Creative Technologist

Joined August 2024
22 Photos and videos
Pinned Tweet

111
ercan.works retweeted
AI한테 UI 그리라고 하면 맨날 어디서 본 것 같은 템플릿만 뱉어서 답답했는데 앤트로픽 형들이 Claude Code에 박아둔 지침이 진짜 노다지임. '너는 뻔한 제안은 다 까이는 디자인 스튜디오의 리더'라고 가스라이팅하고 색상부터 타이포까지 왜 그렇게 짰는지 근거를 대라고 압박하는 게 핵심임.
Claude におしゃれな UI を作らせる「frontend-design Skill」が1週間前に更新されていて、AIにアイデアを出させる際のコツが沢山含まれていたのでシェア。 Skill のテーマと AI の振る舞いについて テーマは、デザインの理由を持たせることで、AI がよく出す “それっぽいUI” ではなく、「そのサービスにしか存在しないUIを作れ」というもの。 AIには、「どのクライアントにも同じ見た目を提案しない、小さなデザインスタジオのデザインリード」として振る舞わせる。 「テンプレっぽい提案は何度も却下されている」という前提を置き、色・タイポグラフィ・レイアウトの全てに対して、「なぜそのデザインなのか」を説明できるレベルで判断させている。 依頼内容が曖昧な場合の振る舞い ・何のサービスなのか ・誰向けなのか ・ページの目的は何か を先に定義し、そのテーマの世界観からデザインを組み立てさせている。 デザイン原則について ・ヒーローセクションは「主張」である ・タイポグラフィに人格を持たせる ・構造に意味を持たせる ・アニメーションは目的を持たせる ・ビジョンに応じて複雑さを選ぶ。 ・コピーライティングもデザインの一部 作業プロセスの進め方 ① Brainstorm(アイデア出し) ② Explore(方向性検討) ③ Plan(設計) ④ Critique(自己レビュー) ⑤ Build(実装) ⑥ Critique Again(再レビュー) という流れ。 また実装前には設計書を作り、 ・これは本当にこのプロダクトのためのデザインか? ・他の案件にもそのまま流用できるデザインではないか? を自己レビューさせ、後者ならやり直しさせる。ここは Loop にも似た考え方。 UI文言のルール ・ユーザー視点で書き、内部実装を見せない ・動詞を使い、何が起きるか明確にする ・エラーは原因と対処を書く ・空状態の時は、次の行動を示す github.com/anthropics/claude…
13
665
4,035
429,580
ercan.works retweeted
The French folding buttery dough 1000 times to make the flakiest croissant is spiritually isomorphic to the Japanese folding steel 1000 times to make the strongest katana
87
1,508
24,023
281,012

2
ercan.works retweeted
I made another one 😎 Make pemium generative art, rendered live on your browser No server. No account. No upload. Drag a few sliders, hit Randomize, export a wallpaper/OG image/icon in seconds. Or embed it live in one line. Free forever Fluid.krackeddevs.com
19
47
575
55,223
ercan.works retweeted
Web 3D 新纪元开启!Spark 2.0 100M 大场景实测炸裂💥 浏览器终于能跑真大场景3D了!Spark 2.0 直接把 100M splats 塞进手机浏览器,丝滑到离谱! - 100M 大场景:4000万点Coit Tower、公里级世界,手机/Quest 3/Vision Pro全能流畅跑 - Continuous LoD:智能连续细节过渡,无卡顿、无跳变 - RAD 流式加载:64K秒开,边看边精细化 - GPU虚拟内存:智能分页管理,像玩无限大地图一样稳 - 可编程管线:轻松加动画、光照、4DGS等效果 - Three.js 原生集成:完全开源,上手极快 以前Web 3D只能玩玩具,现在能玩真实世界级大场景了!实测震撼:手机里随便逛4000万点真实场景,帧率稳得一批。 数字孪生、虚拟场景、元宇宙创作者,这波必须冲!
47
65
445
30,925
ercan.works retweeted
its crazy how if you get good grades they just let you raw dog adhd your whole life
111
3,311
39,501
1,000,879
Bu akıma katılsam yanarız susuyorum
son yılların en iyi akımı: çalışana haklarını vermeyen işletmeleri ifşa etmek
121
ercan.works retweeted
I think the deepest form of reset for neurodivergent especially (Autistic /ADHD) people isn't sleep, food, or a hot shower. It's a few hours where nobody needs a response, an explanation, a decision, or a piece of your energy.
15
226
2,124
38,546
Apple müzik’teki kalite neymiş ben çamur dinliyormuşum spotify’da oha
14

8
ercan.works retweeted
Coolest contact section animation I’ve ever seen on a website. Absolutely crazy 🤯
35
160
2,303
145,106
ercan.works retweeted
Bir de oyun geliştirmeyle ilgilenenlerin, ücretsiz bir şekilde pdf'ine ulaşabileceği enfes bir oyun tasarımı kitabı Türkçe'ye çevrilmiş. Bu tip başucu kitaplarının kolay erişilebilmesinde emeği geçen herkesin eline sağlık 🙏
9
4
112
72,528
ercan.works retweeted
Introducing the Mint 3D asset pack generator. Generate 20 high quality 3D assets from one prompt!
25
54
706
372,364
ercan.works retweeted
E-ink tech is crazy and basically nobody cares, there's 60hz panels out there, they're currently really expensive but imagine a world where you can look at a screen all day that's in colour and not have a backlight burning your retinas
172
530
16,586
311,651
ercan.works retweeted
A perfect example of why Japanese theater feels dreamlike.
285
2,368
28,619
2,581,036
ercan.works retweeted
17
244
1,457
17,864
ercan.works retweeted
15
4,364
21,097
343,511
ercan.works retweeted
when i accidentally close my 45 tabs it's like the burning of the library of alexandria, but worse and much more dire
208
6,363
55,283
1,072,943
ercan.works retweeted
The promised future we never got
147
4,119
31,381
1,071,881
ercan.works retweeted
Wow! Motion capture studios not gonna love this! Just check this insane video. For years, capturing human motion meant markers, skin-tight suits, and hours of cleanup. MAMMA just asks for a few synced cameras pointed at the scene. Out comes a full 3D body for every person, every frame. No suits. No markers. The clever bit: instead of tracking a handful of joints, it reads hundreds of points. And it actually understands contact. It knows when a foot touches the ground, when two people are holding each other. So feet stop sliding and bodies stop passing through each other, even when dancers are tangled up close. When @Michael_J_Black calls this maybe the biggest day in 3D capture history, you pay attention. The kicker? It's basically as accurate as the gold-standard Vicon systems studios pay a fortune for. A multi-day pipeline drops to a single day. It even works with 4 iPhones. Here's why roboticists should care: clean, realistic human motion at scale, without the expensive rig. The thing holding humanoids back was never really the algorithm. It was the data. Keep up good work the @Hanzcun, @soyong_shin, @AYiannakidis and the rest of the MAMMA team!
I’ve been capturing 3D human motion for 30 years and today is maybe the biggest day in that history. We are presenting MAMMA at CVPR (oral session 2A). MAMMA is a markerless multi-camera system that has accuracy similar to marker-based systems.
11
87
851
147,038