first, then use "Image-to-Video" for the close-ups to maintain facial features. Use the "Master Prompt" provided below for the motion cues.

Step 5: Voice Synthesis

Use a tool like ElevenLabs. For the "Woody" voice, use a "neurotic, elderly, New York" preset. For Lucifer, use a "deep, gravelly, charismatic" preset. Ensure the pacing includes pauses for "thinking" (e.g., "uh," "well...").

Step 6: Lip-Syncing

Upload your generated video and the audio to a lip-sync tool (like Sync Labs or Kling's native lip-sync feature). This is crucial for the "realism" factor.

Step 7: Editing & Pacing

Cut the video on the "beat" of the dialogue. When a character finishes a sentence, cut to the reaction of the other character. This mimics professional film editing.

Step 8: Publishing Strategy

Use a "hook" title on the video itself, like "Lucifer has no chill." Post during peak hours for your target region (e.g., 6 PM EST for US audiences).

Growth Playbook: Distribution & Scaling

3 Opening Hook Lines

"The crossover nobody asked for, but everyone needed. ☕️"
"Lucifer just went there... 💀"
"AI is getting too real with these roasts."

4 Caption Templates

The "Technical" Hook: "Kling 3.0 is changing the game. Look at the texture on those horns! 🤯 What should I make next? #klingai #aivideo"
The "Engagement" Question: "Who won this roast: Woody or Lucifer? 👇 Let me know in the comments! #darkhumor #satire"
The "Short & Punchy": "Tea time in the underworld. ☕️🔥 #lucifer #woodyallen #scandal"
The "Value" Play: "How I made this cinematic AI skit in 3 steps. (Check the link in bio for the full tutorial!) 🎬 #indiecreator #aiart"

Hashtag Strategy

Broad: #AI #Cinema #ShortFilm #Comedy (To reach a wide audience)
Mid-tier: #KlingAI #DigitalHuman #DarkHumor #CGI (To reach tech and niche comedy fans)
Niche: #WoodyAllenSatire #LuciferSeries #AIPortrait #IndieCreator (To dominate specific search terms)

FAQ

What tools make it look the most similar?

Kling 3.0 for video generation and ElevenLabs for the voice cloning/synthesis.

What are the 3 most important words in the prompt?

"Cinematic lighting," "Subsurface scattering" (for the red skin), and "Micro-expressions."

Why does the generated face look inconsistent?

Usually because the "Global Lock" isn't specific enough about facial features like "thick black-rimmed glasses."

How can I avoid making it look like AI?

Add "film grain" and "slight camera shake" in post-production to mimic handheld filming.

Is it easier to go viral on Instagram or TikTok?

TikTok favors the "shock" punchline, while Instagram favors the "cinematic aesthetic" of the Kling model.