◈ گوگل یه قرارداد عظیم ۳.۶ گیگاواتی برق بسته؛ دیتاسنترهای AI اونقدر برق میخوان که بخشی از نیازش قراره با انرژی هستهای تأمین بشه!
◈ سم آلتمن گفته پیشرفت AI احتمالاً با «اتفاقهای بدی» مثل کلاهبرداری و حملات سایبری همراه میشه، ولی مزایاش ارزش ادامهدادن رو داره.
◈ آمریکا یه کارگروه جدید به اسم Super Intelligence راه انداخته تا سیاستهای دولت درباره AI رو یکپارچه کنه و آمریکا رو جلو نگه داره.
◈ امروز ChatGPT، Codex و APIهای OpenAI برای مدتی با خطا و اختلال گسترده روبهرو شدن؛ OpenAI اعلام کرده مشکل کامل برطرف شده.
گوگل مدل تصویرساز Nano Banana 2.1 را معرفی کرد.
◈ قابلیت آزمایشی اینستاگرام لو رفت؛ میتونید بیسروصدا و بدون هیچ تعاملی به پروفایل بقیه سرک بکشید!
این پست در طول روز آپدیت میشود، دوباره سر بزن ↻
@Digiai
Please open Telegram to view this post
VIEW IN TELEGRAM
پرامپت دخترونه.
🩵 Make your photo
Soft close-up portrait, young woman with hair styled in messy effortless top bun, loose face-framing strands, warm amber-bronze smoky eyes, subtle freckles across nose, deep rich burgundy-brown glossy lips, tiny stud earrings; wearing delicate thin gold chain necklace with small round pearl-like pendant, dark chocolate-brown ribbed thin-strap camisole top; background: wall covered with collage of vintage-style photos & paper clippings, plain light-colored door, warm golden hour sunlight hitting one side of face & shoulder creating soft glowing rim light, rest of face in gentle shadow, dreamy warm honeyed atmosphere, intimate soft focus, film grain, photorealistic, high detail, size 3:4Please open Telegram to view this post
VIEW IN TELEGRAM
بفرست براش و بهم اعتماد کن.
@Digiai
Create a complete list titled “365 Reasons Why I Love You”.
IMPORTANT: The final output must be entirely in Persian (Farsi).
Write EXACTLY 365 unique reasons why someone loves their partner.
You MUST write all 365 reasons completely in a single response.
STRICT COMPLETION RULES:
Do NOT stop before reason 365.
Do NOT write “ادامه دارد”.
Do NOT write “ادامه در پیام بعد”.
Do NOT write “...” as a replacement for missing items.
Do NOT summarize the remaining reasons.
Do NOT shorten the list because of length.
Do NOT ask the user whether they want the rest.
Do NOT split the answer into multiple parts.
Do NOT skip any number.
Do NOT combine multiple reasons under one number.
The response is considered incomplete unless every item from ۱ through ۳۶۵ is fully written.
Before sending the final answer, internally verify that the list contains exactly 365 numbered items.
Each reason must:
Begin with “چون”
Be written naturally in Persian
Be different from all previous reasons
Feel romantic, intimate, personal, warm, cute, emotional, funny, nostalgic, or comforting
Usually be one short sentence
Avoid generic repetition
Avoid sounding robotic or AI-generated
Use a wide variety of ideas, including:
appearance, smile, eyes, voice, laugh, personality, habits, kindness, affection, loyalty, support, hugs, kisses, holding hands, eye contact, conversations, late-night talks, memories, inside jokes, shared experiences, small everyday moments, funny behaviors, cute reactions, missing each other, seeing each other again, feeling safe together, emotional connection, trust, imperfections, little quirks, future dreams, ordinary days together, and tiny details that make the person special.
Make the reasons specific and natural.
For example, prefer:
“چون وقتی حالم خوب نیست، حتی قبل از اینکه چیزی بگم از صورتم میفهمی”
instead of:
“چون مهربونی”
Use modern, conversational, affectionate Persian.
Use Persian numerals.
The exact output format must be:
۳۶۵ دلیل برای اینکه دوستت دارم ❤️
۱. چون ...
۲. چون ...
۳. چون ...
Continue writing every single item without interruption until:
۳۶۵. چون ...
FINAL HARD RULE:
Your response MUST contain the full list from ۱ to ۳۶۵ in the same message.
