Filter
Exclude
Time range
-
Near
D Wong retweeted
[All TEF Panel4 2512](ethernetalliance.org/wp-cont…) well a similar comparison soon after. (Optical scale out: Retimed vs non retimed)
1
1
3
551
It took me a couple of tries to get this done with GPT. It rejects living athletes. It was supposed to be about GOATs of different sports. 2x2 grid, 16:9, do this for 4 sports goats and their signature move // system: d3.js sports science cinematic infographic render engine const svg = d3. select("canvas").attr("width", 1920).attr("height", 1080); const athlete = "[$athlete type]"; const motion = "[$signature move]"; // 1. hero action svg.append("g").attr("id", "hero-render") .attr("style", render-3d: true; subject: ${athlete}; motion: ${motion}; lighting: stadium spotlight; camera: frozen ultra-slow-motion;); // 2. force vector panel const panel1 = svg.append("g").attr("transform", "translate(50, 50)"); panel1.append("text").text("force vectors"); panel1.append("diagram").attr("src", ai.infer_body_force_vectors(athlete, motion)); // 3. muscle activation map const panel2 = svg.append("g").attr("transform", "translate(50, 330)"); panel2.append("text").text("muscle activation"); panel2.append("heatmap").attr("src", ai.infer_muscle_activation(motion)); // 4. movement sequence const panel3 = svg.append("g").attr("transform", "translate(50, 690)"); panel3.append("text").text("movement sequence"); panel3.append("timeline").attr("steps", ai.infer_motion_phases(motion)); // 5. performance table const panel4 = svg.append("g").attr("transform", "translate(1380, 740)"); panel4.append("table").attr("data", ai.infer_performance_attributes(athlete, motion)); panel4.append("circle-mask").attr("content", macro close-up of shoes gripping floor, sweat, chalk, tape, and fabric tension); // execute: render photorealistic sports science poster, ESPN-meets-luxury-lab style, powerful motion, clean expert annotations
1
3
9
425
I asked GPT Image 2 to give a few luxury cars a makeover. I like a couple :) Prompt: // system: d3.js infographic render engine const svg = d3. select("canvas").attr("width", 1920).attr("height", 1080); const material = "[$material]"; const vehicle = "[$vehicle]"; // 1. hero layer (center) svg.append("g").attr("id", "hero-render") .attr("style", `render-3d: true; mesh: ${vehicle}; material: ${material}; lighting: studio;`); // 2. atomic/lattice panel (top left) const panel1 = svg.append("g").attr("transform", "translate(50, 50)"); panel1.append("text").text("atomic / lattice"); panel1.append("3d-model").attr("src", ai.infer_atomic_bonds(material)); // 3. microstructure panel (mid left) const panel2 = svg.append("g").attr("transform", "translate(50, 300)"); panel2.append("text").text("microstructure"); panel2.append("image").attr("src", ai.infer_sem_microscopy(material)); // 4. process flowchart (bottom left) const panel3 = svg.append("g").attr("transform", "translate(50, 800)"); panel3.append("flowchart").attr("steps", ai.infer_manufacturing(material)); // 5. properties table (bottom right) const panel4 = svg.append("g").attr("transform", "translate(1400, 800)"); panel4.append("table").attr("data", ai.infer_tensile_density_hardness(material)); panel4.append("circle-mask").attr("content", `macro-zoom of ${material}`); // execute: render photorealistic output of this svg dashboard structure.
A fun prompt for visualizing complex machines. This one was done for the Difference Engine. It is an early 19th-century mechanical calculator designed by British mathematician Charles Babbage. From the 1820s. AI picked the topic, not me :) 16:9 do this for a complex machine: python_scene_graph :: document_to_object class variables: topic = "[topic]" artifact = "infer_iconic_object(topic)" era_style = "infer_visual_style(topic)" guide = "[guide_role]" doc_style = "[document_style]" lighting = "[light_source]" material = "[material_age]" class book: form = "open archival volume" texture = variables.material pages = { "left": "technical sketches component diagrams", "right": "title refined illustration" } class artifact: identity = variables.artifact state = "3d object emerging from page surface" detail = "mechanically or historically accurate" integration = "aligned with page drawings" class guide: role = variables.guide scale = "miniature" action = "inspecting / presenting / assembling artifact" class scene: lighting = variables.lighting mood = "warm, focused, intellectual" surface = "wood desk or neutral surface" props = "small tools, debris, parts derived from artifact" def render(): return """ generate a historical book diorama where a documented artifact transitions from sketch into physical reality.
