Grow with Dani

🎥 Digital Content Creator 🤖 I help creators build hyper-realistic AI Influencers & faceless YouTube channels that make money📍 New videos weekly HIGGS-POMRI

[Grow with Dani @GrowWithDani]

🚨 I hooked you up.

Watched my AI Influencer tutorial? Now build it yourself 👇

🎁 Get 1 FREE month of MindStudio Individual ($20 value)
Use code: GROWWITHDANIMS
Build the exact AI Agent here:
👉 https://t.co/iDZtlyHvvm...

Want to support the channel while following the workflow?
👉 https://t.co/lSHeP7jDcP

Who’s launching their AI Influencer this week? 🔥👇

#AIInfluencer #MindStudio #AIAutomation

[Grow with Dani @GrowWithDani]

I created my AI influencer using GPT Image 2 on Higgsfield AI.

If you want to try it, here's my affiliate link (https://t.co/PeOq4FNGWC)

Follow me to learn how to create an ai influencer from scratch.

Here's the prompt for the first image:

Setting: A rural countryside property in late afternoon golden hour light. A rustic dark wood fence runs behind her, with a dirt path, dry grass, scattered leaves, and mango trees on either side. Rolling hills and distant mountains are visible in the background under a clear blue sky with soft clouds.

Subject: A young woman in her mid-20s with sun-kissed tan skin, hyper-realistic visible skin texture with natural pores and micro-texture, no airbrushing. Blonde balayage hair styled in a high double top-knot bun with loose face-framing strands. Soft dewy makeup with nude-pink glossy lips and light contour. Wearing a strapless white ribbed tube top with a lettuce-edge hem, paired with a matching white ribbed mini skirt sitting low on the hips, cinched with a gold vintage-coin chain belt. Accessorized with a delicate gold chain necklace with a small red heart pendant and small gold hoop earrings. White flip-flop sandals.

Pose: Standing upright facing slightly off-camera, chin lifted, gazing toward the horizon with a relaxed, confident expression. Both hands clasped behind her back, shoulders relaxed, weight evenly balanced on both legs, casting a soft shadow on the ground in front of her.

Lighting: Warm golden-hour backlighting from the upper right, sun flare filtering through the tree leaves, soft rim light on her hair and shoulders, gentle warm fill light on her face, natural warm color temperature around 3200-3500K.

Camera: Full-body vertical shot, eye-level angle, shot on an 35mm-equivalent lens, shallow depth of field with the background softly blurred, sharp focus on the subject.

Image quality: Photoreal, sharp detail, 8K, natural skin rendering, true-to-life color grading, no plastic or waxy skin look.

Do not include: text, watermark, logos, extra people, blurry face, distorted hands, unnatural skin smoothing.

[Grow with Dani @GrowWithDani]

Same face. Same vibe. Every single time.

Full body angles, face close-ups, selfie shots, accessory details — all locked in from one prompt.

That's the power of a character sheet. No more random outputs. No more inconsistency. Just clean, repeatable results.

I build mine with Nano Banana Pro on Higgsfield — the quality is unreal.

Create manually → https://t.co/PeOq4FNGWC
Build with AI automation → https://t.co/lSHeP7jDcP
(affiliate links — I earn a small commission at no extra cost to you)

Comment "PROMPT" + follow me and I'll send you the exact prompt 👇

#AIInfluencer #NanoBananaPro #HiggsField

[Grow with Dani @GrowWithDani]

Creators are about to start seeing HIGHER VIEW COUNTS 👀

Starting August 24, YouTube will count a view from the SECOND a video starts playing — with no minimum watch time.

This will apply to:

• Long-form videos
• Shorts
• Live streams

Your original view metric isn’t disappearing either. YouTube will keep it as “Engaged views” inside Advanced Analytics, so you can still see how many people actually chose to keep watching.

Important: this DOESN’T mean higher earnings or easier YPP eligibility.

It’s mainly a new way of measuring your true exposure across YouTube — and creators will likely see their public view counts grow faster.

Big change for creators and brand deals. 👀

[TeamYouTube @TeamYouTube]

Starting 8/24, views will count the moment a video begins to play (w/ no minimum watch time) across all formats (VOD, Shorts, & Live). We’re standardizing views to reflect creators’ true exposure across YouTube, making it easier to show their value to brand partners.

This update won't impact creator earnings or YPP eligibility.

Details: https://t.co/dtFACmLJLM

[Grow with Dani @GrowWithDani]

@higgsfield_ai love it! I already tried it out and created my own ai influencer https://t.co/nEfOKrcKr3

[Grow with Dani @GrowWithDani]

my ai influencer is finally alive 🤖✨ thanks to seedance 2.5, now available in higgsfield at 1080p!

dropping the exact prompt I used below 👇

try it here🔗 https://t.co/PeOq4FNGWC

here's how I built it: started with 4 character images in gpt image 2, brought them into seedance 2.5 inside higgsfield as reference elements, generated the voice in elevenlabs, uploaded it as an mp4 reference, and once the clip came out I gave it a quick edit pass in capcut. nothing fancy, just a clean workflow.

