Ari Carrington

My heart melts every time i see someone with down syndrome. They look so gentle and kind. Pure happiness everytime they smile. This beautiful photo essay is titled “Here I am” by Hilary Gauld in celebration of #worlddownsyndromeday this past 21 March. The portrait series aims to challenge the misconception that adults with Down syndrome do not live into old age. "Here I Am" features adults and seniors alongside their childish selves, celebrating the progress of the Down syndrome community and highlighting the healthcare system's growing dependence on unpaid caregivers. As they age, they are at greater risk for certain health conditions that require additional care, such as Alzheimer's disease. It is crucial to ensure that individuals with Down syndrome and their carers have access to comprehensive and unbiased information and government support.

@Dheepanratnam @Just_sharon7 You’ve got something here amazing! You’re the type of generous creator i like to follow. Its hard to find creators who learn and share their process. So thank you for that

@ari757 @Just_sharon7 I’m testing ways to bypass the celebrity filters (especially likeness blocks on faces like Sydney Sweeney etc.) by using non-photorealistic inputs

AI is wild Chatgpt 2.0 images + Seedance 2.0 tried to generate a red carpet AI video using this custom Sydney Sweeney character sheet. the result looks exactly like "we have sydney sweeney at home" 😭 A Simple Prompt used to test - Use the provided character reference for the facial features and body structure. create a 15 second red carpet multi shot walk with elegant designer wear

@Dheepanratnam Can you share the prompt for the storyboard?

@Dheepanratnam @Just_sharon7 What theory are you testing? If you dont mind me asking?

@Just_sharon7 I am testing a theroy 😋 its always a great model

@h_khachatrian @krea_ai Can Krea 2 use ref image? For character consistency?

Krea 2 new model is insane! @krea_ai

@pixelbyus Is this using chatgpt? Maybe ive been staring at ai too long. But i can almost see the mesh watermark on the image. Its like a fine net

Sydney Sweeney in 8K Ultra Realistic Close Up { "prompt": { "style": "ultra-realistic beauty close-up portrait photography, editorial skincare aesthetic, natural skin texture emphasis", "aspect_ratio": "4:5", "quality": "8k photorealistic, hyper detailed skin pores, realistic lip texture, crisp eye detail, cinematic beauty lighting", "camera": { "body": "Sony A7R IV", "lens": "105mm macro portrait lens", "settings": "f/4.0, ISO 100, ultra sharp focus" }, "lighting": { "type": "soft diffused natural beauty lighting", "tone": "warm neutral tones", "effect": "subtle highlights on lips and skin texture, gentle shadows, realistic complexion" }, "subject": { "gender": "female", "age": "young adult", "appearance": { "hair": "light brown-blonde hair with soft natural waves framing the face", "eyes": "icy blue eyes with detailed iris texture and long natural lashes", "skin": "fair skin with visible pores, realistic texture, subtle freckles and natural imperfections", "makeup": "minimal makeup, soft pink nude tones, mascara, natural brows", "lips": "full glossy pink lips with reflective transparent lip gloss, wet shine finish, ultra detailed lip texture" }, "expression": "soft pout expression looking directly into camera", "pose": "extreme close-up centered facial portrait" }, "background": { "type": "soft blurred neutral background", "depth_of_field": "shallow cinematic bokeh" }, "rendering": { "detail": "high micro skin detail, realistic peach fuzz, authentic beauty photography realism", "color_grading": "soft warm editorial tones", "sharpness": "ultra crisp facial detail" }, "negative_prompt": [ "plastic skin", "over smoothing", "cartoon effect", "AI artifacts", "extra facial features", "distorted lips", "unrealistic skin", "heavy makeup", "blurred eyes", "oversaturated colors" ] } }

@georgesttock Ugc

The UGC "Cheat Code" for 2026 MakeUGC is the secret weapon brands use to scale ads without a single camera. The 30-Minute Workflow: Step 1: Create your "Perfect Creator" using GPT Images 2.0 Step 2: Animate them inside MakeUGC using Seedance 2.0. You get 4K quality, native lip-sync, and "handheld" physics that look 100% real. Step 3: Add "iPhone grain" in CapCut for that final authentic touch. Why it wins: Same face, same product, 100 different hooks in one afternoon. Want the full Guide Comment “UGC” and I’ll DM them to you!