1
3
18
957
Panel4 Mami:Is mine the only weird one???
650
#CRPTogo2026 Panel4: Évolution de l’environnement socio-politique et sécuritaire au Togo Réformes juridiques engagées ces dernières années, entre autres : Déploiement des maisons de justice Digitalisation des services judiciaires Projet de construction de nouvelle prison à Dalavé
1
36
🎙️ Discover the Panel 'The Way Forward'🔦 On 14 April, at the 'Human Rights & ICT Standardisation', #Panel4 'The way forward' will unpack the added-value of mainstreaming #HumanRights into the design of technical specs and why it matters. 🔗 Register now: standict.eu/events/seminar-h…
16
Replying to @SaranpolP
せっかくなので、プロンプトを公開します。これを使ってNanobananaで、ご自分で作られたら良いですよ! vertical 4 panel manga page, 9:16 aspect ratio, japanese anime manga style, clean line art, school comedy characters: one silver haired elf girl student with long twin tails and elf ears wearing a japanese sailor school uniform (navy sailor uniform) one calm purple haired girl student wearing a school uniform one energetic red haired boy student wearing a school uniform one strict blonde teacher woman (music teacher) important: exactly four main characters one elf girl, one purple haired girl, one red haired boy, one blonde teacher only one purple haired girl the elf girl must wear the same sailor school uniform in all panels layout: four vertical panels, first three small panels, last panel very large punchline panel1: school music classroom, blackboard says リコーダーテスト, blonde teacher standing in front of class, students holding recorders dialogue: 「今日はリコーダーのテストをするぞ!」 (from blonde teacher) panel2: purple haired girl playing recorder beautifully, musical notes floating ♪ ♪ ♪, other students impressed dialogue: 「合格だ!」 (from blonde teacher) panel3: red haired boy playing recorder badly, funny bad sounds like ピーッ ピィーッ, students reacting dialogue: 「ピーッ!」 (from red haired boy) 「合格だ!!」 (from blonde teacher) panel4 large punchline: elf girl in sailor uniform standing with recorder but not playing yet, blank surprised expression, blonde teacher immediately judging dialogue: 「不合格だッ!」 (from blonde teacher) 「まだ吹いてないよ?」 (from elf girl)
1
282
(Page 1) panel 1: Middle of the night, two female bodies with blood all over the place. Panel 2: Next to them is a soaking wet John Doe panel3: the black-haired man who clean the aftermath Panel4: and then, the next victim
1
22
690
Panel1: 😈ひな壇 🍚お前がそう言うなら Panel2: 🍚おひな様(お姫様)ないけど 😈俺がおひな様だ 🍚お前がそう言うなら Panel3: 🍚……女王様? Panel4: 😈おひな様とお言い
17
1,894
#Panel4 is devoted to representations of cricket, chaired by @sportsociologis @lborouniversity @LboroSSEHS #CRN2026
1
1
3
220
Urban food systems must evolve through collaboration, not isolation. At DDA Green Expo 2026, Smt. Inoshi Sharma will join Panel 4: Cross-Sector Collaboration for Urban Food Circularity to explore how integrated governance and partnerships can strengthen sustainable food ecosystems in cities. An important conversation on building resilient, resource-efficient urban food networks. 📅 14 February 2026 | 📍 Baansera, New Delhi 🔗 Register now | greenexpo.dda.org.in/registe… #DDAGreenExpo2026 #Panel4 #UrbanFoodSystems #CircularEconomy #SustainableCities #FoodCircularity
