RK (@rukku) profile picture

Code. Comics. Cartography. Information junkie. Researcher at @stamina4space (formerly @phlmicrosat). @maptimediliman

Quezon City
Joined December 2007
Photo by RK (@rukku)Photo by RK (@rukku)Photo by RK (@rukku)Photo by RK (@rukku)Photo by RK (@rukku)Photo by RK (@rukku)Photo by RK (@rukku)Photo by RK (@rukku)Photo by RK (@rukku)Photo by RK (@rukku)
Pinned Tweet
Here's one of community pantries in the Philippines 🕷️ rkda.xyz/spoke/
2
1
28
日本デザインセンターが手がけたサウンドアプリ「033」を、11月16日にリリースしました。 「033」は、世界各地で採録した自然や人々の営みの音を、その背景にある物語とともに届けるアプリです。 フィンランドの図書館に流れる静寂、知床の流氷を歩く音、京都の老舗で職人が餡子をつくる音など、さまざまな土地で出会った音を、現地の写真や旅の記録を綴った「採音日記」とともに収録しています。 仕事や勉強に集中するとき、心を落ち着けたいとき、思考を自由に巡らせたいとき。音を通じて、遠く離れた場所の風景や人々の暮らしに触れてみてください。 アプリはiOS App Storeで配信中です。Android版は後日公開予定。 Appダウンロードはこちら:apps.apple.com/jp/app/033/id… 033 公式サイト:033.audio Product & Business Lead:宮田 洋平 Art Direction, Design:嶋口 智洋 Concept making, Writing:長瀬 香子、廣瀬 智 Planning:伊藤 ひなの Field Recording, Sound Editing:堀 修生*、宮田 洋平 Photography:真島 一樹 Produce:佐藤 大幹 App Development:UNIBA INC.* *社外
11
1,580
189
8,704
618,608
it's funny, my blog post bloom filters was so impactful a screenshot was in Ilya's tenure case, but absolutely ignored in any academic evaluation. I've had more people come up to me about blog posts than any paper I've written. EVEN distinguished papers Academia NEEDS to change
I am increasingly of the opinion that the academic paper is no longer a good medium for disseminating CS research. I think we should move to (1) blog posts and (2) books.
7
22
3
533
21,953
Just read a paper that made me literally say "wow" out loud 🤯 — it uses everyday street-view images (like Google Street View) to measure tree trunk diameter and height with near-LiDAR accuracy. No expensive scanners, no manual field work. The trick? Link[doi.org/10.1080/10095020.202…]
4
16
606
We built a tool for our own internal team at AnswerAI to let us learn, explore, & develop mastery with the help of AI: SolveIt. Today, we’re releasing it to the public for the 1st time. fast.ai/posts/2026-11-16-sol…
65
184
24
1,599
105,008
外网最近很火的一个学习方法:MIT 学霸用 AI,在48小时内快速吃透一门陌生学科。 第一步不是直接问 ChatGPT“给我总结一下”,而是先建立一个足够可靠的 AI 知识库。 他上传了6本该领域的经典教科书、15篇研究论文,以及这门学科的全部课堂讲义。 这样 AI 的回答就不再只是网上东拼西凑,而是基于指定资料帮他梳理知识。 资料准备好以后,他只问 AI 三个问题。 1⃣第一个问题:这个领域有哪些被普遍接受的核心共识?请整理出它的底层概念、关键原则和整体框架,并解释这些知识之间的关系。 这个问题的作用不是总结材料,而是先画出整门学科的知识地图。你需要先知道哪些是地基、哪些是分支,再决定时间应该花在哪里。 2⃣第二个问题:这个领域目前存在哪三个根本性争议?请分别说明各方的观点、依据和反对理由。 一个领域真正有价值的地方,往往不是大家都背得出来的结论,而是专家还在争论什么。看懂这些分歧,才能理解不同理论的边界、证据和隐藏假设,不会把课本上的每句话都当成绝对真理。 3⃣第三个问题:根据这些资料生成10道测试题,覆盖核心概念、因果关系、实际应用和主要争议,先不要提供答案。 接下来关掉资料,自己回答。答完以后再把答案交给 AI,让它逐题批改,并继续追问: “我的回答遗漏了哪些关键点?哪些地方只是记住了结论,却没有真正理解?” 这套方法的精髓,不是让 AI 替你读书,而是让它帮你完成三件事:通过领域共识建立知识骨架,通过领域分歧看清理论边界,再通过测试题暴露自己的认知漏洞。 很多人所谓的 AI 学习,只是把“自己看总结”变成了“让 AI 生成总结”,大脑依然没有参与加工。 这套方法反过来利用 AI 增加学习难度:逼你比较、判断、回忆、输出,再接受纠错。虽然过程更痛苦,却更接近真正有效的学习。 当然,48小时不可能让人真正精通数学、物理这类需要长期训练的学科。 但它确实能让你快速搭起知识框架,找到最重要的问题,避免在陌生领域里漫无目的地看资料。AI负责整理地图,你负责真正走一遍。 ocw.mit.edu/
100
856
43
7,754
333,479
Metro Manila: 4 blocks of land. 124 blocks of people Every block = 0.1% of the Philippines 👇🏻
2
27
1
296
18,750
Exploring landscapes through AI and satellite data. A5 Pyramids uses Google’s AlphaEarth satellite embeddings and PCA to reveal hidden spatial patterns, similarities and differences across the landscape. a5-pyramids.belian.earth/
6
61
1,914
"but how do you balance everything?" i don't. every week, i neglect a different part of my life.
