Hi guys!
Welcome to another week’s newsletter!
Machine Learning is everywhere in our daily lives. It makes our lives more insightful and more personalized! With the enormous amount of data created by humans today, we rely on machine learning to analyze and make sense of the data and to create algorithms that optimize our experiences.
Let’s take YouTube videos for example,
Hundreds of hours of content are uploaded onto YouTube videos each minute. Recommendation uses deep neural networks to deliver optimized recommendations depending on your user history and context. Millions of videos narrow down to hundreds of candidates and ultimately to dozens of video recommendations relevant to users.
Another example is Facebook and its personalized ad generations,
Facebook generates personalized ads by studying user behavior using machine learning models. The 1.9 billion daily users’ video clicks, posts, likes/dislikes, and ad preferences on Facebook contain an incredible amount of data that greatly surpasses human’s ability to analyze them. Facebook uses Artificial intelligence and Machine learning as the automated solutions to process this big data; Machine learning provides a system that processes and learns from this data to produce targeted ads, posts, and videos that users may prefer.
Here are some 10 interesting stats of Facebook as of 2021 (for example Month Active Users MAU, or Daily Active Users DAU, approx hours spent per user, etc..) Check it out here.
https://www.oberlo.com/blog/facebook-statistics
Google search, Image recognition, Cybersecurity, Cloud computing, Cancer detection, Autonomous driving are all real applications of machine learning today. AI and Machine Learning are truly everywhere.
Moore’s law says that the number of transistors or computer capacity doubles every 2 years and cost will continue to reduce. Our technology has been growing exponentially under Moore’s law for more than 50 years since 1965. My first computer had no colors, no windows (it uses DOS mode). It was not user-friendly and it weighed more than a bag of rice. I remember my first colored monitor, it changed my life! Ask your parents what life was like without an iPhone, cable TV, Google search engine, or access to the internet!
Will Moore’s law end? It is debatable. Regardless if it ends or not, we are at the next phase of technological breakthroughs that are much bigger than the internet.
Artificial intelligence and Machine Learning.
The world today is more connected than ever! Thanks to Artificial intelligence and Machine Learning.
We have the 5G network (10Gbps vs 100mbps in 4G) that offers instant download with minimal latency.
We have the Starlink satellites (SpaceX’s launch of 12,000 satellites into lower orbits with 30,000 more on the way) provide internet to less urban areas.
We have the Microsoft Mesh, the cloud-based-mixed-reality virtual meeting that will connect people from many part of the world to meet in holographic virtual avatars.
To me, Machine learning is not magic, it is beyond magic! It is more fun and It is more mind-boggling! It is also already here, for better or worse. It will continue to transform our experiences of the world, whether we accept it or not.
Remember google’s DeepMind project AlphaGo that beat the best Go player? It had no problem beating humans learning from human played games.
After AlphaGo, there came the Alpha Zero project where AI learns from its own games! The zero actually means no human knowledge or interventions in the loop. Check out this quick video. It is amazing how machines can learn from themselves!
https://www.youtube.com/watch?v=7L2sUGcOgh0&ab_channel=DeepMind
Should we be afraid of AI?
Maybe.
“If AI has a goal and humanity just happens to be in the way, it will destroy humanity as a matter of course without even thinking about it…It’s just like, if we’re building a road and an anthill just happens to be in the way, we don’t hate ants, we’re just building a road.” Elon Musk
We will be in the technological singularity. The only way we can beat Artificial intelligence is to be in symbiosis with Artificial intelligence.
It is a mathematical fact:
AI+Human > AI
I hope you are as inspired by AI as I do.
Check out this video from the Robotics exhibition in Germany 2021.
You will see Robots enabled by Artificial intelligence & Machine Learning: DNA Robots to deliver drugs, robot mobile assistants, robots for mining oxygen on Mars, drones to avoid obstacles without pre-made maps, robotics that recognize and sort parts, robot companions etc.
One last item about the summer program!
One of my students at Harvey Mudd College (HMC) sent me an email about a 6-week course that runs from June 14th to July 23rd. The course is called “Math of Big Data” https://math189bigdata.github.io)
This course will cover the mathematics behind big data and machine learning. This will be a course over zoom. This course is project-based and can be taken by high school students. The Bonus: At the end of the course, students who do well may continue onto Machine Learning research with Professor Weiqing Gu at HMC. If you are interested, let me know. Here is the registration and application process for HMC courses. https://www.hmc.edu/summer-session/#ss-application-process
You can also email me if you have further questions.
Thank you and see you next week!
