A lot of you probably have already seen videos about it, but I want to bring you some new and interesting use cases and value propositions that you might not see in other places.

你們中的很多人可能已經看過相關視頻，但我想給你們帶來一些新的、有趣的使用案例和價值主張，這些可能是你們在其他地方看不到的。

So for those of you who don't know, this is Gen 3, which is produced by RunwayML up here.

對於不瞭解的人來說，這是 RunwayML 生產的第 3 代產品。

Now, RunwayML is special in the AI video space because they were the first ones to make the actual first commercial video generation model.

現在，RunwayML 在人工智能視頻領域很特別，因為他們是第一個製作出第一個實際商業視頻生成模型的人。

And now they have Gen 3 Alpha, so the third iteration.

現在他們有了第三代 Alpha，所以是第三次迭代。

It's a step towards building general world models.

這是朝著建立通用世界模型邁出的一步。

And yeah, right off the bat, you can see this thing is damn impressive.

是的，從一開始，你就能看出這東西非常了不起。

I mean, really, really good.

我是說，真的，真的很棒。

Edges close to Sora.

靠近索拉。

I think in the motion department, it does struggle a tad.

我認為，在運動方面，它確實有些吃力。

But I mean, all of these examples are really, really quite impressive.

但我的意思是，所有這些例子都非常、非常令人印象深刻。

Take a look at this one, for example.

例如，看看這個。

I think it doesn't have necessarily the same fidelity that Sora has, but it looks really, really cool and the motion is super impressive.

我認為它的逼真度未必能和《索拉》相比，但它看起來真的非常酷，動作也令人印象深刻。

Or even something like this.

或者甚至像這樣

This has a little bit better detail in my mind, and it's really cool to see how temporally consistent the buildings are as they pass by the camera.

在我看來，這幅畫的細節更好一些，而且當建築物從鏡頭前經過時，它們在時間上的一致性也非常好。

And I mean, you can do very special effect-esque style things as well.

我的意思是，你也可以做一些特效式的東西。

The possibilities, as always with AI video, are truly endless.

人工智能視頻的可能性永遠是無窮無盡的。

So apparently, it's been trained with highly descriptive, temporally dense captions, enabling imaginative transitions like, you know, a street being flooded with water.

是以，很顯然，它經過了高度描述性、時間密集型字幕的訓練，能夠實現富有想象力的過渡，比如，你知道，一條街道被水淹沒了。

Not something you're going to see every day.

這可不是每天都能看到的。

Or, you know, we have like this drone shot that moves through a castle.

或者，你知道，我們有像這樣的無人機鏡頭，穿過城堡。

This looks pretty cool.

這看起來很酷。

Looks like real GoPro footage.

看起來像真的 GoPro 錄像。

I mean, that is really impressive.

我的意思是，這真的令人印象深刻。

That almost looks legit.

這看起來幾乎是合法的。

Like, I would believe that that is an actual video, not some AI generated thing.

比如，我會相信那是一段真實的視頻，而不是人工智能生成的東西。

Here we can see some water glistening on a window through a train as it passes by.

在這裡，我們可以透過火車看到窗外閃爍的水光。

Again, wouldn't really be able to tell right off the bat that that is AI.

同樣，我們也無法一眼看出那就是人工智能。

Here is another one that people found really, really impressive.

這是另一個讓人印象深刻的例子。

This is clearly, you know, some more GoPro footage of someone walking through some tunnels, and then it pans over and you see a runway.

很明顯，這是更多的 GoPro 錄像，顯示有人走過一些隧道，然後鏡頭一轉，你會看到一條跑道。

Look at the effect of the flashlight as it moves over the graffiti on the runway.

看看手電筒在跑道塗鴉上移動時的效果。

I mean, you can tell it's a different texture there.

我的意思是，你可以看出那裡的質地不同。

It's really impressive stuff.

這真是令人印象深刻的東西。

I mean, no doubt in my mind, all of these examples blow me away, and I cannot wait until we get access.

我的意思是，毫無疑問，所有這些例子都讓我大開眼界，我迫不及待地想知道我們能不能進入。

That's the thing.

這就是問題所在。

We still don't have access to this, but it appears that access is going to be coming to Gen 3 very, very soon.

我們仍然無法訪問，但似乎很快就能訪問第 3 代。

And I can imagine people are going to be willing to pay a lot for access to this thing because it is the closest thing to Sora that we've seen so far in terms of like prompt following, coherency, and temporal stability.

我可以想象，人們會願意花大價錢來購買它，因為它是迄今為止我們所見過的最接近《索拉》的東西，比如及時跟蹤、連貫性和時間穩定性。

It is something else.

這是另一回事。

Photorealistic humans, that's obviously a really big deal when you're trying to tell stories.

