So this is simultaneously really impressive and really frightening at the same time.

這同時給人留下了深刻的印象，也讓人感到非常害怕。

And it's hitting me in ways that I didn't really expect.

它以我始料未及的方式衝擊著我。

So, do you remember Will Smith eating spaghetti?

你還記得威爾史密斯吃意大利麵的影片嗎？

Do you remember when this was what AI generated videos look like?

你還記得 AI 生成的影片是這個樣子的嗎？

Remember when we said, OK, this AI stuff is cool and all but clearly, there's a long way to go before there's any need for concern.

還記得我們說過：「好吧，這個 AI 的東西很酷，但很明顯，要想讓人放心，還有很長的路要走。」

Well, welcome to the future, people, because this is also an AI-generated video.

歡迎來到未來，因為這也是一部由 AI 製作的影片。

And so is this. Completely synthesized out of thin air by computers.

還有這個。完全是由電腦憑空合成的。

This one too.

這個也是。

This is not real. Absolutely ridiculous, how far we've come in literally one year.

這不是真的。非常荒謬，就在這一年居然進步了這麼多。

This does feel like another ChatGPT DALL-E moment for AI and maybe I'm overreacting because OK, I'm a video creator.

這確實感覺像是另一個 ChatGPT DALL-E 時刻，可能是因為我是一個影片創作者，我有點過於激動。

So an AI that's actually doing my job, maybe that feels a little more threatening.

因此，如果 AI 真的在做我的工作，也許會讓人感覺更具威脅性。

So I'm particularly impressed by it.

所以我對它印象特別深刻。

But also this stuff is really good.

不過，這個，這個東西真的很不錯。

So today, Sam Altman and OpenAI announced a new model called Sora and it can generate full up to one-minute video clips from just text input.

今天，Sam Altman 和 Open AI 發佈了一款名為 Sora 的新模型，它只需輸入文字，就能生成長達一分鐘的完整影片。

So the same way DALL-E was able to understand our text input and turn it into a photo realistic or stylized image or whatever you want.

與 DALL-E 相同，能夠理解我們的文字輸入，並將其轉換為逼真的或風格化的圖像或任何你想要的圖像。

Same thing with Sora.

Sora 也是一樣。

But now since it's videos, it also needs to "understand" how all these things like reflections and textures and materials and physics all interact with each other over time to make a reasonable looking video.

但現在既然是影片，它還需要「理解」反射、紋理、材料和物理等所有這些東西如何隨著時間的推移相互作用，以製作出看起來合理的影片。

And of course, right away, there's a bunch of examples on their website that are crazy.

當然，他們的網站上馬上就有一堆瘋狂的例子。

Now, before I show you these, I just need you to keep this in mind.

現在，在我向你們展示這些之前，我只需要你們記住這一點。

You're about to watch a bunch of AI-generated videos and you know that you're about to watch a bunch of AI-generated content.

你即將看到的是一堆由 AI 生成的影片，你知道你即將看到的是一堆由 AI 生成的內容。

So your brain, you're already looking for this stuff and it's not perfect.

所以，你的大腦已經在尋找這些東西，但它並不完美。

You will find imperfections but not everybody who sees AI-generated content on the internet knows to be looking for that.

你會發現不完美之處，但並不是每個在網上看到 AI 生成的內容的人都知道要尋找不完美之處。

So also keep that in mind.

所以也要牢記這一點。

This is also the worst that this technology is going to be from here on now.

這也是這項技術今後最糟糕的情況。

So, OK, here's one of the videos. There's no audio to any of these clips, but the prompt for this one is a stylish woman walks down a Tokyo street filled with warm glowing neon and animated city signage.

好了，這是其中一段影片，這些片段都沒有音頻，但這段影片的指令是，一位時尚女性走在東京的街道上，街道上到處都是溫暖的霓虹燈和動畫城市標誌。

She wears a black leather jacket, a long red dress and black boots.

她身穿黑色皮夾克、紅色長裙和黑色靴子。

This video is already miles ahead of where we were.

