/agents
- debug
- code-review
- planner
- Generator
- Evaluator
玩電腦的,其實是被電腦玩的,所以,有人在機櫃上面放乖乖,而且不能放到過期,過期的話它就不乖了,你看,這是不是被電腦玩弄?所以,被電腦玩的都希望能諸事大吉!
FB看到的,留存~~
https://www.facebook.com/share/p/18k3xY6Eef/
一位麻省理工學院學生使用NotebookLM在48小時內掌握了整個領域
1. Most people use NotebookLM for summaries. Boring.
He dropped in textbooks, papers, and a full lecture series.
Then he ignored the facts and went for the "how."
“What mental frameworks do the top researchers in this field actually use to solve problems?”
1. 大多數人只用 NotebookLM 做摘要。這太無趣了。
他丟進了教科書、論文和完整的系列講座紀錄。
接著,他忽略了單純的事實資訊,直搗核心問「方法論」:
「這個領域的頂尖研究員,實際上是用什麼樣的思考框架來解決問題的?」
2. His next move: Finding the friction.
“Where do the leading voices here completely disagree? Give me the best evidence for both sides.”
In 20 minutes, he had a map of the intellectual battleground.
That kind of clarity usually takes a PhD student a year to build.
2. 他的下一步:尋找矛盾與衝突。
「這裡的領軍人物在哪些點上完全無法達成共識?請給我雙方最核心的論據。」
短短 20 分鐘內,他就理出了一張知識戰場的地圖。
這種程度的洞察力,通常需要博士生花上一整年才能建立。
3. Then the stress test:
“Write 10 questions that separate someone who understands this from someone just memorizing it.”
He spent 6 hours answering them. Every time he slipped up, he’d ask:
“Break down exactly why I’m wrong and what concept I’m missing.”
3. 接著是壓力測試:
「請出 10 個問題,用來區分出哪些人是真的『理解』,哪些人只是『死背』。」
他花了 6 個小時回答這些問題。每當他出錯,他就會追問:
「請精確地拆解我為什麼錯了,以及我遺漏了哪個關鍵概念?」
4.48 hours later, he was holding his own in a theory debate with his advisor.
It’s the same tool everyone has access to. The difference was the interrogation strategy.
He didn't treat it like a search bar; he treated it like an immortal professor who’s read everything.
4. 48 小時後,他已經能在理論辯論中與指導教授平起平坐。
工具是每個人都能使用的,差別在於提問策略(Interrogation Strategy)。
他沒有把 AI 當成搜尋列,而是把它當成一位讀過萬卷書、長生不死的博學教授。
5. A semester vs. 48 hours isn’t about working harder.
It’s about asking the right questions, in the right sequence, to the right "brain."
5. 一個學期 vs. 48 小時,這無關乎誰更努力。
這關乎於:在正確的序列下,向正確的「大腦」提出正確的問題。
我的觀察與建議
這份案例體現了 「第一原理思考」 在 AI 時代的威力。如果你也想嘗試這種學習方式,除了 NotebookLM,這種「蘇格拉底式提問法」在 Gemini 或 Claude 上同樣非常強大。
需要我幫你針對某個特定領域,設計一套類似的「48 小時深度提問清單」嗎?
