https://www.facebook.com/share/p/1DfQoodzZ8/

RT @esorhjy: 如果你本來就是 ChatGPT 的使用者，就算不是程式開發者，也很推薦試試看 Codex。 延續之前提過的，我現在用 Codex 維護一個不用筆記軟體、純用 AI 管理的外部資料庫系統。最近我又讓它多做一件事： 每天早上自動根據我最近關注的主題（外部資料庫中建立的新筆記、永久筆記），去 blog、YouTube、論壇抓固定數量的新資料，整理好原文、摘要，等我起床後直接看。 這一次連 RSS 閱讀器、稍後閱讀等工具都不需要另外去操作。如圖，一張是我一開始寫給 Codex 的需求，一張是它實際跑完之後，整理給我看的總覽大綱（另外還有每一篇的完整原文、摘要筆記）。 #以前如果要做到這件事，我大概要自己去操作筆記、 RSS、稍後閱讀軟體，研究自動化工具，想辦法把它們串起來。 但這次我做的事情其實很單純：就是在 Codex 中用一段對話，先把需求講清楚（如圖）。 我跟 AI 大概說明了下面幾個工作流程，例如： 搜尋主題要從我最近的筆記與貼文主題去推斷 來源要分成 blog、YouTube、論壇 每天抓固定數量的內容 先放暫存區，不要直接混進正式資料庫 我要先看 review summary，再決定哪些正式收錄 第二天自動抓新資料時，前一天沒收錄的暫存內容就清掉 剩下那些技術問題，像是怎麼分析我最近的筆記、怎麼設定搜尋主題、怎麼判斷對象與範圍，或是怎麼抓文章、字幕、討論串，以及怎麼整理成原文與摘要，幾乎都交給 Codex 自己去處理了（他會自己寫小工具，或是設定好流程自己用內建技能處理）。 用自然語言描述完我需要的工作流程，接著讓 AI 自己寫程式、跑測試，把整條自動化流程搭建起來。 如果是非工程師的用戶，也可以試試看，因為在這個案例中，真正的門檻不是會不會寫程式。而是練習把自己要的流程、邊界、例外和驗收條件講清楚。 以前我們常常是先學很多工具怎麼操作，才有辦法慢慢拼出一條流程。現在的 AI 工作時代則是： #先把你要的流程說清楚，#再讓AI去幫你把工具和自動化搭起來。 如果你沒有程式背景，但本來就是 ChatGPT 使用者，我真的覺得可以試試看。 先下載 Codex （ https://t.co/RQa6wCPLjL ），開一個新對話，指定一個新的資料夾給它，先從一條很小的流程開始就好。例如： 幫你整理每天固定追的資料 幫你做一份 review summary 幫你維護一個自己的 md 資料夾 我最近也把這套做法整理成一篇比較完整的文章，歡迎參考： AI 結合卡片盒筆記法，人不再操作軟體，用對話流程讓 Codex 搭建資料整理系統：我的兩個月實測心得https://t.co/ni6uoOgbhU #ChatGPT #Codex #AI工作流 #第二大腦

RT @drkeithsiau: Anatomy of the shoulder 📍

RT @OGdukeneurosurg: Peripheral nerve injuries

RT @OGdukeneurosurg: Peripheral nerves

節鈔 楞嚴經

RT @TrackYourHeart: Diagnostic algorithm and triage in acute coronary syndrome

蘭亭十三跋-1 & 2

RT @bsla_32: WBCに向けて、台灣代表の選手名鑑を作成しました。 ・全30選手+首脳陣名鑑 ・戦力分析 ・チア+団長紹介 ・応援歌歌詞カード ・対戦国紹介(日韓は全選手ミニ寸評付き) で全27ページです。 観戦のお供にご活用ください！ https://drive.google.com/file/d/1fXQBGhTBUHMQgSqzfjGmueWKiBteUfjj/view?usp=drivesdk #WBC #台灣

RT @jointstaffpa: 令和8年2月28日(土)、中国軍の情報収集機（Y-9）x2機が東シナ海から沖縄本島と宮古島との間を通過して太平洋に至り、奄美群島沖まで飛行したことを確認し、これらに対して #航空自衛隊 の南西航空方面隊の戦闘機を #緊急発進 させ対応しました。#JASDF #ｽｸﾗﾝﾌﾞﾙ