這段影片已經遠遠超出了我們的預期。

It has accurate lighting, it has materials, it has skin tones, movements, even has reflections all over the place.

它有準確的照明，有材質，有膚色，動作甚至到處都有反光。

Now, of course, if you look at it for more than about 10 seconds very closely, there are lots of giveaways like this dude in the background kind of looks like he's gliding in a weird way.

當然，如果你仔細觀察超過 10 秒鐘，就會發現很多蛛絲馬跡，比如背景中的這個人看起來就像是在以一種奇怪的方式滑行。

The frame rates and the reflections in the water are for some reason lower than the rest of the video.

由於某些原因，幀頻和水中的倒影比影片的其他部分要低。

The camera movement overall is just a bit inconsistent and it just, I don't know, it just kind of feels a little bit off, but then again, and this is where we were one year ago.

鏡頭的移動整體上有點不連貫，我不知道，就是感覺有點不對勁，不過話說回來，一年前我們的效果是這樣。

So just keep that in the back of your head for all this.

所以，你要時刻牢記這一點。

OK. How about this one?

好的。這個怎麼樣？

This is another one which has a long prompt about a camera following behind a white vintage SUV with a black roof rack as it speeds up a steep dirt road.

這是另一部有很長指令的作品，講述的是一架攝影機跟在一輛白色復古越野車後面，帶著黑色車頂行李架，在陡峭的土路上飛馳。

This is also, again, really good.

這也同樣非常好。

It kind of looks a little more video gamey because of how rock solid the drone footage is but clearly very usable.

由於無人機拍攝的畫面堅如磐石，它看起來有點像電玩，但顯然非常實用。

Here's another one, a litter of golden retriever puppies playing in the snow, their heads pop in and out of the snow covered in it.

這是另一張照片，一窩金毛獵犬小狗在雪地裡玩耍，它們的頭在雪地裡忽隱忽現。

It's so good.

太厲害了。

It feels like the physics of the fur and the ears and everything with the snow flying around in slow motion is incredible.

感覺毛皮、耳朵和雪花在慢動作中飛舞的物理效果非常不可思議。

I've looked through all the sample videos on OpenAI's website and clearly these are the hand picked best ones that they chose to share where they just put in some text and then get a video and don't modify it.

我瀏覽了 OpenAI 網站上的所有示例影片，很明顯，這些都是他們精心挑選的最佳影片，他們選擇分享，只是輸入一些文字，然後得到影片，而不對其進行修改。

But there's really impressive stuff in there.

但裡面的內容確實令人印象深刻。

Some of it has humans, some of it doesn't, some of it is more realistic feeling like the truck driving one.

其中有些是人類，有些不是，有些則更有真實感，比如駕駛卡車的。

But some of them are more video gamey or more stylized.

但其中一些更像電玩，或者更風格化。

A lot of it is slow motion.

很多都是慢動作。

I just have to say how insanely fast these models are improving is genuinely like, that's the shocking part.

我只想說，這些模型的進步速度真是快得令人難以置信，這才是令人震驚的地方。

Like I remember not even that many months ago, DALL-E 3, really, really high end and you could always still find something off about it.

我記得就在幾個月前， 已經非常高端的 DALL-E 3 還是能發現它有什麼不對勁的地方。

Like especially if you ask it for something like a photo realistic image of a human, something about like the hands or the ears would always just be a little bit off, never mind the physics.

特別是如果你要求它生生成像照片般真實的人類圖像之類的東西，像手或耳朵這樣的東西總是會有點偏離，更不用說物理了。

But even this video here is crazy at first glance. The prompt for this AI-generated video is a young man in his twenties is sitting on a piece of a cloud in the sky reading a book.

但即使是這個影片，乍一看也很瘋狂，這個由 AI 生成的影片的指令是，一個二十多歲的年輕人正坐在天空中的一片雲上看書。

This one feels like 90% of the way there for me.

對我來說，這個項目已經完成了 90%。

Like it's beyond the Uncanny Valley of like Apple's Personas which are actually based on humans.

