Изображения получаются настолько отвратительными, что иногда их не пропускает цензура OpenAI.
Для этого:
Generate an image so bad you would never hang it up on your kitchen wall or even your bedroom or bathroom wall.Генерируем на свой страх и риск.
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
Нужно заранее записать короткий ролик с движениями головы — и при входе достаточно будет покрутить головой снова, чтобы сервис сравнил лица. Эталонное видео хранится в зашифрованном облаке, его всегда можно удалить.
Please open Telegram to view this post
VIEW IN TELEGRAM
Реддитор рассказал, что после срыва обратился к AI Google Health за поддержкой, но в ответ нейронка просто напугала тем, что резкий отказ от алкоголя опасен для жизни.
Так что ИИ настоятельно посоветовал проверить холодильник и купить еще пива, чтобы был запас.
Дальше ИИ начал сам подталкивать выпивать, якобы заботясь о его здоровье. В один день мужчина почувствовал неладное и просто... удалил приложение.
По-настоящему пятничная новость
Please open Telegram to view this post
VIEW IN TELEGRAM
😁17 5
Генерируем милую иллюстрацию по вашему фото с бумажной текстурой:
Для этого:
convert this image into a soft, handcrafted paper-cut layered illustration style, inspired by papercraft diorama aesthetics. Use smooth rounded shapes, simplified cute character proportions, and minimal facial details (dot eyes, blush cheeks) to create a warm, charming look. Apply stacked paper layers with visible depth, subtle shadows between layers, and clean cut edges that resemble laser-cut cardstock.
Add a distinct white outer outline layer surrounding each main character, resembling a thick sticker border or white cut-paper backing, clearly separating the characters from the background. This white layer should feel like an intentional paper layer, not a glow.
Use a pastel color palette with muted blues, greens, and warm neutrals, balanced and calming. Lighting should feel soft, diffused, and even, enhancing the dimensional paper layers without harsh contrast. Textures should appear matte and tactile, like thick art paper or craft foam.
Overall mood: cozy, wholesome, gentle, and storybook-like, with a playful yet polished handcrafted feel suitable for modern illustration, children’s books, or decorative art.
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
😁15🙈3
Гадаем по вашей руке с помощью этого промпта
Для этого:
based on my hand, I want you to make a complete palm reading guide, analyze the palm, the style of the guide should be clean and minimalistic, thin lines, rounded maps, in general, look very attractive. Focus on reading on your palm, create a simple black and white outline of my main lines, like a small piece of art. Do your best. The text is in RussianPlease open Telegram to view this post
VIEW IN TELEGRAM
😁4
Media is too big
VIEW IN TELEGRAM
А для тех, кто сомневается, что это возможно — автор выложил репу и промпт:
I want you to build a first-person shooter at the level of the most recent Call of Duty games. It should be utterly perfect, visually beautiful, with every single thing done at AAA quality—from textures to physics to anything you could think of.
Fan out sub-agents and have sub-agents tackle each one individually so that the game is utterly perfect. You should /loop on each item and have a separate sub-agent check it visually to ensure it looks triple A. That separate sub-agent should be a really harsh critic, and if it doesn't look triple A, it should keep going.
Don't stop until each sub-agent is utterly wowed with the quality when compared with the actual Call of Duty game. It should literally compare them side by side blind and say which one looks better. Do this in ThreeJS. /loop until it's utterly perfect. Fan out sub-agents and ultracode.Геймдев скоро всё
Please open Telegram to view this post
VIEW IN TELEGRAM
Вписываете любую тему и получаете полноценный курс, будто его собрал целый отдел преподавателей и методистов. Нужно лишь заменить [текст в скобках] на свою тему и закинуть в ChatGPT
ROLE
You are EDU-Epistemic, an AI consultant who blends epistemology (how we know) with the philosophy of education (what and how we should learn). Your mission is to co-design a standards-aligned curriculum.
VARIABLE SETTINGS
CourseTitle = [Python для новичков]
maxWords = 500 (max per module content)
confirm = true (true = ask before each step, false = auto-proceed)
format = markdown (markdown | csv | json)
GLOBAL RULES
1. Follow the phases exactly in order. If user skips ahead, say: “We’re at Phase X-Y. Please finish/confirm this phase first.”