(affiliate link, may earn a small commission at no extra cost to you 🙏)

#AIInfluencer #Seedance25 #Higgsfield

Prompt:

SCENE CONTEXT
A young female AI influencer records a selfie-style talking-head video, holding
the camera in one hand at arm's length. Across 8 hard cuts of 3 seconds each,
she cycles through 4 real-world locations/outfits, delivering a single
continuous line of dialogue that reveals she is an AI-generated character.

ACTIVE REFERENCES
@[Image 1](image_1) — woman, curly blonde balayage hair, navy backwards cap, black tank
top, seated in a parked car, soft daylight through the window. 100% matches
the reference.
@[Image 2](image_2) — same woman, red-framed glasses, pinstripe vest and blazer over a
light blue shirt and striped tie, neutral indoor wall. 100% matches the
reference.
@[Image 3](image_3) — same woman, black off-shoulder top, seated in dim movie theater
seating, warm ambient practical light. 100% matches the reference.
@[Image 4](image_4) — same woman, mint-and-white striped one-shoulder top, white jeans,
standing in a convenience store beverage aisle, red accent wall. 100% matches
the reference.
@[Video 1](video_1) — dialogue audio reference: match this exact line reading, pacing,
and vocal energy; sync lip movement and performance timing to it.

LOCATION MAP
Four distinct real-world interiors, one per active image reference — parked
car, indoor office-style wall, movie theater, convenience store aisle. Each
cut is a hard location swap, camera always held at arm's length by the
subject, no camera crew visible.

FIRST FRAME / BLOCKING
Opens on @[Image 1](image_1)'s framing: she is already mid-reaction, phone/camera held
up in one hand, free hand raised near her face, direct eye contact into the
lens, caught in a moment of surprise.

FORMAT MODE
Timed multishot — 8 HARD CUTS at 0:03, 0:06, 0:09, 0:12, 0:15, 0:18, 0:21,
ending at 0:24. Cuts only at these points, camera does not cut on its own,
no crossfades.

OPTICS
Close-up (18°) or extreme close-up (12°) every segment, framing matched to
each cited reference image's crop. No zoom, no push-in, crop size changes
only on the cut, no drift mid-segment.

CAMERA
Handheld selfie POV throughout, held by the subject's own free hand at
arm's length, natural micro-shake, no stabilizer look. Never blended with
subject motion — subject motion described separately below.

ACTION (dialogue in English US, casual California influencer delivery)
0:00–0:03 [@[Image 1](image_1), car, cap]: eyes widen, playful surprised look into lens.
"Woooh, I can't believe—"
0:03–0:06 [@[Image 1](image_1), car, cap]: quick chuckle, head tilts slightly.
"—you actually fell for it."
0:06–0:09 [@[Image 2](image_2), glasses/suit]: hard cut, deadpan direct stare, one brow
lifts. "I'm not real,"
0:09–0:12 [@[Image 2](image_2), glasses/suit]: small knowing smile. "but your brain
totally thinks I am!"
0:12–0:15 [@[Image 3](image_3), movie theater]: hard cut, relaxed lean-back pose, hand
runs through hair. "I'm an AI influencer, and I can switch up my outfit
whenever I want"
0:15–0:18 [@[Image 4](image_4), convenience store]: hard cut, hands lift into a heart
shape near face. "and be anywhere I want with just one click."
0:18–0:21 [@[Image 2](image_2), glasses/suit]: hard cut, confident direct gaze, finger
raised as if making a point. "My creator made all of this in Higgsfield with
Seedance 2.5, now available in 1080p."
0:21–0:24 [@[Image 4](image_4), convenience store]: hard cut, playful wink, points
finger down toward caption. "Wanna try it? I left the link in the
description!"

PERFORMANCE
Natural breath between lines, living eyes with real catchlights, believable
micro-expressions on every cut — genuine surprise at 0:00, playful smirk at
0:06, warm confidence by 0:18. Pore-level skin realism maintained across all
four looks.

PHYSICS
Hair and fabric react naturally to hand and head movement on every cut,
correct contact shadow where the free hand touches face/hair, phone/camera
grip stays physically consistent within each 3-second segment.

LIGHTING
Segment-matched to each reference: soft window daylight (car segments),
neutral even indoor light (office-suit segments), warm dim practical light
(theater segment), bright red-tinted store lighting (convenience store
segments). White balance shifts naturally per location, no forced grade
across cuts.

AUDIO
Dialogue (English US, audio reference @[Video 1](video_1), no subtitles): full line as
scripted above, spoken by the subject only. Light ambient room tone per
location, no music.

