◈ گوگل و OpenAI با هم شریک شدن! مدل‌های GPT-6 قراره از طریق Google Cloud در اختیار شرکت‌ها قرار بگیرن؛ یعنی رقبای ChatGPT حالا شریک تجاری هم هستن!

◈ کلاد Duet معرفی شد! دستیار جدید Anthropic می‌تونه کارهات رو پیگیری کنه، گزارش روزانه بده و با تأییدت وظایف مختلف رو انجام بده.

◈ هوش مصنوعی یه حفره امنیتی مهم در PHP پیدا کرد! مدل‌های GPT-5.6 و Gemini 4 به کشف آسیب‌پذیری کمک کردن؛ AI داره به ابزار جدی شکار باگ تبدیل می‌شه.

◈ ربات انسان‌نمای تسلا وارد تولید شد! تسلا تولید Optimus رو در کالیفرنیا شروع کرده؛ قدمی مهم برای آوردن ربات‌های هوشمند به دنیای واقعی.

◈ هوش مصنوعی متا حسابی محبوب شده! اپ Muse در آمریکا از مرز ۱۰ میلیون دانلود در ۴۰ روز گذشته و خریدکردن یکی از کاربردهای پرطرفدارشه.

این پست در طول روز آپدیت می‌شود، دوباره سر بزن ↻

همراه اول و ایرانسل از امروز (اول آبان) تعرفه بسته‌های اینترنت، مکالمه و پیامک را تا ۴۰ درصد افزایش دادن. بسته‌های خریداری‌شده قبلی تا پایان اعتبار بدون تغییر استفاده می‌شوند.

Edit the uploaded photo to create ONE photorealistic cosmetic nose-reshaping preview of the same person. Generate the edited image directly.
The nose MUST look visibly reshaped compared with the original. Aim for a moderate, clearly noticeable transformation that remains natural and proportionate to this specific face. Do not return an unchanged or barely altered nose.
Adapt the changes to the visible nose: smooth a prominent bridge hump if present, refine a broad or rounded tip, moderately reduce excessive apparent width, and gently adjust a downward-pointing tip if visible. Select only changes relevant to the photograph; do not apply every modification automatically. Keep realistic nostril shapes, natural contours, and believable proportions. Avoid an extremely small, pinched, deeply scooped, or sharply upturned nose.
Change ONLY the nose and the immediately surrounding shadows needed to integrate the edit. Preserve the exact same person, all other facial features, skin texture, expression, hairstyle, camera angle, lighting, clothing, background, and framing. Do not beautify the rest of the face.
Before delivering, compare the original and edited nose. If the reshaping is not clearly noticeable at normal viewing size, revise the nose edit until it is visible without exaggeration.
This is an aesthetic simulation, not a medical assessment or a prediction of an achievable surgical outcome. Do not claim to be a surgeon.
Output one edited portrait, without a collage or text overlay. Add this short caption separately:

«شبیه‌سازی ظاهری؛ نتیجهٔ واقعی جراحی ممکن است متفاوت باشد.»
If image editing is unavailable, say so clearly instead of returning the original photo or claiming it was edited.

◈ مایکروسافت یه هوش مصنوعی پزشکی ساخته که توی آزمایش‌های تشخیص بیماری‌های سخت، بیش از دو برابر پزشک‌های عمومی دقت داشته!

◈ کانوا مدل تصویرساز اختصاصی خودش به اسم Craft رو معرفی کرد؛ توی رتبه‌بندی Design Arena از مدل‌های تصویری گوگل و مایکروسافت جلو زده.

◈ رویترز هشدار داده هوش مصنوعی داره باگ‌های خطرناک نرم‌افزارها رو خیلی سریع‌تر پیدا می‌کنه؛ تعداد آسیب‌پذیری‌های کشف‌شده هم نسبت به پارسال دو برابر شده!

◈ اوبر اعلام کرده هر ۳۰ هزار کارمندش رو به ChatGPT Work مجهز می‌کنه تا هوش مصنوعی توی کارهای روزمره شرکت کمکشون کنه.