就像它超越了蘋果公司的人物角色的恐怖谷理論一樣，實際上是基於人類的。

This is a made-up person.

這是一個虛構的人。

I mean, his eyes are kind of weird and the motion of the pages in the book are kinda odd and yeah, obviously he's in a cloud and that's a giveaway, but like, the lighting and the shadows and the skin tones and then all the realism of the textures on the shirt and the way the shirt and the pants move and the hair, they're all really impressivee.

我的意思是，他的眼睛有點奇怪，書頁的運動也有點奇怪，是的，很明顯他是在雲裡，這是個線索，但是像燈光、陰影、膚色，還有襯衫上的紋理、襯衫和褲子的運動方式以及頭髮，都非常逼真。

And then for this one, they typed in a movie trailer featuring the adventurers of the 30-year-old spaceman wearing a red wool-knitted motorcycle helmet.

然後，這個，指令是生成一個電影預告片，主角是一位戴著紅色羊毛針織摩托車頭盔的 30 歲太空人的冒險家。

Blue sky, salt desert, cinematic style, shot on 35 millimeter film and the close-ups of his face, the fabrics on the helmet, the film grain through every shot and the cinematic style, this is one of the most convincing AI-generated videos I've ever seen.

藍天，鹽漠，電影風格，35毫米膠片拍攝，臉部特寫，頭盔上的布料，每個鏡頭的膠片顆粒和電影風格，這是我見過最有說服力的AI生成影片之一。

Minus maybe the weird physics of that dude walking kind of in fast motion.

也許是因為那人走路的姿勢有點快，所以物理效果有點怪。

So Sam Altman, if you follow him on Twitter, he's going through a whole bunch more of like people's requests and posting a bunch more generated videos.

Sam Altman，如果你在推特上關注他的話，他正在處理一大堆像人們一樣的請求，並發布更多生成的影片。

And so if you want to check out his profile, you can see those.

如果你想查看他的個人檔案，就可以看到這些內容。

But here's the thing about these AI-generated videos now as good as they've gotten to this point.

但這些 AI 產生的影片現在已經達到了這樣的程度。

They can and will pass as real videos to people who are not looking for AI-generated videos.

他們可以並將作為真實影片傳遞給那些沒有特別要看 AI 影片的人。

Now that is obviously insanely sketchy during an election year in the US and also terrifying for a bunch of other internet related reasons.

在美國大選年，這顯然是極其粗略的，而且由於其他一些與網路相關的原因，這也令人恐懼。

But it's also perfect for stock footage.

但它也非常適合拍攝素材。

Like there are already all kinds of presentations and advertisements and then powerpoints that are in need of oddly specific stock videos and these AI-generated videos are already good enough to 100% pass for that purpose.

就像已經有各種各樣的簡報和廣告，然後是需要奇怪的特定庫存影片的幻燈片，而這些 AI 生成的影片已經足夠好，可以 100% 通過該目的。

Like look at this one, this one with the waves at Big Sir, this drone shot.

就像看看這個，這個是 Big Sir 的波浪，這個無人機拍攝的。

Honestly, if I saw this on Twitter, I wouldn't even think twice.

老實說，如果我在 Twitter 上看到這張照片，我根本不會多想。

I'd be like, "Oh nice drone shot, dude."

我會想說：「哦，無人機拍得不錯。」

Wouldn't even think about AI if I wasn't pixel peeping at like the way the water was moving like this, this is a totally usable video in an ad for some California-based product.

如果我不是像這樣仔細檢查水流動的方式，我甚至不會想到 AI，這是加州某些產品廣告中的一個完全可用的影片。

And that has all sorts of implications for the drone pilot that no longer needs to be hired for all the photographers and videographers whose footage no longer needs to licensed to show up in that ad that's being made.

這對無人機駕駛員和所有攝影師和攝像師都會產生各種影響，因為他們不再需要獲得許可就能將拍攝的畫面出現在正在製作的廣告中。

It's already that good.

已經這麼好了。

There's other stuff like this wall of TVs, which would be a totally expensive and difficult thing to shoot with a camera and all these old expensive props.