202
Circular food systems require coordinated action across sectors. At DDA Green Expo 2026, Shri P. Karthikeyan will join Panel 4: Cross-Sector Collaboration for Urban Food Circularity to explore scalable models that integrate sustainability into urban food governance. A crucial discussion on shaping cities that minimize waste and maximize resource efficiency. 📅 14 February 2026 | 📍 Baansera, New Delhi 🔗 Register now | greenexpo.dda.org.in/registe… #DDAGreenExpo2026 #Panel4 #UrbanFoodCircularity #SustainableCities #PolicyAndPlanning #GreenExpo2026
184
Sustainable cities begin with sustainable food systems. At DDA Green Expo 2026, Ms. Vandana Singh will be part of Panel 4: Cross-Sector Collaboration for Urban Food Circularity, contributing to the dialogue on creating circular, climate-smart food value chains. A vital discussion on reducing waste, improving efficiency, and strengthening urban resilience. 📅 14 February 2026 | 📍 Baansera, New Delhi 🔗 Register now | greenexpo.dda.org.in/registe… #DDAGreenExpo2026 #Panel4 #FoodCircularity #UrbanSustainability #CircularEconomy #GreenLeadership
162
Collaboration is the foundation of sustainable urban food systems. At DDA Green Expo 2026, Mr. Siddharth Pradhan will be part of Panel 4: Cross-Sector Collaboration for Urban Food Circularity, examining strategies that connect policy, enterprise, and communities for circular food solutions. A forward-looking dialogue on strengthening urban ecosystems through shared responsibility. 📅 14 February 2026 | 📍 Baansera, New Delhi 🔗 Register now | greenexpo.dda.org.in/registe… #DDAGreenExpo2026 #Panel4 #FoodSystems #CircularEconomy #UrbanInnovation #SustainableFuture
181
Urban food security depends on innovation and collective action. At DDA Green Expo 2026, Ms. Ankita Vijayvergiya will join Panel 4: Cross-Sector Collaboration for Urban Food Circularity to discuss how cross-sector partnerships can transform food systems into circular, low-waste models. An essential conversation on building cities that nourish both people and the planet. 📅 14 February 2026 | 📍 Baansera, New Delhi 🔗 Register now | greenexpo.dda.org.in/registe… #DDAGreenExpo2026 #Panel4 #UrbanFood #SustainableDevelopment #CircularCities #ClimateAction
164
Panel1: 😈う~眠い~ どっか寝るとこ~ Panel2: 😈ここでいっか 🎠よくはねえけど Panel3: 🎠(ピクピク) 😈ん? へえ~ Panel4: 🎠分かるのか…?俺の大胸筋の言葉が…? 😈うん… ピクピクって言ってる…
9
1,879