129
39,771
966
217,525
1,967,475
"Really? You're gonna send Steam framerate data?" Dude if Steam asks for a picture of my cock I'm sending it idc glory to the valve empire
It's happening. Steam is starting to collect Frame Rate survey data to display estimated FPS per game based on hardware. No hiding now.
363
15,454
149
164,618
1,490,566
Everyone has drawn this chart in their head and is making a generational bet on which line is true.
🤖 Made with AI
249
467
177
8,114
586,650
From a design standpoint, have your agent build a custom template with Quarto for your decks. All the benefits of Claude but you avoid the design-slop Avoiding Claudeslop in text is way harder, it’s ingrained. Usually better to just write it yourself.
>sit in a presentation on a topic I’m interested in by someone I want to hear from >the slides are obvious claudeslop I’m not going to call the presenter out directly here but it really does make the presentation worse, and people do notice. provided as general advice
1
25
3,319
>sit in a presentation on a topic I’m interested in by someone I want to hear from >the slides are obvious claudeslop I’m not going to call the presenter out directly here but it really does make the presentation worse, and people do notice. provided as general advice
39
25
9
1,066
41,697
20 years of design in 1 tweet
253
3,768
455
34,812
4,829,893
american keigo
this has noticeably changed how i speak at work. He's not familiar with the process.
30
2,572
50
15,443
311,167
DEAR SA ONITSUKA STAFF NA NAG-DEMO NITO, SANA BUONG-BUO YEAR-END BONUS MO! 😭 AKALA KO THREE SEPARATE SHOELACES ANG KAILANGAN PARA SA TRIPLE BOW LOOK, ISANG SINTAS LANG PALA! GRABE, ANG GENIUS NG HACK NA ’TO!
22
3,853
192
41,728
1,146,598
Is he still alive?
Kalian suka gelang kaya gini gakk?
748
6,121
351
65,047
4,597,479
Introducing Manimble: a free, open-source slide editor in which every slide is a Manim program (so underlying the deck is Python). It alows for editing like in Keynote or PowerPoint while seamlessly incorporating AI edits / animations. pachterlab.github.io/Manimbl… 1/🧵
31
153
13
1,201
79,870
XでGrokAIの開発にかかわってるアジア人女性、理系の姫を具現化したような見かけだなと思ったら、声がめっちゃカッコよくてビビった。英語の発音ってやっぱ大事だわ。
89
974
135
14,546
2,659,596
This is probably my best advice: Cultivate fun. Make sure you don’t look anxious and sad and jaded and overthinking in front of other people. Have fun. Look for fun. Force it upon yourself. Be happy. Choose to be happy. Happiness is a sign of abundance. It attracts everything
54
3,555
124
33,692
872,022
THE ENGINEER BEHIND GROK BOT DROPPED A 17-MINUTE GUIDE. AI agents write her code while she sleeps. and almost none of it breaks. Bookmark this for the weekend. 17 minutes. 3,000 PRs. 11,000 for the whole team. Agents merge their own fixes. her secret: she doesn't trust the AI. she builds rules it can't break. full talk below ↓
ANTHROPIC ENGINEER DROPPED A 17-MINUTE CLAUDE CODE MASTERCLASS. This is how Anthropic's own devs actually use Claude Code now. Bookmark this for the weekend. 17 minutes. Real workflows. No fluff. What actually works. Sessions → CLAUDE.md → Permissions → Auto mode → Safety → Scale
11
96
9
1,068
229,546