還有其他東西，比如這面電視牆，如果用相機和這些昂貴的舊道具來拍攝，那將是一件既昂貴又困難的事情。

But if you can just generate it, this well with reflections and the environment and everything else around it, I mean, why do it any other way?

但是，如果你能通過反射、環境和周圍的一切來生成這麼好的效果，我的意思是，為什麼還要用其他方法呢？

It's also very capable of historical themed footage.

它還能拍攝歷史主題鏡頭。

So this is supposed to be California during the gold rush.

這裡應該是淘金熱時期的加州。

It's AI-generated but it could totally pass for the opening scene in an old western with the right music over it.

它是由 AI 生成的，但如果配上合適的音樂，完全可以作為老式西部片的開場戲。

How long until an entire ad, every single shot is completely generated with AI or what about an entire YouTube video or an entire movie?

整個廣告的每一個鏡頭都完全由 AI 生成，或者整個 YouTube 影片或整部電影的每一個鏡頭都由 AI 生成，這種情況還要多久就能達成？

I'm tempted to say like, we're a long way away from that because, you know, this still has flaws clearly and there's no sound and there's a long way to go with the prompt, engineering to iron these things out.

我很想說，我們離這個目標還有很長的路要走，因為，你知道，它顯然還存在缺陷，沒有聲音，要解決這些問題，還有很長的路要走。

But then again, the spaghetti was like a year ago.

不過話又說回來，威爾史密斯吃意大利麵就在一年前而已。

Now actually like that OpenAI on their website, they show some of the downfalls too of this particular model and because who would know better than the people who have been using it.

現在其實 OpenAI 在其網站上展示了這個特定模型的一些缺點，因為誰會比使用它的人更清楚。

This is a very private tool, by the way, right now, it's in super limited access.

順便說一下，這是一個非常私人的工具，目前訪問權限超級有限。

So it's in the hands of red teamers, which basically means people testing it, pushing the limits, trying to break it and a few trusted creators.

所以它掌握在紅隊的手中，基本上是指測試它、挑戰極限、嘗試破壞它的人，以及一些值得信賴的創作者。

But they have found plenty of weird edge stuff like this clip here of a bunch of Gray Wolf pups. Looks normal at first, but then it's pretty clear that something's kind off with the way they're just kind of appearing out of nowhere and walking through each other.

但是他們發現了很多奇怪的邊緣情況，像是這個一群灰狼幼崽的片段。 一開始看起來正常，但後來很明顯有點奇怪，它們似乎是從無中出現，而且彼此之間還穿過。

That's kind of weird.

這有點奇怪。

Or this clip of a guy running on a treadmill, which I mean, I don't really have to say much more about why this one's weird.

或是這個人在跑步機上的片段，我想我應該不用多說為什麼這個奇怪，但這是我最喜歡的一個。

But this is my favorite one. Again.

接下來這個是我最喜歡的。再說一次。

So again, just try to try to put yourself in the mind of someone who's not expecting AI. You're just scrolling through Facebook or Twitter or something, right?

再說一次，試著把自己當作沒有預期會看到 AI 的人。 你只是在滑 Facebook 或 Twitter 之類的。

So you just see this video.

你剛好看到這部影片。

So, first, I just want you to watch this clip as if it's just a stock video you found of a grandma celebrating her birthday and just try to try to think like, "I wonder what birthday she's celebrating."

首先，我只是想讓你看這個片段，就像是你找到的一個素材影片，展示一個慶祝生日的奶奶。 試著思考一下：「我想知道她在慶祝什麼生日。」

Right?

對吧？

I don't know. How old do you think she is?

我不知道。你覺得她多大了？

60? 65?

60？65？

Maybe it's the big 70. She seems to really like that cake.

也許是 70 大壽。她似乎真的很喜歡那個蛋糕。

Now, did you see it?

現在，你看到了嗎？

Did you catch that?

你發現了嗎？

I'm gonna play it again.

我要再播一次。

But this time watch the video knowing that AI-generated photos and videos have trouble accurately doing hands.