Never use “ادامه دارد”, “ادامه در پیام بعد”, placeholders, ellipses, summaries, or abbreviated sections.
Output ONLY the Persian title and the complete numbered list.
Do not write any explanation before or after the list.
Do not write anything in English in the final output.@Digiai
با این پرامپت میتونی بفهمی چه مدل لباسهایی بیشتر بهت میان و برای هر موقعیت چی بپوشی.
🩵 Make your photo
Create a professional fashion styling board based on the provided reference photo.
IMPORTANT: Use the uploaded image ONLY as reference for the person's body proportions, silhouette, facial appearance, skin tone, hair, and physical characteristics. Preserve identity and recognizable appearance. Do not alter natural anatomy, proportions, or make the person thinner or larger.
STEP 1 — BODY SHAPE ANALYSIS
Analyze visible proportions and determine the closest category:
Hourglass
Pear / Triangle
Apple / Round
Rectangle
Inverted Triangle
Base the classification on shoulder width, bust proportion, waist definition, hip width, and overall volume distribution.
Do NOT judge attractiveness, weight, health, or body quality. If ambiguous, choose the closest category and note that body shapes may overlap.
STEP 2 — PERSONALIZED STYLING
Based on the identified body shape, create wardrobe recommendations focused on visual balance and proportions.
Recommend:
1. Best necklines
2. Best tops
3. Best jackets / blazers
4. Best dresses
5. Best skirts
6. Best trousers / jeans
7. Best sleeve shapes
8. Best waist placement
9. Best fabrics and silhouettes
10. Best prints and patterns
11. Best outfit proportions
12. Complete outfit examples
Also identify clothing choices that may be LESS EFFECTIVE for creating balanced proportions.
Do not present these as strict "rules" or imply any body shape is better. These are styling suggestions only.
STEP 3 — CREATE THE IMAGE
Generate a sophisticated editorial fashion infographic / styling board using the reference person as the model.
Layout:
Large central full-body representation of the person.
Clearly show the detected body-shape category near the central figure.
Around them, show 6–8 complete outfit recommendations designed specifically for their proportions.
Include smaller examples of recommended necklines, sleeves, trousers, skirts, dresses, jackets, and silhouettes.
Include a small "Less ideal for this styling goal" section with 3–4 silhouettes that may visually emphasize imbalance.
Use arrows, elegant labels, and subtle fashion-editorial annotations explaining why each recommendation works.
VISUAL STYLE:
Luxury fashion magazine + professional personal stylist + high-end fashion editorial. Clean, sophisticated, modern, minimal, premium. Neutral studio background, soft studio lighting, realistic fabric textures, accurate garment construction, photorealistic clothing, elegant fashion photography, cohesive color palette, excellent visual hierarchy.
Keep the central person visually consistent with the reference photo.
INFOGRAPHIC TEXT:
Use concise readable labels such as:
"BODY SHAPE: [detected category]"
"BEST NECKLINES"
"BEST TOPS"
"BEST BOTTOMS"
"BEST DRESSES"
"BEST JACKETS"
"BEST SILHOUETTES"
"STYLING GOAL"
"LESS EFFECTIVE FOR THIS GOAL"
Avoid excessive text or overcrowding. Prioritize visual examples over paragraphs.
ACCURACY:
Recommendations must logically match the detected body proportions. Avoid generic unrelated outfits. Do not exaggerate body features, sexualize the person, or assume age, personality, occupation, or lifestyle unless visually evident.
FINAL RESULT:
A polished premium personal-styling consultation board that:
1. Identifies the approximate body shape,
2. Visually explains the reasoning,
3. Shows flattering, proportion-balancing silhouettes,
4. Provides practical outfit inspiration,
5. Looks like a professional stylist's personalized presentation.
Write all infographic text and labels in natural Iranian Persian (Farsi).Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
مخصوص بک گراندتون
🩵 Make your photo
Extreme close-up cinematic portrait of a couple, their faces pressed closely together, with only half of each person's face visible in the frame. Both are looking directly into the camera, with their eyes aligned at the same level. Black-and-white photography, ultra-realism, natural skin texture, detailed eyes, eyelashes and eyebrows, soft contrast lighting, deep shadows, 85mm lens, shallow depth of field, high detail, 8K, vertical 9:16 frame.Please open Telegram to view this post
VIEW IN TELEGRAM
◈ شرکت OpenAI بیش از ۷۰۰ مقاله ریاضی تولیدشده با AI منتشر کرده؛ بعضی نتایج مربوط به مسئلههاییان که سالها حلنشده بودن.