Mr. Daniel
嗨，大家好！
歡迎收看另一周的新聞通訊！
機器學習在我們的日常生活中無處不在。 它使我們的生活更有個性化和洞察力！ 在當今人類創造的海量數據的基礎上，我們依靠機器學習來分析和理解數據並創建可優化體驗的算法。
以YouTube視頻為例，
每分鐘將數百小時的內容上載到YouTube視頻中， 並且使用深度神經網絡根據用戶歷史和上下文提供給客戶最優質的推薦。 從數百萬個視頻中縮小到數百個用戶，最終縮小到數十個與用戶相關的視頻在進行推薦。
另一個例子是Facebook及其個性化的廣告世代。
Facebook通過使用機器學習模型研究用戶行為來生成個人個性化廣告。 Facebook上每天有19億用戶的視頻點擊，帖子，喜歡/不喜歡以及廣告偏好設置，其中包含的數據量是非常驚人的，大大超過了人類對其進行分析的能力。 Facebook使用人工智能和機器學習作為自動解決方案的方式來處理這些大數據。 機器學習提供了一個系統，該系統處理該數據並從中學習，以產生用戶可能喜歡的定向廣告，帖子和視頻。
以下是截至2021年Facebook的10項有趣的統計信息（例如，月活躍用戶MAU或每日活躍用戶DAU，每位用戶花費的大約小時數，等等。）你可以在此處查看相關訊息。
https://www.oberlo.com/blog/facebook-statistics
谷歌搜索，圖像識別，網絡安全，雲端計算，癌症檢測，自動駕駛都是當今機器學習的真正應用。 人工智能和機器學習在我們生活中無處不在。
摩爾定律說，晶體管或電腦的數量每兩年增加一倍，成本將繼續降低。 自1965年以來，根據摩爾定律，我們的技術指數增長，已超過50年。我的第一台電腦是沒有顏色，沒有window視窗系統（它使用DOS磁碟作業模式），使用上不順手，並且它的重量超過一袋大米， 但是記得我的第一台彩色顯示器，它改變了我的生活！ 詢問您的父母，如果沒有iPhone，有線電視，Google搜索引擎或無法訪問互聯網，他們的生活會如何！
摩爾定律會終止嗎？ 這值得商榷。 不管它是否結束，我們都處於下一階段的技術突破，這一突破要比互聯網大得多。
人工智能和機器學習。
今天的世界比以往任何時候都更加緊密！ 多虧了人工智能和機器學習。
我們擁有5G網絡（10Gbps與4G中的100mbps），可提供最低延遲縮短即時下載時間。
我們擁有Starlink衛星（SpaceX將12,000顆衛星發射到更低的軌道，還有30,000顆正在發射），可為較少的城市地區提供互聯網。
我們擁有Microsoft Mesh，這是一種基於雲端的混合現實虛擬會議，它可以把來自世界各地的人們聯繫起來使用全像影像一起開會。
對我來說，機器學習不是魔法，它已超越！ 而且更有趣，更令人感到驚喜！ 無論好壞，它都已經在這裡了。 無論我們是否接受，它將繼續改變世界和我們的經驗。
還記得谷歌的DeepMind AlphaGo擊敗最佳圍棋棋手嗎？ 毫無疑問，它從與人類玩遊戲中學習知識。
在AlphaGo之後，出現了Alpha Zero項目，在此AI可以從自己的遊戲中學到東西！ Zero實際上意味著沒有人類知識及人類也無法介入這項目中。請觀看此視頻— 機器如何向自己學習，這真是太神奇了！
https://www.youtube.com/watch?v=7L2sUGcOgh0&ab_channel=DeepMind
我們應該害怕人工智能嗎？
也許。
“如果人工智能有一個目標，而人類恰好妨礙它的前進，那麼它就會理所當然，毫不考慮地將其摧毀，……這就像，如果我們正在建造一條道路，而蟻丘恰好在 順便說一句，我們不討厭螞蟻，我們只是在築路。” 伊隆·馬斯克（Elon Musk）
我們將處於科技上的奇異點。 如果要我們擊敗人工智能的唯一方法那就是與人工智能共存。
這是一個數學事實：
AI +人類> AI
希望您像我一樣受到AI的啟發。
請觀看2021年德國機器人展上的這段視頻。
您將看到具有人工智能和機器學習功能的機器人：DNA機器人用於運送藥物，機器人移動助手，用於在火星上開採氧氣的機器人，無人駕駛飛機以避開沒有預先製作的地圖的障礙物，識別並分類零件的機器人技術，機器人同伴等。
關於暑期課程的最後一項！
我在哈維穆德學院（HMC）的一個學生給我發送了一封電子郵件，告知今年哈維穆德學院提供6月14日至7月23日的為期6週的暑期課程。 該課程稱為“大數據的數學”（https://math189bigdata.github.io）
本課程將涵蓋大數據和機器學習背後的數學，這門課將會讓學生拓展不同的學習領域。 本課程以專題研究項目為主，高中生可以參加。 優點是：課程結束後，表現出色的學生可以有機會與HMC的顧偉清教授繼續進行機器學習研究。 如果你感興趣，請告訴我。 以下是HMC課程的註冊和申請程序。 https://www.hmc.edu/summer-session/#ss-application-process
如果您還有其他問題，也可以給我發送電子郵件。
謝謝，下週見！
丹尼爾