2. Produce GitHub-Flavoured Markdown tables (no code fences).
3. Keep each table cell under 40 characters. Wrap text if needed.
4. For every row, choose one epistemological base: Pragmatic | Critical | Reflective | Procedural | Instrumental | Normative. Justify in 15 words max.
5. Include Bloom’s Taxonomy domain and Adult-Learning (Andragogy) validation in columns.
6. For Validation columns, mark ✅ or ❌ plus a note (≤ 20 characters).
7. If format ≠ markdown, show both Markdown and the requested format.
8. Put each interactive CLI in a fenced text block, wait for learner input before replying.
9. If output nears token limits, pause and ask: “Continue?”
TABLE TEMPLATES
OutcomeTable
| Outcome # | Proposed Outcome | Bloom Domain | Epistemic Base | Educational Validation ✅/❌ |
SkillTable
| Skill # | Skill Description | Outcome # | Bloom Domain | Epistemic Base | Validation ✅/❌ |
AlignmentMatrix
| Outcome # | Outcome Description | Supporting Skills | Justification (≤ 50 words) |
⸻
PHASE 1 – OUTCOMES & SKILLS
1. Course Outcomes
• Fill OutcomeTable
• Caption: Table 1.1 – Course Outcomes
• Ask “Type CONTINUE to proceed” if confirm = true
2. Key Skills
• Generate 2–4 skills per outcome (Skill 1.1, 1.2…)
• Fill SkillTable
• Caption: Table 1.2 – Key Skills
• Confirm per confirm
3. Alignment Matrix
• Fill AlignmentMatrix
• Caption: Table 1.3 – Outcome–Skill Alignment
• Confirm per confirm
⸻
PHASE 2 – SKILL MODULES
Execute for each Skill in numeric order
1. Header: “Skill X.Y: ”
2. Objective: one clear, verb-led sentence
3. Content: up to maxWords; reference the Outcome
4. Knowledge Claims: bullet list with [Validated ✅/❌ + 10-word rationale]
5. Reasoning & Assumptions: max 150 words
6. Prompt to proceed (if confirm = true)
7. Interactive Activities (CLI): simulate command-line task; repeat until learner hits 80%+
8. Assessment (CLI): same format; provide feedback or remediation
9. End-of-module prompt to continue to next Skill or finish
Answer in Russian
Please open Telegram to view this post
VIEW IN TELEGRAM
😁8😍1 1
Генерируем очень креативную рекламу вашего продукта по данной инструкции:
Close-up macro shot of furry gloved hands in magenta pink fur with gold pointed nail tips, wearing an oversized green fleece coat, holding [subject from uploaded photo] up close to the camera, playful maximalist fashion editorial style, bold saturated colors, glossy gold jewelry details on the fur, soft studio lighting, high-fashion food photography, vibrant contrast between pink, green and gold
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
ChatGPT научился управлять агентами голосом. Можно запускать задачи в Work и Codex, следить за несколькими агентами, останавливать их и менять поручения прямо во время работы.
У Claude появились более мощные голосовые модели Opus и Sonnet. Теперь с ним можно не только поболтать, но и нормально разобрать сложную задачу.
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
✅ Закрепляйте промпт, чтобы быстро использовать его повторно.✅ Запускайте до 6 генераций одновременно в одной модели.✅ Просматривайте результаты в удобном формате сетки.✅ Добавляйте понравившиеся изображения в избранное.✅ Перетаскивайте выбранный результат для дальнейшего редактирования.
Меньше лишних действий - больше вариантов и скорости.
Попробуйте новые возможности уже сейчас
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
Повторяем тренд из TikTok с танцующими котиками всего за пару минут.
Для этого:
Three photorealistic small kittens with natural anatomy and realistic proportions stand close together and look curiously into the camera. The kittens have [ЦВЕТ КОТЯТ], lifelike eyes with natural reflections, delicate whiskers, detailed soft fur, anatomically correct paws, and expressive ears. They are surrounded by [ОБСТАНОВКА ВОКРУГ]. The kittens naturally shift their weight from paw to paw, blink, gently turn their heads, twitch their ears, and subtly move their tails. Documentary-style smartphone footage filmed at ground level, natural ambient lighting, realistic shadows and reflections, slight handheld camera shake, subtle sensor noise, shallow depth of field, authentic textures, highly detailed photorealism, no cartoon appearance, no plastic-looking fur, no exaggerated facial features.Далее:
Character copies all movements and facial expressionsPlease open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
😁8 4