RT @unkaido_koten: [書の古典] 趙眘(ちょう しん) 草書后赤壁賦巻：南宋（孝宗）・遼寧省博物館蔵

RT @Object_Zero_: Taiwan Taiwan currently produces 90% of the world’s advanced silicon chips. The Taiwanese semiconductor industry consumes 7.2 GW or power each year, and each year it produces enough silicon chips to consume 45 GW of power. Nvidia’s AI chips like H100 and B200 account for 5-6 million units, which is 4.5 GW of the 45 GW of compute produced by Taiwan. Nvidia’s AI chips recently overtook the global server CPU market for “GW of compute”, and Nvidia’s AI chips recently production doubled in the past 12 months from 2.2 GW in 2024. So for every 1GW of power generation on the island of Taiwan, Taiwan is able to produce around 6 GW of compute. By 2030, every 1 GW of power generation on the island of Taiwan will result in 20 GW of compute manufactured. Meaning Taiwan will be able to produce around 144 GW of chips of which 40 GW will be CPU, servers, etc, etc and around 104 GW will be Nvidia AI chips. This represents a 2,300% increase in the energy needed to power Nvidia’s product output. In the whole world today, there is 20 GW of AI. Nvidia is projecting the following: 2026 = 4.5 GW of chips sold 2027 = 10.0 GW of chips 2028 = 22.0 GW of chips 2029 = 40.0 GW of chips 2030 = 80.0 GW of chips It’s not that they are making 20x more chips, they will be making 3x more chips and the chips will each demand 7x more power. This is the Grid shockwave that is coming, and Elon is manoeuvring to use SpaceX to catch Jensens massive overproduction capacity, that the market otherwise cannot deploy. Where does the revenue come from? Well corporate payroll is $20 trillion is USA alone and is $60 trillion globally. AI is not an app, it’s an economy. It is coming for corporate payroll. It doesn’t need to take 100.0% of corporate payroll, it might only take 10-20%. But that’s $6-12 trillion of revenue for the AI industry. Which is more than enough to pay for the chips. Prices don’t really matter here, all that matters is capacity and schedule. This is the supersonic tsunami. It will knock unemployment to 25% it will deflate almost all knowledge and services wages / income, and it will spike commodities and manufacturing sharply. The epicentre of this is supersonic tsunami is Taiwan. Taiwan is also the geostrategic flashpoint of the world’s two preeminent powers. Good luck Taiwan! I’m sure whoever loses the AI race will be benevolent and gracious in defeat, and accepting of their new century of subordination. 😬

RT @NextScience: 🧠 MIT recently completed the first brain-scan study on ChatGPT users—and the results are deeply revealing. Rather than boosting brain function, prolonged AI use may be dulling it. Over four months of cognitive data suggest we might be measuring productivity all wrong ⤵️ In MIT’s study, participants had their brains scanned while using ChatGPT. → 83.3% of users couldn’t recall a single sentence they’d written just minutes earlier. → In contrast, those writing without AI had no trouble remembering. Brain connectivity dropped sharply—from 79 to 42 points. → That’s a 47% drop in neural engagement. → The lowest cognitive performance among all user groups. Even after stopping ChatGPT use in later sessions, these users showed continued under-engagement. → Their performance remained lower than those who never used AI. → This suggests more than dependency—it’s cognitive weakening. Beyond the scans, educators flagged the writing itself. → Essays were technically solid, but often called “robotic,” “soulless,” and “lacking depth.” Here’s the paradox: → ChatGPT makes you 60% faster at completing tasks… → But it reduces the mental effort required for learning by 32%. The top-performing group? → Those who began without AI and added it later. → They retained the best memory, brain activity, and overall scores. Using ChatGPT can feel empowering—but it may quietly offload your thinking. → You gain speed, but lose engagement. → You get answers, but stop learning how to think. The takeaway isn’t to avoid AI—but to use it intentionally. → Use it to assist, not replace your mind. → Build cognitive strength—not dependency. MIT’s early study on AI and the brain lays out the stakes. The way we use these tools matters more than ever.

RT @Doctors_GUILD: CROUP VS ACUTE EPIGLOTTITIS

RT @tvbsworldtaiwan: Taiwan showcases F-16V jets with new Sniper targeting pod at Chiayi Air Base, with pilots ready to scramble in 6-15 minutes. ★ https://news.tvbs.com.tw/english/3113496 ★

RT @nervemed: Tuberculosis Treatment Guidelines

RT @unkaido_koten: [篆書の古典] 傅山(ふ ざん) 天龍禅寺五言詩篆書軸：明末清初・上海博物館蔵

RT @netflix: Alex Honnold’s selfie from the top of Taipei 101 after his historic free solo. #SkyscraperLIVE

1-3 懷素 〈自敘帖〉Huaisu, Autobiography https://youtu.be/g_r8V_B-hPU?si=gfMjuGyJTpN0hXAj 來自 @YouTube

RT @Doctors_GUILD: Causes of Sudden Painless Loss of Vision Mnemonic: “CARROT” 🔹 C – Central retinal artery occlusion 🔹 A – Amaurosis fugax 🔹 R – Retinal detachment 🔹 R – Retrobulbar neuritis (early) 🔹 O – Occipital lobe infarct 🔹 T – Temporal arteritis Painful loss of vision? → Think optic neuritis, uveitis, glaucoma