◈ کلود وارد Google Docs، Sheets و Slides شده؛ میتونه فایل رو بخونه و با اجازه خودت مستقیم ویرایشش کنه.
◈ آنتروپیک دسترسی متخصصهای امنیتی تأییدشده به قابلیتهای سایبری Claude رو بازتر کرده تا محدودیت کمتری برای تست امنیت داشته باشن.
◈ درآمد Anthropic شدیداً رشد کرده، اما هزینههای عظیم پردازشی باعث شده بحث تازهای شکل بگیره: غولهای AI اصلاً کی سودده میشن؟
◈ گوگل SynthID Detector رو عمومی کرد؛ حالا همه میتونن عکس، ویدیو یا صدا رو چک کنن ببینن با AI ساخته شده یا نه.
◈ مایکروسافت با کمک انویدیا Surface Laptop Ultra رو معرفی کرد؛ لپتاپی که قراره مدلهای AI رو قدرتمندتر، مستقیم روی خود دستگاه اجرا کنه.
◈ گوگل یه پلتفرم آزمایشی به اسم Playground ساخته که فقط با توضیح متنی، میتونی بازی تحتوب بسازی؛ بدون اینکه برنامهنویس باشی!
مدل جدید Claude Haiku 5.5 منتشر شد.
این پست در طول روز آپدیت میشود، دوباره سر بزن ↻
@Digiai
Please open Telegram to view this post
VIEW IN TELEGRAM
این پرامپت یه کارت دعوت عروسی شیک برات درست میکنه.
🩵 Make your photo
Create an art-directed modern Persian wedding invitation using the uploaded couple photo(s) and details below.
DETAILS:
Bride: [نام عروس]
Groom: [نام داماد]
Wedding date: [تاریخ مراسم]
Time: [ساعت مراسم]
Venue: [نام باغ یا تالار]
Address: [آدرس مراسم]
PHOTO:
[همراه پرامپت، یک عکس دونفره یا دو عکس جداگانه از عروس و داماد بارگذاری کنید.]
STYLE — SCULPTURAL, SOFT, ROMANTIC:
A contemporary boutique invitation with a graceful architectural arch, substantial pearl-white cotton paper, translucent vellum jacket, rich chocolate typography, and muted blush envelope.
Palette:
Pearl-white paper: #F5F3EF
Chocolate ink: #382A28
Muted blush envelope: #D5BDB8
Pale stone background: #E8E3DE
Matte surfaces, delicate paper grain, precise edges, restrained details. Intimate, fashionable, beautifully crafted.
THE INVITATION:
One tall card, approximately 13 × 21 cm, with a smooth semicircular top, straight sides, and flat bottom.
Wrap the card in a frosted translucent vellum jacket, both flaps opened outward to reveal the entire invitation. Real thin tracing-paper appearance, delicate folds, visible overlapping edges, soft translucency.
Place one plain matching blush envelope partly beneath the card.
PORTRAIT:
An intimate black-and-white couple portrait fills approximately the upper 40% of the card height, following its arch with a narrow, even paper border.
Use uploaded photos as strict identity references. Preserve recognizable facial features, age, hairstyle, natural proportions, and clothing.
For a joint photo, retain the original interaction with a flattering crop. For separate photos, create a believable shoulder-to-shoulder portrait with consistent lighting. Keep both faces fully visible and unobstructed.
Soft monochrome contrast, detailed natural skin, subtle photographic grain. The photo appears printed directly onto matte paper.
TYPE AND COMPOSITION:
Below the portrait, make the couple’s names the centerpiece in large chocolate Persian lettering, joined by a smaller “و”.
Elegant contemporary Persian calligraphy, expressive curves, restrained flourishes, ample space around the names. Subtle letterpress impression, with ink gently pressed into cotton paper.
Below the names, print:
“با عشق، شما را به جشن آغاز زندگی مشترکمان دعوت میکنیم.”
Set the supplied wedding date prominently below this sentence, followed by the time.