@ayzalnooor24521 How come there is a gemini watermark. Lol

GPT Image 2 Prompt :- Asymmetrical cinematic film-strip style collage featuring a young woman using the exact face from the reference image, preserving identity and natural skin texture. The composition is arranged like a rolling timeline of memories captured in overlapping film frames and negatives. Each frame shows different emotional micro-moments: laughing while walking, looking out a window, fixing hair in mirror, holding coffee, and soft candid smile.

@azed_ai Can it use reference image / ref face?

Krea 2 is here and it's awesome 😍

@bmx_ai13 I think you did a terrific job. this is not on you but im starting to recognize the chatgpt watermark now in all their generation - its the super subtle pattern in the grain of the digital imaging. Dont know how to explain it really lol. I think nb2/nbp hides theirs better

Early 1990s Flash Camera Portrait GPT image 2 on ChatGPT Prompt Template. Use the uploaded image as the main reference. Transform the uploaded photo into a realistic candid portrait with an early 1990s digital camera aesthetic. Preserve the subject’s identity, facial features, pose, outfit, and overall composition, but restyle the image with harsh blown-out flash highlights, subtle red-eye effect, low-resolution image quality, raw snapshot imperfections, nostalgic flash-filter styling, and a vintage timestamp look. The final image should feel candid, imperfect, and authentic, like an old retro party or personal snapshot taken with an early consumer digital camera. Keep the background dark or naturally subdued when appropriate, maintain a direct-flash look, and give the image a raw, unpolished, memory-like atmosphere. Include: - harsh direct camera flash - overexposed or blown-out highlights - subtle red-eye effect - low-resolution / soft digital detail - slight grain or noisy texture - authentic retro snapshot feeling - vintage date/timestamp aesthetic in one corner - candid, natural, imperfect energy Avoid: - cartoon or anime style - overly polished beauty retouching - studio lighting - ultra-sharp modern DSLR look - glossy AI skin - text, logos, watermarks, or graphic overlays other than the timestamp aesthetic - distorted anatomy or altered identity Make the aspect ratio 3:4

@ai_gezgini im falling in love with all your work!

StoryBoard & Prompt: 👇 Create a seamless cinematic one-take video using two reference images. REFERENCE USAGE: Image 1 = shot-plan and camera-language reference only. Use it only for the action flow, angle progression, and cinematic structure. Do NOT recreate the storyboard grid, borders, captions, text, or sketch style. Image 2 = woman identity reference. Use it for strict identity match: same face, eyes, skin tone, hair, body proportions, and overall appearance. The woman in the final video must clearly match Image 2. CORE SCENE: A dramatic deep valley / canyon with two rocky cliffs. A single thick bamboo pole is stretched across the gap between the cliffs. The woman is balancing on the bamboo in a perfect handstand. The scene must feel tense, elegant, athletic, and visually thrilling. AUDIO / MUSIC: Start with tension-building cinematic music: low pulsing bass, subtle rising strings, distant wind, and a suspenseful heartbeat-like rhythm. As the camera moves closer and the woman shows control, the music gradually becomes more confident and powerful. During the orbit and aerial flip, add a bold, stylish, adrenaline-driven beat that highlights her self-confidence. The final standing pose should land with an empowering musical hit, then soften into a confident, victorious tone. No dialogue.

Seedance 2.0 ile sabah sporu🔥 İki uçurumun arasında tek bir bambu… Geniş açıyla başlayan gerilim, kameranın hızla yaklaşmasıyla iyice yükseliyor. Kadın ellerinin üzerinde dengede. Önce kontrol. Sonra özgüven. Ters şınav, under-bar kamera, orbit hareketi, havada takla ve finalde ayakta gülümseyen güçlü bir karakter. Sabah motivasyonu bazen alarm sesiyle değil, “Ben yaparım.” hissiyle başlar. 🎬 StoryBoard & Prompt: 👇

@ai_gezgini Hi there. I was wondering if you have the link / prompt to create the character ref sheet for the lady?