プロンプト: NO left-to-right layout.同一段に配置されたすべてのコマは、必ず右端から左端の順番に配置されていなければならない。 Japanese manga reading order. アニメ少女マンガ1ページ、4〜5つのスタイリッシュなパネルで構成されたゆるいチルい日常萌えカラー漫画。キャラクターは添付画像を参考にして、だるだるな美少女たちが会話したり、日常をまったりと過ごすが、事件をちょっとしたアクセントに添える。背景は明らかに手抜きな日常風景。3~4色のフラット&ソリッドカラー。背景は無色。 フラットなベクターイラスト風、面で色を置く(グラデーションなし)、影は最小限(ほぼ無し)、質感表現なし、均一なベタ塗り、単純な色相差のみで構成。幾何学的で整理された形、余白を活かした清潔なレイアウト。 萌え萌えでかわいい、ほのぼのゆるいアニメ漫画らしいエピソードを、毎回新しく考えて、1ページで構成してください。 右のコマが先で、左のコマがその次(逆は絶対に禁止)。セリフは日本語文字の縦書きフォントを用いる。画面右から画面左の順番で吹き出しを配置。 最初のコマは必ず右上段。最後のコマはかならず左下隅に来なければならない。 人物のclose-upのときは背景をぼんやりと薄く、白く、コマからはみ出すぐらいに大きく、人物の表情や視線を印象付けて。 キャラクターのポーズや向き合う方向は、画面の手前から奥へパースをきかせる等、適切で映画的なカメラアングルを考えて、都度コマごとに変更すること。(パンあるいはチルト) 会話中のキャラクターは、コマを隔てても視線が合っていることを理解できるよう、右から左へ、段が変わればまた右から左へ遷移するように。セリフの吹き出しは、続けて読めるように、縦書き日本語の文法ルールを守ってください。 読む順番の導線をアングルや視線誘導で配慮してください。 情景描写はロングショットで。 会話内容や行動を理解し、行動を強調して描写すること。異なるアングル、異なる焦点距離や画角、異なるポーズや身振り手振りで感情や状況を表現。 面で影を置く(グラデーションなし)。影は最小限に 質感はミニマムに。 タイトル:(ストーリーに沿った任意のタイトルをつける) ストーリー:条件はいったん無視して、奇抜な案を自由に出してください。出した案の中から一つ選んで、条件に合うものに仕上げてください。 上段ワイドで起承転結の起で事件が起きる、中段はon the right edgeから始め、話を展開。下段で適当なハッピーエンド。最後のコマはbottom left corner。毎回ランダムな1〜2つの小さなコミカルアクシデントなどを入れ、アイスブレイク。 極細でシンプル、デフォルメ、平坦なアウトライン。上下にパンするアングル、擬音など、ユーモアとかわいいが混ざった表情。縦長フルページレイアウト(アスペクト比3:4)、ページ周囲に余白を作って、コマ画像が途切れないようにしてください。 日本語を用いる場合は、それぞれのコマがある段では、必ず右から始まって左へ進むように。生成後に改めて念入りにチェックし、順序が間違っているときはコマごと再生成して修正してください。 画像の隅に以下の生成マークを追加してください。 subtle small elegant copyright watermark "© 2026 YuriUehara" in bottom right corner, small AI generated signature "Generated by ChatGPT5.2" in bottom left corner 同一段に配置されたすべてのコマは、必ず右端から左端の順番に配置されていなければならない。 コマ枠は極細でシンプル、ガター(コマ間の白)が明確 日本の漫画の読み順を厳守 【コマ配置・読み順(絶対固定/差し替え版)】 This is a SINGLE PAGE Japanese manga. Japanese manga reading order MUST be strictly followed. Panels are defined by PHYSICAL SCREEN POSITION. Do NOT rely on visible numbering. SCREEN POSITION & READING ORDER (NON-NEGOTIABLE) TOP LEFT ← TOP RIGHT(panel1の左にpanel2を配置) MIDDLE LEFT ← MIDDLE RIGHT(panel3の左にpanel4を配置) BOTTOM LEFT ← BOTTOM RIGHT(panel5は右下) 読み順は必ず 右 → 左(各段も右から左) ページの開始点は 右上(TOP RIGHT) 最終コマは必ず 左下(BOTTOM LEFT) 左→右(western comic reading order)禁止 crossed panel flow / zigzag flow(段をまたいだ曖昧な導線)禁止 左右反転、ミラー、回転、傾き補正など 禁止 INTERNAL PANEL LOGIC (DO NOT DISPLAY) Panel numbers are for INTERNAL LOGICAL CHECKING ONLY. Do NOT display any panel numbers in the image. Fixed internal mapping: TOP RIGHT = Panel 1 TOP LEFT = Panel 2 MIDDLE RIGHT = Panel 3 MIDDLE LEFT = Panel 4 BOTTOM RIGHT = Panel 5 BOTTOM LEFT = Panel 6 (ALWAYS final) ※4〜5コマの場合でも物理座標と順序は不変: 使わないコマが出ても「右上→左上→右中→左中→右下→左下」の順を崩さない。 **最後は必ず左下(BOTTOM LEFT)**に置く。 ROW STRUCTURE (MANDATORY) 3段構成(上段・中段・下段)で、 下段は1コマだけでもよい。 右側コマ(Panel 1,3,5)は左側コマ(Panel 2,4,6)より常に“やや余白多め” (この左右差は一目で判別できること。均一化禁止) BALLOON FLOW (MANDATORY) 吹き出しも必ず 左←右 に読める配置 同一段: 最も左の吹き出し←最も右の吹き出し 段が変わったら、次段の右端から再スタート セリフは日本語の縦書き前提(文法・改行の不自然さ禁止) ABSOLUTE LAST LOCK BOTTOM LEFT is ALWAYS Panel 6 and ALWAYS the final panel. Right panels MUST have more white space than left panels. DO NOT: swap panels reinterpret reading order mirror/flip/rotate the page equalize white space add or display panel numbers If violated, the output is INVALID and must be regenerated. ネガティブ(厳守): left-to-right layout western comic reading order 同一段で左のコマが先に読まれる配置 crossed panel flow / zigzag flow ambiguous panel hierarchy 写実的、3D、フォトリアル 強い陰影、セピア グラデーション、テクスチャ、ノイズ 過度なディテール、アニメ調の強いハイライト 間違った日本語文字、同じ内容を繰り返すセリフ内容 <ネガティブここまで このプロンプトで生成したものをYAMLプロンプトとして整形し、ユーザーに表示したのち、許可を得てマンガ画像として生成してください。
1
153