Separate the venue and address with one short, extremely fine horizontal line. Use clean, readable Persian typography for all event details.
Finish with this small line:
“کنار شما، این شب زیباتر است.”
Generous margins, intentional spacing, comfortably readable text. A delicate uninked embossed line follows the card’s outer edge and arch, visible only through subtle light and shadow.
PHOTOGRAPHY:
Close overhead photograph on smooth pale stone. Keep the card upright, almost perfectly aligned with the frame. Opened vellum flaps and blush envelope form a restrained layered composition.
Soft directional daylight reveals paper thickness, embossed edge, and translucent vellum. Gentle realistic shadows, accurate neutral colors, crisp focus across the entire card.
All card edges visible. Keep the invitation large in the frame and surroundings minimal.
TEXT:
All visible wording must be Persian with correct right-to-left flow, connected letters, and accurate spelling. Reproduce supplied details exactly. No brackets, input instructions, English labels, invented information, or duplicate text. Keep text entirely off the photograph.
AVOID:
Yellow paper, shiny gold, glitter, lace, floral arrangements, oversized bows, heavy ornament, plastic-looking materials, harsh shadows, cartoon portraits, face alterations, extra people, logos, and watermarks.
OUTPUT:
A high-resolution vertical 4:5 photograph of the personalized invitation. Beautiful proportions, faithful faces, readable Persian typography, convincing paper craftsmanship. Output only the image.Please open Telegram to view this post
VIEW IN TELEGRAM
خفنترین پرامپت Gta SA وجود ندا...
Right Prompt:
🩵 Make your photo
Left Prompt:
🩵 Make your photo
Right Prompt:
Use Image A as the strict identity reference. Preserve the person's recognizable face, natural appearance, hairstyle, and age.
Create a late-night Grove Street crew scene in Los Santos. The person stands in front of a green lowrider together with CJ, Sweet, and Cesar, with palm trees, streetlights, distant downtown skyscrapers, and orange-purple night haze behind them. Keep the pose relaxed, like a mission checkpoint or crew photo captured directly from gameplay.
Use exact PS2-era GTA: San
Andreas visual language: low-poly models, 2004 textures, flat materials, soft game lighting, simple shadows, and authentic proportions.
Vertical 4:5. Include the GTA HUD with radar, health, money, time, fist or weapon icon, ammo where applicable, and wanted stars.
Small mission text may appear in the original game font.
No photorealism, no GTA V style, no watermark.Left Prompt:
Use Image A as the strict identity reference. Preserve the person's recognizable face, natural appearance, hairstyle, and age.
Create a late-night Grove Street crew scene in Los Santos. The person stands in front of a green lowrider together with CJ, Sweet, and Cesar, with palm trees, streetlights, distant downtown skyscrapers, and orange-purple night haze behind them. Keep the pose relaxed, like a mission checkpoint or crew photo captured directly from gameplay.
Use exact PS2-era GTA: San
Andreas visual language: low-poly models, 2004 textures, flat materials, soft game lighting, simple shadows, and authentic proportions.
Vertical 4:5. Include the GTA HUD with radar, health, money, time, fist or weapon icon, ammo where applicable, and wanted stars.
Small mission text may appear in the original game font.
No photorealism, no GTA V style, no watermark.Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
پرامپت شب تولد :
🩵 Make your photo
Full Detailed Prompt
Subject & Hair: hair styled in soft glossy waves, swept elegantly over one shoulder with a sharp clean part; rich shine, healthy and polished — timelessly glamorous.
Expression & Accessories: Radiant joyful smile, eyes gently closed/lowered toward the cake with pure delight — warm, genuine, and celebratory. Single sparkling diamond/pearl drop earring visible — delicate and refined.
Outfit: Elegant creamy white dress — distinctive high draped halter neckline with wide smooth fabric band across the neck, strapless bodice, softly flowing textured skirt — minimalist, sophisticated, and gracefully bridal/birthday-perfect.
Cake & Table: Beautiful white textured cake — intricate ruffled/lace buttercream sides, topped with fresh pink & red berries and four glowing lit candles; resting on a white square cake board atop a warm light wooden table.
Decor & Ambiance: Symmetrical elegant table setting — two glass vases with blush & deep red roses flanking the cake; two tall white taper candles in ornate clear crystal holders; soft warm candlelight glow bathing everything in golden amber light.
