And in today's video, we're going to unravel some complexities of artificial intelligence or AI.

在今天的視頻中，我們將揭開人工智能（AI）的神祕面紗。

We'll explore the different types of AI, everything from traditional AI to general AI, and clarify some common misconceptions along the way.

我們將探討從傳統人工智能到通用人工智能等不同類型的人工智能，並順便澄清一些常見的誤解。

So let's jump straight in.

那我們就直接開始吧。

First, let's talk about traditional AI.

首先，我們來談談傳統的人工智能。

This includes algorithms that you encounter every day, such as recommendation systems on Netflix or Amazon.

這包括你每天都會遇到的算法，如 Netflix 或亞馬遜的推薦系統。

These systems analyze your past behaviors to predict and recommend movies or products you might like.

這些系統分析你過去的行為，預測並推薦你可能喜歡的電影或產品。

They operate using straightforward data processing and matching algorithms, making predictions based on your previous interactions.

它們採用直接的數據處理和匹配算法，根據你以前的互動情況進行預測。

If we now step up the level of complexity, we get to supervised or imitation learning, which powers sophisticated systems like self-driving cars.

如果我們現在提高複雜度，就能實現監督學習或模仿學習，為自動駕駛汽車等複雜系統提供動力。

So supervised learning means training an AI with data that is already labeled, or learning from copying human behaviors like we see in Tesla cars that monitor what drivers do.

是以，監督學習意味著用已經標註的數據來訓練人工智能，或者通過模仿人類行為來學習，就像我們在特斯拉汽車上看到的那樣，它可以監控駕駛員的行為。

And self-driving cars learn to navigate real-world roads safely by analyzing vast amounts of labeled data and monitoring what people are doing, so they're able to react to elements like stop signs and pedestrians.

自動駕駛汽車通過分析大量標籤數據和監控人們的行為，學會在現實世界的道路上安全導航，從而能夠對停車標誌和行人等因素做出反應。

In contrast to supervised learning, we have unsupervised learning.

與監督式學習相反，我們還有無監督式學習。

This is an AI that doesn't rely on labeled data, it identifies patterns and relationships on its own, which is crucial for discovering hidden insights in data without any prior knowledge of what you're looking for.

這是一種不依賴於標記數據的人工智能，它能自行識別模式和關係，這對於在不預先知道要尋找什麼的情況下發現數據中隱藏的洞察力至關重要。

This capability is especially useful in market segmentation and anomaly detection.

這種功能在市場細分和異常檢測中尤其有用。

Another type of AI is reinforcement learning, which learns through trial and error, primarily using feedback from its own actions and experiences, rather than from explicit teaching.

另一種人工智能是強化學習，它主要利用自身行動和經驗的反饋，而不是明確的教學，通過嘗試和錯誤來學習。

In the context of self-driving cars, reinforcement learning is used in simulated environments where the system learns to make decisions by experiencing virtual scenarios, thus improving its algorithms before ever hitting the real road.

在自動駕駛汽車方面，強化學習用於模擬環境，系統通過體驗虛擬場景來學習決策，從而在真正上路之前改進算法。

Next, let's delve into generative AI, a truly exciting frontier of AI development.

接下來，讓我們深入探討生成式人工智能，這是人工智能發展的一個真正令人興奮的前沿領域。

Generative models like ChatGPT for text or MidJourney for images can create new, realistic text, images, sound or video.

文本生成模型 ChatGPT 或影像生成模型 MidJourney 可以創建新的、逼真的文本、影像、聲音或視頻。

They operate by predicting the next word or pixel based on their training data.

它們的工作原理是根據訓練數據預測下一個單詞或像素。

Although these tools generate impressive results, it's crucial to remember they do not truly understand the content they create, they simply generate predictions based on patterns.

雖然這些工具能產生令人印象深刻的結果，但必須記住，它們並不能真正理解它們所創建的內容，它們只是根據模式生成預測。

Finally, we reach something that doesn't yet exist, the concept of Artificial General Intelligence or AGI.

最後，我們將討論一個尚不存在的概念，即人工通用智能（AGI）。

AGI would be capable of performing any intellectual task that a human can, with the ability to understand and conceptualize the world on a human level.

人工智能將能夠完成人類所能完成的任何智力任務，並能以人類的水準理解和構思世界。

This would involve processing contextual nuances, abstract thinking, planning and even experiencing emotions, capabilities far beyond the reach of today's AI.

這將涉及處理上下文的細微差別、抽象思維、規劃甚至情感體驗，這些能力遠遠超出了當今人工智能的能力範圍。

So while today's AIs can perform tasks from driving cars to writing articles, it lacks a deep understanding of the world.

是以，儘管今天的人工智能可以執行從駕駛汽車到撰寫文章等任務，但它缺乏對世界的深刻理解。

AGI however remains a theoretical leap that would fundamentally change how machines interact with their environment and with us.

然而，AGI 仍然是一個理論上的飛躍，它將從根本上改變機器與環境和我們的互動方式。

Thank you for tuning in.

感謝您的收聽。

If you are fascinated by the evolution of AI and want to stay updated on the journey towards AGI, don't forget to like, subscribe and hit the notification bell.

如果您對人工智能的發展非常著迷，並希望隨時瞭解通往 AGI 的最新進展，請不要忘記點贊、訂閱並按下通知鈴。

And of course check out my book Genitive AI in Practice.

當然，還可以看看我的新書《基因人工智能實踐》（Genitive AI in Practice）。

I will see you in my next video where we continue to explore the transformative technologies that are shaping our future.

我們將在下一期視頻中繼續探討塑造未來的變革性技術。