STYLE
8K photorealistic, no CGI or illustration look, natural smartphone-selfie
image quality, subtle handheld grain, true-to-reference skin and hair
rendering in every segment.

OUTPUT SETTINGS
9:16 vertical, 24 seconds total, real-time throughout, no slow-motion.

POSITIVE LOCKS
Identity stays 100% consistent with each cited image across its segments —
same face, same hair, same outfit within that image's cuts, no bleed between
outfits mid-segment. Hands stay anatomically correct in every gesture. Lip
sync stays locked to @[Video 1](video_1)'s performance throughout.

[Grow with Dani @GrowWithDani]

@hive_ai indeed, it's actually labeled ai-generated

[Grow with Dani @GrowWithDani]

my ai influencer is finally alive 🤖✨ thanks to seedance 2.5, now available in higgsfield at 1080p!

dropping the exact prompt I used below 👇

try it here🔗 https://t.co/PeOq4FNGWC

here's how I built it: started with 4 character images in gpt image 2, brought them into seedance 2.5 inside higgsfield as reference elements, generated the voice in elevenlabs, uploaded it as an mp4 reference, and once the clip came out I gave it a quick edit pass in capcut. nothing fancy, just a clean workflow.

(affiliate link, may earn a small commission at no extra cost to you 🙏)

#AIInfluencer #Seedance25 #Higgsfield

Prompt:

SCENE CONTEXT
A young female AI influencer records a selfie-style talking-head video, holding
the camera in one hand at arm's length. Across 8 hard cuts of 3 seconds each,
she cycles through 4 real-world locations/outfits, delivering a single
continuous line of dialogue that reveals she is an AI-generated character.

ACTIVE REFERENCES
@[Image 1](image_1) — woman, curly blonde balayage hair, navy backwards cap, black tank
top, seated in a parked car, soft daylight through the window. 100% matches
the reference.
@[Image 2](image_2) — same woman, red-framed glasses, pinstripe vest and blazer over a
light blue shirt and striped tie, neutral indoor wall. 100% matches the
reference.
@[Image 3](image_3) — same woman, black off-shoulder top, seated in dim movie theater
seating, warm ambient practical light. 100% matches the reference.
@[Image 4](image_4) — same woman, mint-and-white striped one-shoulder top, white jeans,
standing in a convenience store beverage aisle, red accent wall. 100% matches
the reference.
@[Video 1](video_1) — dialogue audio reference: match this exact line reading, pacing,
and vocal energy; sync lip movement and performance timing to it.

LOCATION MAP
Four distinct real-world interiors, one per active image reference — parked
car, indoor office-style wall, movie theater, convenience store aisle. Each
cut is a hard location swap, camera always held at arm's length by the
subject, no camera crew visible.

FIRST FRAME / BLOCKING
Opens on @[Image 1](image_1)'s framing: she is already mid-reaction, phone/camera held
up in one hand, free hand raised near her face, direct eye contact into the
lens, caught in a moment of surprise.

FORMAT MODE
Timed multishot — 8 HARD CUTS at 0:03, 0:06, 0:09, 0:12, 0:15, 0:18, 0:21,
ending at 0:24. Cuts only at these points, camera does not cut on its own,
no crossfades.

OPTICS
Close-up (18°) or extreme close-up (12°) every segment, framing matched to
each cited reference image's crop. No zoom, no push-in, crop size changes
only on the cut, no drift mid-segment.

CAMERA
Handheld selfie POV throughout, held by the subject's own free hand at
arm's length, natural micro-shake, no stabilizer look. Never blended with
subject motion — subject motion described separately below.

ACTION (dialogue in English US, casual California influencer delivery)
0:00–0:03 [@[Image 1](image_1), car, cap]: eyes widen, playful surprised look into lens.
"Woooh, I can't believe—"
0:03–0:06 [@[Image 1](image_1), car, cap]: quick chuckle, head tilts slightly.
"—you actually fell for it."
0:06–0:09 [@[Image 2](image_2), glasses/suit]: hard cut, deadpan direct stare, one brow
lifts. "I'm not real,"
0:09–0:12 [@[Image 2](image_2), glasses/suit]: small knowing smile. "but your brain
totally thinks I am!"
0:12–0:15 [@[Image 3](image_3), movie theater]: hard cut, relaxed lean-back pose, hand
runs through hair. "I'm an AI influencer, and I can switch up my outfit
whenever I want"
0:15–0:18 [@[Image 4](image_4), convenience store]: hard cut, hands lift into a heart
shape near face. "and be anywhere I want with just one click."
0:18–0:21 [@[Image 2](image_2), glasses/suit]: hard cut, confident direct gaze, finger
raised as if making a point. "My creator made all of this in Higgsfield with
Seedance 2.5, now available in 1080p."
0:21–0:24 [@[Image 4](image_4), convenience store]: hard cut, playful wink, points
finger down toward caption. "Wanna try it? I left the link in the
description!"