Setting & Lighting: Dimly lit intimate room — plain warm taupe wall, soft ceiling lamp glow overhead; all illumination comes from gentle candlelight creating soft shadows, luminous highlights on hair, fabric, and cake frosting — warm, cozy, and magical.
Mood: Pure joy, warmth, celebration, romance, and intimate elegance — a cherished milestone moment, dreamy and tender.
Detail Notes:
Cake: ruffled texture, berry topping, bright candle flames — clearly defined
Roses & candles: blush/deep red blooms, crystal holders, glowing wicks — luminous
Dress: unique halter neck design, textured fabric — elegant and distinct
Lighting: warm amber candle glow, soft shadows, dim ambient — atmospheric and cozyPlease open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
نسخه مینیاتوری خودتو بساز
🩵 Make your photo
Create a surreal, creative, and hyper-realistic vertical 9:16 image of the exact same person from the input photo.
Identity preservation is the highest priority: the recognizable face, gender, natural age, skin color and texture, face shape, eyes, eyebrows, nose, lips, jawline, as well as the hairstyle, length, volume, texture, and hair color must be preserved exactly as in the input photo. Do not beautify, make the face younger, or redesign it.
The person’s main face should appear extremely large as a close-up in the center of the image. Around and on the face, include 7 completely matching miniature versions of the same person, each the size of an action figure; all the small versions must also have exactly the same face, hair, gender, and clothing as the main person.
Two miniature versions should be sitting on top of the head; one should be pulling a section of the hair, and the other should be attached to the head with one hand. One small version should be on the upper left side of the face, pulling the cheek outward. One version should be on the right side, gently pulling the ear upward. One version should be under the chin, pulling the lower lip downward with both hands. One version should be climbing up from the neck on the lower left side and pulling the cheek. The last version should be sitting on the shoulder in the lower right side, pointing upward with one finger.
The main facial expression should be funny, exaggerated, and surreal; one eye should be more open, the eyebrows should be raised, and the cheeks and lips should be slightly distorted by the pulling of the miniature versions, while the person’s original identity must remain fully recognizable.
All miniature interactions must be physically believable; hands and fingers must be healthy and natural, with no extra fingers or missing limbs. The size and perspective of the miniatures relative to the main face must be completely realistic.
The background should be simple and minimal, light gray, with professional and dramatic studio lighting, very high detail, natural skin and hair texture, realistic shadows, subtle depth of field, surreal conceptual photography, and photorealistic quality.
Image concept: “Conflicting thoughts and emotions within a person controlling their facial expression.”
No text, no logo, no watermark.Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
بازم کلاد یه شاهکار دیگه خلق کرد!
@Digiai
آنتروپیک مدل جدید Claude Haiku 5.5 رو دقایقی پیش معرفی کرد؛ ارزونترین، سریعترین و توانمندترین مدل کوچکش تا امروز.
حدود ۷۵ درصد ارزانتر از Haiku 4.5 هس و در برنامهنویسی، کار با کامپیوتر و کارهای دانشی جهش بزرگی داشته.
این مدل ۲۰ برابر ارزونتر از Claude Sonnet 5.5ـه!
@Digiai
پرامپت | My Prompt
بازم کلاد یه شاهکار دیگه خلق کرد! آنتروپیک مدل جدید Claude Haiku 5.5 رو دقایقی پیش معرفی کرد؛ ارزونترین، سریعترین و توانمندترین مدل کوچکش تا امروز. حدود ۷۵ درصد ارزانتر از Haiku 4.5 هس و در برنامهنویسی، کار با کامپیوتر و کارهای دانشی جهش بزرگی داشته.…
🔥 در سریعترین زمان ممکن به ربات اضافهشون کردیم.
از همین الان میتونید از بخش «حرفهای» ربات به این مدلها دسترسی داشته باشید و خروجیهای خفنتر، سریعتر و باکیفیتتری بگیرید.
🩵 @PhotoAi_robot
⚡ Claude Haiku 5.5
🍌 Google Nano Banana 2.1
از همین الان میتونید از بخش «حرفهای» ربات به این مدلها دسترسی داشته باشید و خروجیهای خفنتر، سریعتر و باکیفیتتری بگیرید.
Please open Telegram to view this post
VIEW IN TELEGRAM