Prompt: Create a 15-second cinematic action video. REFERENCE USAGE: Image 1 = strict reference for the female protagonist. Use her face, blue eyes, long straight black hair, pale skin, slim body type, and overall look. She must remain a realistic live-action woman. Image 2 = strict reference for the male antagonist. Use his exact anime identity: messy black hair, sharp blue eyes, youthful face, slim athletic build, dark navy T-shirt, black shorts, sneakers, necklace, and handgun. IMPORTANT STYLE RULE: The male character from Image 2 must stay fully anime during the entire video. He must NOT become realistic, photoreal, live-action, or humanized. Keep him clearly anime in every shot: anime face, anime eyes, anime hair, anime proportions, cel-shaded / anime rendering. The female character must remain realistic live-action. Keep the contrast intentional: realistic woman vs anime man. Do not copy the reference layouts, panels, borders, labels, or text. Use them only for identity, outfit, and weapon consistency. SCENE: The chase happens on Istanbul streets at night. It must clearly feel like Istanbul, not a generic city. Use narrow sloped streets, older apartment facades, Turkish storefronts, wet asphalt, parked cars, yellow taxis, street lamps, road reflections, and dense Beyoğlu / Karaköy style atmosphere. Cinematic night lighting, handheld energy, shallow depth of field, realistic motion blur. 0:00–0:03 Wide establishing shot of an Istanbul street at night. The realistic woman runs desperately through the street, weaving between parked cars and moving traffic. Behind her, the anime man chases her at high speed while firing a handgun. Muzzle flashes light the wet street. Bullets streak past her. Dynamic shaky tracking shot. 0:03–0:04 She gets hit in the shoulder, stumbles forward, and clutches the wound while trying to keep running. The male antagonist keeps advancing. He remains fully anime. 0:04–0:07 She crashes onto the wet asphalt and rolls. Shocked and exhausted, she looks up as the anime man approaches and tries to fire again, but the gun is empty. Only clicking sounds. He throws the handgun aside. She uses the chance to push herself back up. 0:07–0:12 Close-quarters combat begins. Fast punches, dodges, blocks, elbows, and kicks. Include a low-angle near-miss spinning kick, a close-up blocked strike, and a quick counterattack from the woman. The fight is aggressive and cinematic. The woman stays realistic; the man stays anime. 0:12–0:13 The anime man lands a powerful spinning back kick to her chest. Show the impact in slow motion. Water droplets and dust spray from the wet street. 0:13–0:15 She is thrown backward into a parked car, smashing the windshield. Glass scatters. She collapses against the vehicle, defeated. The anime man stands victorious in the middle of the Istanbul street in a strong combat stance as the camera slowly pushes in. Final shot: he is still fully anime. STYLE: Cinematic action, Istanbul street atmosphere, wet reflective asphalt, high-contrast night lighting, realistic impact physics, handheld camera movement, no gore. NEGATIVE: Do not turn Image 2 into live-action or photoreal. Do not use a generic city. No text, subtitles, logos, watermark, gore, dismemberment, extra fighters, duplicated characters, distorted anatomy, or extra weapons after the gun is thrown away.

Seedance 2.0 ile Real + Anime aksiyon denemesi 🎬 İstanbul sokaklarında geçen 15 saniyelik sinematik bir kovalamaca sahnesi. Gerçekçi dünya içinde: → ultra gerçekçi kadın karakter → anime erkek karakter → gece İstanbul sokakları → silahlı takip → yakın dövüş → sinematik final darbesi En önemli nokta şu: Kadın ve çevre tamamen gerçekçi kalırken, erkek karakter video boyunca anime stilini kaybetmiyor. Seedance 2.0 bu mixed-style aksiyon nasıl sonuç veriyor. 🎬

@mxvdxn Sorry same question. Hope you can just be direct. Hows the platform with uploading a direct face ref? Not talking about small faces…. Just a direct face ref

You might remember this from a meme, but it’s actually a typical day in Southeast Asia 😆 Never leave this region • Seedance 2.0 Omni Reference via @Hailuo_AI

@spwfeijen MAKEUGC

GAME OVER for manual UGC production. 🚀 99% of teams pay $15K+ for "authentic" videos that fatigue in 30 days. Stop begging teams for shaky iPhone footage. Claude x MakeUGC automates it all: - Claude scripts winning hooks - MakeUGC generates hyper-realistic video instantly Go from idea to live ad in <120 seconds. Want the full automation playbook? Comment "MAKEUGC" below! (must be following for DM)

@dom_lucre Are you seeing the community note you got? Did you literally do something illegal here?

🔥🚨RECENT: This mom is going viral after celebrating Mother’s Day and sharing this thoughtful message with her two sons. Mother: “Being the mom to these two boys is one of the greatest joys of my life. Your love and respect for each other, your playful spirits, quick wit, endless energy, and kind hearts make me so proud every single day. What I treasure most is how deeply you value family and the way you show up for one another. You both bring so much life, laughter, strength, and love into our lives. You complete our family in the most beautiful way, and being your mom is a gift I never take for granted.”