PERFORMANCE
Natural breath between lines, living eyes with real catchlights, believable
micro-expressions on every cut — genuine surprise at 0:00, playful smirk at
0:06, warm confidence by 0:18. Pore-level skin realism maintained across all
four looks.

PHYSICS
Hair and fabric react naturally to hand and head movement on every cut,
correct contact shadow where the free hand touches face/hair, phone/camera
grip stays physically consistent within each 3-second segment.

LIGHTING
Segment-matched to each reference: soft window daylight (car segments),
neutral even indoor light (office-suit segments), warm dim practical light
(theater segment), bright red-tinted store lighting (convenience store
segments). White balance shifts naturally per location, no forced grade
across cuts.

AUDIO
Dialogue (English US, audio reference @[Video 1](video_1), no subtitles): full line as
scripted above, spoken by the subject only. Light ambient room tone per
location, no music.

STYLE
8K photorealistic, no CGI or illustration look, natural smartphone-selfie
image quality, subtle handheld grain, true-to-reference skin and hair
rendering in every segment.

OUTPUT SETTINGS
9:16 vertical, 24 seconds total, real-time throughout, no slow-motion.

POSITIVE LOCKS
Identity stays 100% consistent with each cited image across its segments —
same face, same hair, same outfit within that image's cuts, no bleed between
outfits mid-segment. Hands stay anatomically correct in every gesture. Lip
sync stays locked to @[Video 1](video_1)'s performance throughout.

[Grow with Dani @GrowWithDani]

Six months ago AI influencers still looked "off." Robotic movement, dead eyes, mismatched lipsync.

Not anymore.

Just tested Seedance 2.5 through @apob_ai and the result genuinely stopped me mid-scroll. Full demo on my channel 👉 https://t.co/W83NQCJwr0

[Grow with Dani @GrowWithDani]

If you want to build your own AI influencer like this, Apob AI is the tool I use to bring characters like her to life 👇

🔗 https://t.co/tKQOiMU4ij

Want the full breakdown of how I made mine? Watch here: https://t.co/wbwwKiK6YT

affiliate link, I may earn a commission
@apob_ai

[Grow with Dani @GrowWithDani]

@leohuynh139 yeah! I'm looking forward to it

[Grow with Dani @GrowWithDani]

My AI influencer just came alive on screen and I'm still processing it 😅

Ran her through Seedance 2.5 inside Apob AI and the lipsync + micro-expressions finally cleared the uncanny valley. No more stiff head-bob, no more mismatched mouth movement — just natural, human-feeling motion.

Full demo + breakdown 👉 https://t.co/hHqI5Sd4Np

Tool I used to make it happen (affiliate link, may earn a small commission at no extra cost to you 🙏) 👉 https://t.co/NmuG4UPezo

#AIInfluencer #Seedance25 #ApobAI

[Grow with Dani @GrowWithDani]

Tested Apob AI's new Storyboard to Video feature — auto-generates character sheets + outfit refs and stitches multi-scene content together for 60 credits.

Genuinely one of the faster AI influencer workflows I've used. Free to try 👉 @apob_ai (affiliate link, may earn a small commission at no extra cost to you 🙏) https://t.co/WIGmqtUsxC

[Grow with Dani @GrowWithDani]

The AI influencer "consistency problem" everyone complains about is 100% solvable now.

Lock the voice first. Create reusable elements for outfit/background. Tag them into every brief.

That's it. That's the whole trick. Broke it down step by step here 👇
https://t.co/hHqI5SdCCX

@apob_ai

[Grow with Dani @GrowWithDani]

Unpopular opinion: don't start your AI influencer with the face.

Start with the niche. A hot AI character with nothing to be known for just... exists. No reason to follow.

Pick the niche → pick the personality → THEN build the face around it. Full workflow: https://t.co/hHqI5SdCCX

[Grow with Dani @GrowWithDani]

6/
Full step-by-step (from GPT Image 2 base image to final 4K video, plus the monetization talk) is in the tutorial:

https://t.co/W83NQCJwr0
@apob_ai

Code DANI15 for a discount on Apob AI 👉 https://t.co/ou97rxzag7

[Grow with Dani @GrowWithDani]

1/
I built an AI influencer with the SAME face, voice, and outfit in every single clip. No drift, no random changes.

Here's exactly how 🧵

https://t.co/W83NQCJwr0

Code DANI15 for a discount on Apob AI 👉 https://t.co/ou97rxzag7

[Grow with Dani @GrowWithDani]

5/
Pro tip: AI still messes up sometimes. Don't skip reviewing your storyboard before hitting generate — pull the weak shots, keep the strong ones.