人工智慧概論 Introduction to Artificial Intelligence

Yes. There are no prerequisites: no programming, Git, or command-line background is required. You start by making something useful and interesting, then learn how to hand a task to AI, preview what it proposes, and check the result.可以。這門課零先修，不要求程式設計、Git 或命令列背景。你從做出有趣又有用的東西開始，再學會怎麼把任務交給 AI、預覽它要做的動作、檢查結果。

Payment: never—no API key, paid plan, command line, or phone verification. Accounts: three are worth setting up before term—GitHub, one agent desktop app, and one chat AI (see "What should I prepare before term starts?"); none are graded, and equivalent paths cost you nothing if they fall through. The only public artifact is the Week-4 GitHub Pages exercise, done in a public repo with synthetic content and taken down in class; if you'd rather not publish, follow the instructor's demo site for the same credit. Your university account signs you into e-Learning for submissions and Zuvio for roll call—not an extra sign-up, and it costs nothing.付費一律不用：不要求 API key、付費方案、命令列或電話驗證。帳號有三個請在課前辦好——GitHub、agent 桌面 app（二擇一）、一個聊天型 AI（見「開學前要先準備什麼？」）；這些不計分，辦不成有等值路徑、不扣一分。公開作品只有第 4 週的 GitHub Pages 練習用公開 repo＋合成內容、當堂教撤下；不想公開改看教師示範站，同分。學校帳號用來登入 e學苑交作業、Zuvio 點名，不是額外註冊也不要錢。

No. Claude Academy and OpenAI Academy are only a source for the instructor's preparation and optional extension for you. They are not graded, no certificate or account screenshot is collected, and they are never stacked on top of the three class hours.不用。Claude Academy 與 OpenAI Academy 只當教師備課來源與學生自選延伸，不計分、不收證書，也不收帳號截圖，更不會疊在課內三小時之外。

No. Conditional tools are not the shared core. Use a school-approved tool, instructor-generated material, or a fully offline path—as long as you leave equivalent work and evidence, the same rubric applies, and not using a given product costs you no marks.不用。條件式工具不是共同核心。你用學校核可的工具、教師預生成的內容或完全離線的路徑，只要留下同等的作品與證據，就用同一份評分標準評分；沒用那個產品不會扣分。

Assignments are 20% of the grade. They are set according to teaching progress and materials, so not every week has one; the content, the number of pieces and the submission route are announced in class and on e-learning. Participation is 10% and is based on attendance and in-class learning records.作業佔總成績 20%，依教學進度與教材安排，不預設每週都有；內容、次數與繳交方式於課堂及 e學苑公告。平時成績 10% 看出席與課堂學習紀錄。

The midterm is 30% and takes the form of a midterm AI maker gallery whose scope is the skills of Weeks 2–7 (Week 1 orientation and Week 8 multimodal are not examined). You bring one improved piece of work, attributable and verifiable evidence of the improvement, and a personal judgment card. Week 8 multimodal work may strengthen your display, but it is not examined.期中佔 30%，形式是期中 AI 作品展，範圍＝第 2–7 週的能力；W1 導覽與第 8 週多模態都不考。你帶一件改良過的作品、具名可驗證的改良證據，以及一張個人判斷卡。第 8 週的多模態作品可以用來加強展示，但不是必考。

The final is 40%: an AI workflow project combining at least one primary and one supporting AI capability, three tests, and one failure with a recovery. You may extend and deepen work from earlier weeks—two brand-new tools are not required, and technical difficulty, paid tools, or public deployment earn no extra credit on their own. Selected work from previous terms is shown in class for reference.期末佔 40%，是一個 AI 工作流專題：至少整合一項主要與一項輔助的 AI 能力、三筆測試，以及一次失敗與復原。可以沿用並深化前面幾週的作品，不要求兩套全新工具；技術難度、付費工具與公開部署本身不加分。課堂會展示過去學期的優秀作品當參考。

Follow the pre-course checklist and set up four things: a typing-friendly setup (classroom PCs are available if you have no laptop), a GitHub account (used in Week 4), one agent desktop app (Antigravity needs a personal Google account and age 18+; otherwise the ChatGPT desktop app), and one chat AI. Three reminders: once installed and signed in, stop—don't burn your quota early; use a personal Google account for Antigravity, not your school email; the Gemini student plan is worth one application, but it requires a payment method and auto-renews after a year—you can finish this course without it. None of this is graded or a gate; you can attend Week 1 without it.照課前準備清單先弄四樣：能打字的環境（沒筆電可用教室機）、GitHub 帳號（W4 用）、agent 桌面 app 二擇一（Antigravity 要滿 18 歲並用個人 Google 帳號，或 ChatGPT 桌面版）、一個聊天型 AI。三個提醒：裝好登入成功就停手、先別玩掉額度；Antigravity 別用學校信箱；Gemini 學生方案值得申請一次，但要綁付款方式、一年後不取消會自動收費——不申請也能修完這門課。這些都不計分、不是入場門檻，沒做完也可以來上課。

This section has an official cap of 40 students. Come to the first class; adding and dropping, including instructor consent, follow the methods and deadlines announced by the Office of Academic Affairs—please check the official page linked under "How to enrol" above for the schedule and rules.本班官方限修 40 人。第一堂課先來上；加退選與加簽一律依教務處公告的方式與期限辦理——時程與規則請看上方「怎麼選課」連結的官方頁面。

You may learn and create with AI, but you must be able to explain your process, verify its output, and hold the integrity and safety limits. This course uses public or synthetic data only—no personal data, real grades, or unlicensed photos and audio.可以用 AI 學習與創作，但你要能解釋過程、會查證輸出、守誠信與安全底線。課堂只用公開或合成資料，不輸入個資、真實成績或未授權的照片與聲音。