@mikefutia Ugc

Seedance 2.0 + Claude Code is f*cking insane 🤯 I built a Claude skill that creates UGC ads on demand. One product + one prompt = the AI creator, the script, the scene-by-scene shot list, and the finished video. All inside Claude Code. Perfect for DTC brands and agencies who can't afford to keep paying $500-$1,500 per UGC video and waiting 2 weeks for revisions. This skill eliminates the entire loop: → Tell Claude the product, ad angle, and length → Skill writes the GPT Image 2.0 prompt to generate the AI creator from scratch → Skill writes every scene prompt, dialogue line, and delivery direction → Pipes it into Seedance 2.0 with character + product + voice locked → Speed up + caption in CapCut → Ship the ad in 20 minutes No more paying $11 per video on Arcads. No more 2-week revision cycles. No more PR boxes to creators who ghost you. What you get: → Perfect character consistency across every scene → Voice consistency that holds clip-to-clip → Real product fidelity using your actual product photo as a reference → Multi-scene day-in-life, testimonial, and action-shot formats out of the box Built 100% with a Claude skill + Seedance 2.0. I recorded a full step-by-step tutorial showing the exact workflow so you can build these AI UGC ads yourself. Want the full breakdown? > Like this post > Comment "UGC" And I'll send it over (must be following so I can DM)

@ImagineArt_X Nahh… the bottleneck is just money🤣🤣🤣💸💸💸

@OurLifeWithX @omo_ollanimi @grok They went to college / uni. Parents grief when a kids grownup and leave home for furthee studies or work

Hey @grok please explain this video 🥹.

@ariaxawan Its really really beautiful. But its not ultra-realistic though. Looks more like beautiful painting. Still…. Stunning!

GPT Image 2 on Chatgpt Prompt: Ultra-realistic high-fashion editorial portrait of a beautiful young Asian woman standing confidently in a monochromatic white urban environment, wearing modern luxury white streetwear fashion. Oversized white bomber jacket, fitted white crop top, layered white cargo pants with metallic chains, white tactical sneakers, silver jewelry, subtle glossy makeup, soft matte skin texture, long voluminous ash-brown wavy hair with wispy bangs flowing naturally in the wind. A massive realistic albino python emerges behind and around her body, partially concealed within thick swirling white smoke and atmospheric fog. The snake appears lifelike with detailed scales, realistic muscle tension, natural reflections, visible texture, sharp eyes, and subtle movement. The environment is almost entirely white — white architecture, white smoke, white clothing, white textures, white lighting — creating a clean monochromatic aesthetic. Large distressed black “1989” typography dominates the background in bold industrial font, the only dark element in the entire composition. Strong contrast between the black numbers and the all-white scene. Shot like a luxury fashion campaign, realistic photography, cinematic urban atmosphere, natural skin texture, highly detailed eyes, realistic hair strands, authentic fabric folds, realistic smoke physics, dramatic directional lighting, shallow depth of field, crisp focus, realistic proportions, modern editorial photography, Vogue magazine aesthetic, street luxury styling, high dynamic range, photorealistic, ultra detailed, 85mm lens, soft cinematic shadows, studio-quality color grading, real-world textures, professional fashion photography, no fantasy elements, no CGI appearance, no cartoon style, no anime aesthetic.

@mynewshub Take down this post. Its a she lah! She refused help. You tak buat research ke. Aiyooo. Benda ni dah kisah lama okay.

Mohon bantuan semua 🙏 Beliau sudah lama menjual tisu di hadapan Pavilion. ​ Koh Lee Peng merupakan bekas atlet para renang Malaysia yang pernah mengharumkan nama negara di persada antarabangsa. Beliau telah mewakili negara dalam pelbagai kejohanan serantau, termasuk ASEAN Para Games, di mana beliau berjaya meraih banyak pingat emas dan perak. ​ Setelah bersara daripada arena sukan, beliau kini menghadapi cabaran hidup dan tekanan ekonomi. Buat masa ini, beliau menyara kehidupan dengan menjual tisu di sekitar kawasan Bukit Bintang. Walaupun mempunyai kekurangan fizikal, beliau tetap memilih untuk berdikari dan berusaha sendiri demi kelangsungan hidup. Bahan : Threads sharon_choon

@ai_artworkgen Well, i know the new MJ model is great with details and artistic flare. But what i REALLY wanna know is how they can handle human face reference to create consistent character.

DEE.TAILS. INSANE. // Midjourney