MySQL/MariaDB 備份資料庫在還原時,若發生 FK 錯誤的狀況時,可以用以下的方式備份資料庫,在匯入還原時就可以正常匯入。
mysqldump --single-transaction --routines --triggers --disable-keys databaseName -uUserName -pPassword > backup.sql
C:\Users\wangjenmin\AppData\Roaming\Rime 資料夾裡的 default.custom.yaml 檔案,加入以下藍色文字,可以設定候選字有九個,也可以關閉 shift 的切換中英文的動作。
customization:
distribution_code_name: Weasel
distribution_version: 0.17.4
generator: "Rime::SwitcherSettings"
modified_time: "Sat Nov 29 11:12:31 2025"
rime_version: 1.13.1
patch:
"menu/page_size": 9
schema_list:
- {schema: scj6}
ascii_composer/switch_key: {
Caps_Lock: noop,
Control_L: noop,
Control_R: noop,
Eisu_toggle: clear,
Shift_L: noop,
Shift_R: noop
}
C:\Users\wangjenmin\AppData\Roaming\Rime 資料夾裡的 scj6.schema.yaml ,底下藍色的設定值改為 1 ,可以預設為 英文 模式。
# Rime schema settings
# encoding: utf-8
schema:
schema_id: scj6
name: 快速倉頡
version: "0.9"
author:
- 麥志洪 <makchehu@netvigator.com>
description: |
快速倉頡輸入法第六代
網址: http://www.scj2000.com
Copyright (C) 2002-2009 麥志洪 <makchehu@netvigator.com>
授權方式: GPL version 3
dependencies:
- luna_quanpin
switches:
- name: ascii_mode
reset: 1
states: [ 中文, 西文 ]
- name: full_shape
states: [ 半角, 全角 ]
- name: simplification
states: [ 漢字, 汉字 ]
- name: ascii_punct
states: [ 。,, ., ]
候選字會依據使用量大的時候往前移動,很麻煩,可以做以下設定
新建或修改 scj6.custom.yaml 檔案,內容如下
patch:
translator/enable_user_dict: false
參考以下內容
https://tools.wingzero.tw/article/sn/3591
這是 Vibe Coding 推廣者的說法
https://gizmodo.com/even-the-inventor-of-vibe-coding-says-vibe-coding-cant-cut-it-2000672821
在 Ubuntu 下使用 Docker 安裝 OpenVAS
參考來源: https://medium.com/@ma7moudsabra/install-gvm-vulnerability-scanner-on-ubuntu-24-04-using-containers-fd38935b7121
apt repository.# Add Docker's official GPG key:
sudo apt-get update
sudo apt-get install ca-certificates curl
sudo install -m 0755 -d /etc/apt/keyrings
sudo curl -fsSL https://download.docker.com/linux/ubuntu/gpg -o /etc/apt/keyrings/docker.asc
sudo chmod a+r /etc/apt/keyrings/docker.asc
# Add the repository to Apt sources:
echo \
"deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.asc] https://download.docker.com/linux/ubuntu \
$(. /etc/os-release && echo "$VERSION_CODENAME") stable" | \
sudo tee /etc/apt/sources.list.d/docker.list > /dev/null
sudo apt-get update如果是使用 Linux Mint 22.1 的環境,上面的 $VERSION_CODENAME 這裡會不正確,手動改為 noble 就可以了。
2. Install the Docker and dependencies packages.
sudo apt-get install gnupg docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin3. Verify that the installation is successful by running the hello-world image:
sudo docker run hello-world
4. Manage Docker as a non-root user
sudo groupadd docker
sudo usermod -aG docker $USER
newgrp docker5. Verify that you can run docker commands without sudo.
docker run hello-world6. For downloading the Greenbone Community Edition docker compose file, a destination directory should be created.
export DOWNLOAD_DIR=$HOME/greenbone-community-container && mkdir -p $DOWNLOAD_DIR7. Download the file
cd $DOWNLOAD_DIR && curl -f -L https://greenbone.github.io/docs/latest/_static/docker-compose-22.4.yml -o docker-compose.yml
8. To allow remote access to the Greenbone Web Interface, you need to modify the docker compose file to configure the web server (gsad) to listen on all network interfaces.
gsa:
image: greenbone/gsa:stable
restart: on-failure
ports:
- 127.0.0.1:9392:80 #before
volumes:
- gvmd_socket_vol:/run/gvmd
depends_on:
- gvmd
----
gsa:
image: greenbone/gsa:stable
restart: on-failure
ports:
- 9392:80 #After
volumes:
- gvmd_socket_vol:/run/gvmd
depends_on:
- gvmd9. Start the Greenbone Community Edition container.
docker compose -f $DOWNLOAD_DIR/docker-compose.yml -p greenbone-community-edition up -dOpenVAS is now installed and running on your Ubuntu 24.04 LTS system. To access the OpenVAS web interface:
1. Open a web browser and navigate to http://localhost:9392.
2. Log in using the default credentials: