60 minutes explores the potential dangers and possibilities created by recent dramatic advances in artificial intelligence.

60分鐘》探討了最近人工智能的巨大進步所帶來的潛在危險和可能性。

The A I market is expected to become a more than $42 billion industry this year.

預計今年A I市場將成為一個超過420億美元的產業。

According to the company pitchbook, an A I like chat GP T can learn or appear to learn based on the stuff that you feed into it.

根據該公司的宣傳手冊，我喜歡哈拉的GP T可以根據你輸入的東西來學習或看起來是在學習。

60 minutes correspondent Scott Pelley was given access to Google's campus in California and their A I lab in London to see their new technology and also meet with the CEO.

60分鐘》記者斯科特-佩利獲准進入谷歌在加州的園區和他們在倫敦的A I實驗室，觀看他們的新技術，還與首席執行官會面。

Here's a preview, how great a risk is the spread of disinformation A I will challenge that in a deeper way.

這裡有一個預告，虛假資訊的傳播有多大的風險A我將以更深的方式挑戰。

The scale of this problem is going to be much bigger, bigger problems.

這個問題的規模會更大，更大的問題。

Sundar Pacha says with fake news and fake images, we saw the potential with Google's experimental A I that creates pictures out of words.

桑達爾-帕查說，隨著假新聞和假影像的出現，我們看到了谷歌實驗性的A I的潛力，它可以從文字中創造出圖片。

This is text in pictures.

這是圖片中的文字。

Exactly.

正是如此。

Eli Collins is a vice president at Google.

Eli Collins是谷歌公司的副總裁。

You can generate a picture of pretty much most things you imagine.

你可以生成一幅幾乎是你想象中的大多數事物的圖片。

And I'm happy to put it in a prompt.

而且我很樂意把它放在一個提示中。

If you've got something, let's try pink taxi cabs on Fifth Avenue, right pink taxi cabs.

如果你有東西，讓我們試試第五大道上的粉紅色計程車，對粉紅色計程車。

Instantly we saw a shade unknown on New York cabs.

瞬間，我們看到了紐約計程車上未知的陰影。

Next Collins gave us a first look at an A I system.

接下來，柯林斯讓我們第一次看到了一個A I系統。

Google has not released yet experimental text to video which instantly answered our prompt.

谷歌還沒有發佈實驗性的文本轉視頻，它立即回答了我們的提示。

Golden Retriever with wings.

有翅膀的金毛獵犬。

Oh, look at that.

哦，看看這個。

As with Google's chat bot barred.

正如谷歌的哈拉機器人被禁止一樣。

This has safety filters.

這有安全過濾器。

For example, it doesn't create images of people.

例如，它不創造人的影像。

It will be possible with a to create, you know, a video easily where it could be Scott saying something or me saying something and we never said that and it could look accurate.

它將有可能創建，你知道，一個視頻很容易，它可能是斯科特說的東西或我說的東西，我們從來沒有說過，它可以看起來準確。

But you know, at a societal scale, you can cause a lot of harm.

但是你知道，在社會範圍內，你可以造成很大的傷害。

There are deeper risks people worry about, you know, which is at some point that humanity loses control of the technology it's developing.

人們擔心的還有更深層次的風險，你知道，那就是在某些時候，人類失去了對它所開發的技術的控制。

So those are some of the far out use cases which we need to think about as early as possible and get it right.

是以，這些是我們需要儘早考慮的一些遠期用例，並將其正確化。

You can watch Scott Pelley report on Sunday on 60 minutes right here on CBS.

你可以在這裡觀看斯科特-佩利週日在CBS的60分鐘節目中的報告。