◈ آنتروپیک اپلیکیشن Claude Duet رو معرفی کرده؛ یه قدم تازه برای تبدیل کلود به دستیار هوشمندی که کارها رو با ایجنت‌ها جلو می‌بره.

◈ سرویس 1Password قابلیت جدیدی معرفی کرده که اجازه می‌ده ایجنت‌های هوش مصنوعی بدون دیدن رمز عبورت، وارد حساب‌هات بشن!

◈ دیپ‌سیک طبق گزارش‌ها حدود ۱۵ میلیارد دلار سرمایه جذب کرده؛ رقابت هوش مصنوعی چین و آمریکا حسابی داغ‌تر شده!

این پست در طول روز آپدیت می‌شود، دوباره سر بزن ↻

Create ONE very natural-looking photograph of the two people in the uploaded identity references. Preserve each person’s recognizable facial features, age, facial proportions, skin tone, eye color, hairstyle, hair texture, and any existing facial hair. Keep their identities distinct. Do not beautify, reshape, or blend their faces.

COMPOSITION AND POSE
Vertical 2:3, closely framed from the chest up. The woman is in the foreground, slightly lower and to the right. The man is immediately behind her, slightly higher and to the left, wearing a plain black hoodie with the hood up. He leans gently toward her, his cheek resting lightly against the side of her head. Their heads are close together in a comfortable, believable pose.

The woman wears an oversized washed-black sweatshirt with long ribbed cuffs. She raises one sleeve over her mouth and the lower part of her nose. Most of her hand remains inside the sleeve, with a few relaxed fingers resting naturally across its upper edge. Preserve realistic hand anatomy and fabric contact; do not spread or arrange her fingers decoratively.

Both look toward the phone lens with relaxed eyelids and quiet, neutral expressions. No forced smiles, exaggerated pouting, or fashion-model posing. Keep their faces at plausible relative sizes and similar distances from the camera.

SETTING AND LIGHT
An ordinary dim room with a plain dark charcoal-gray wall, close behind them. Weak, soft ambient light from a nearby window or room light gently illuminates their faces. The man is slightly darker beneath his hood, but his features remain recognizable. Underexpose the scene modestly while retaining believable facial detail.

No studio lighting, dramatic spotlights, rim lights, or cinematic color grading. Keep muted natural skin colors, dark gray-black clothing, soft shadows, and an ordinary indoor atmosphere.

AUTHENTIC PHONE-PHOTO APPEARANCE
Make this feel like a casual photograph two real people took together on a phone. Use a comfortable camera distance and natural perspective, without wide-angle facial distortion. Slightly imperfect framing, mild low-light sensor noise, restrained detail, and gentle optical softness. Keep both faces within the same believable focus range; the nearby wall should not have artificial portrait-mode blur.

Skin must look unretouched: delicate pores, fine facial hairs, subtle unevenness, and only the small imperfections supported by the references. Do not invent freckles, exaggerate pores, deepen wrinkles, sharpen beard hairs excessively, or smooth skin into plastic. Preserve the woman’s existing makeup without intensifying it. Hair should have natural irregular strands and soft edges, not uniformly sharp individual hairs.

Clothing should have ordinary cotton texture, natural folds, and slightly worn ribbed cuffs. Shadows and fabric compression must match where their heads, shoulders, hand, and sleeve touch.

Add a small, unobtrusive “💪❤️” emoji pair near the upper-right corner, like a simple sticker added to a phone photo. No other text.

OUTPUT
One continuous photograph, not a collage or comparison. The result should feel intimate, understated, imperfect, and convincingly real.
Avoid: digital illustration, painting, CGI, doll-like faces, beauty filters, airbrushing, glossy skin, excessive sharpening, HDR, dramatic teal-orange grading, artificial haze, exaggerated eyes or lashes, altered facial hair, duplicated subjects, extra fingers, malformed hands, logos,and watermarks.