但這次觀看影片時要知道，AI 生成的照片和影片很難準確地做出手部動作。

I'll play it again, and now it feels super obvious like every time you watch it, watch a different set of hands, it gets weirder and weirder.

我再播一遍，現在感覺超級明顯，每看一次就看不同雙手，越看越奇怪。

You can watch it like five times and there's dead giveaway after dead giveaway.

你可以看上五遍，但還是會發現一個又一個的蛛絲馬跡。

Not even mentioning the weird inconsistencies with the direction of the wind on the candles.

更不用說蠟燭的風向奇怪的不一致。

But even as I'm saying, all that, even as it's coming out of my mouth, I can't help but remember that 12 months ago, we were critiquing this.

但即使我在說這一切，即使它從我嘴裡說出來，我也忍不住記得12個月前，我們正在批評這個。

So what does this all mean?

那麼，這一切意味著什麼呢？

Well, I mean, there's what it means now and there's what it means for the future.

嗯，我的意思是，有現在的意義，也有對未來的意義。

Now, Sora this thing that they've made is clearly a really impressive video generation AI tool that is both going to fool people and also be very useful.

Sora 制作的這個東西顯然是一個非常了不起的影片生成工具，既能騙人，又非常有用。

There's also a watermark in the bottom corner of every video generated by it.

它生成的每個影片的底角還有一個浮水印。

So if you see one of those videos and ideally it hasn't been cropped out, then that's at least a pretty clear indicator that it's AI-generated.

因此，如果你看到其中一個影片，並且理想情況下它沒有被裁剪，那麼這至少可以清楚地表明它是 AI 生成的。

It's a Sora video.

這是 Sora 的影片。

But I also do think they're going to have to be very careful with this.

但我也認為，他們必須非常謹慎地處理這件事。

They're gonna have a whole bunch of safety stuff to keep in mind.

他們會有一大堆安全問題需要注意。

I think they'll probably have to be even more safe than DALL-E.

我想他們可能要比 DALL-E 更加安全。

Like you shouldn't be able to generate people's likenesses.

就像你不應該生成別人的肖像一樣。

Like you shouldn't be able to make a politician look like they're doing something on video, especially this year, You probably won't be able to make Will Smith eating spaghetti.

就像你不應該在影片中讓政客看起來像在做什麼，尤其是今年，你可能無法讓威爾史密斯吃意大利麵。

But it also definitely means stock video generation is absolutely going to take a dent out of video licensing.

但這也絕對意味著，素材影像的生成絕對會對影片授權造成影響。

Like I can basically guarantee that.

我基本上可以保證這件事。

Like, logistically, why would anyone making something pay for footage of a house in the cliffs when they can generate one for free or for a small subscription price?

從邏輯上講，既然可以免費或以低廉的訂閱價格生成懸崖上的房子，為什麼還要花錢購買影像呢？

Like that is the real scary part of what this tool implies.

這才是這個工具真正可怕的地方。

But in the future, it gets pretty existential, man.

但在未來，它變得非常有存在感。

I mean, OK, if this is trained on all videos that have ever been made by humans, then surely it can't be innovative or creative in ways that humans haven't already been, right?

我的意思是，好吧，如果這是在人類製作過的所有影片基礎上訓練出來的，那麼它肯定不會有人類沒有過的創新或創意，對吧。

I don't know, either way, I'll have all the links below for all the Sora stuff, for OpenAI stuff.

我不知道，無論哪種方式，我都會在下面提供所有 Sora 內容、OpenAI 內容的所有連結。

And I guess I'll talk to you next year when we look back and go, "Remember that first version of Sora and how bad those wolf pups looked when they spawned out of nowhere?"

我想明年當我們回顧過去時我會和你談談：「還記得 Sora 的第一個版本，以及那些不知從何而來的狼崽們看起來有多糟糕嗎？」

Just remember, this is the worst that this technology is going to be from here on out.

請記住，從現在開始，目前這是這項技術最糟糕的情況，情況只會越來越好。

Thanks for watching.

感謝觀看。

Catch you in the next one.

下支影片見。

Peace.

掰。