逼真的人類，對於想要講述故事的人來說顯然是件大事。

Humans are most often a part of a lot of stories that are told in filmmaking, TV shows, etc.

在電影、電視劇等的許多故事中，人類往往是故事的一部分。

So there's some really great-looking humans in here, and obviously that is some bias in the training data there.

是以，這裡有一些非常漂亮的人類，很明顯，訓練數據中存在一些偏差。

They purposefully set out to make sure that it's going to be able to produce some pretty realistic-looking people.

他們有目的地確保它能夠製作出一些外形非常逼真的人。

And they absolutely have done that.

他們絕對做到了。

I mean, look at how cinematic this scene looks, and look at the glare of the sun and everything.

我的意思是，看看這個場景看起來多有電影感，再看看刺眼的陽光和一切。

So impressive.

令人印象深刻。

One thing I will say that I am seeing a lot is pretty much every one of these videos, no matter what it is, if it's a person or like this over here, everything looks like it's being shot in slow motion, which isn't necessarily a downside because, think about it, you can always speed up a video.

我想說的是，我經常看到的一件事是，幾乎每一個視頻，不管是什麼，如果是一個人，或者像這裡這樣，一切看起來都像是慢動作拍攝的，這並不一定是一個缺點，因為想想看，你總是可以加快視頻的速度。

If for some reason it produces videos that look like they're in slow motion, well, we can always just speed up the video.

如果由於某種原因，它生成的視頻看起來像慢動作，那麼我們可以隨時將視頻加速。

Like, that totally looks like it's in slow motion, no?

這看起來就像慢動作，不是嗎？

Maybe they purposefully trained it on slow motion video?

也許他們是故意用慢動作視頻來訓練的？

Like, we could just download these videos and speed them up and they would look completely normal.

比如，我們可以下載這些視頻，然後將它們加速，它們看起來就完全正常了。

I mean, shockingly realistic, but I just find that really intriguing.

我的意思是，現實得令人震驚，但我覺得這真的很有趣。

I mean, there we have like a monster walking around.

我的意思是，我們有一個怪物在走動。

That looks like a shot from a movie legitimately.

看起來真像電影裡的鏡頭。

I mean, it's so impressive.

我是說，它太令人印象深刻了。

The way that the light bounces and reflects off of the fur, it is so darn cool.

光線在毛皮上反彈和反射的方式，真是酷斃了。

Here's like some, I don't know, rock person walking through the woods.

這裡就像一個，我不知道，在樹林裡行走的石頭人。

Like, this thing has an understanding of physics and the world in a way that reminds me of the Sora demos.

比如，這東西對物理和世界的理解，讓我想起了索拉的演示。

This is a true competitor to Sora, I think, more so than anything else we've seen.

我認為，這是索拉的真正競爭對手，比我們看到的其他任何東西都更勝一籌。

And that's really, I think, what is causing all of the stir around Gen 3.

我認為，這才是引起第 3 代轟動的真正原因。

I keep seeing people talk about this.

我一直看到有人在談論這個問題。

I mean, yeah, check that out.

我的意思是，是的，看看這個。

I mean, that's some, that's a pretty good understanding of physics and paper and how it's going to interact in this alleyway with the wind tunnel.

我是說，你對物理和紙張的理解還不錯，知道它在巷子裡和風洞是如何相互作用的。

It's super cool.

太酷了

Here's another one that again looks like it's in slow motion.

這是另一張看起來又像是慢動作的照片。

I mean, why does it look like it's so in slow motion?

我的意思是，為什麼看起來像慢動作？

I don't get that part about it.

我不明白這一點。

But the way he like moves his thumb and everything, that just looks so legit.

但他移動拇指和其他動作的方式，看起來非常合理。

And these are obviously going to be cherry picked here.

很明顯，這些都是經過挑選的。

Detailed close-up of bacteria.

細菌的詳細特寫。

That's so cool.

太酷了

I mean, that looks so legit, man.

我的意思是，這看起來太合法了，夥計。

That's just some crazy, crazy concepts.

這只是一些瘋狂、瘋狂的概念。

It's exciting.

太激動人心了

It is exciting, no doubt.

這無疑令人興奮。

See different styles.

查看不同風格。

Here's like an anime, animated film art style.

這就像卡通片的藝術風格。

Really cool stuff.

真的很酷。

They've really gotten close to solving motion at this point.

在這一點上，他們確實已經接近解決運動問題了。

I mean, it's just, again, here's another one that looks like it's in slow motion.

我是說，這又是一個看起來像慢動作的畫面。

Super cool though.

不過超級酷。

I mean, what would you guys create if you had access to this?

我的意思是，如果你們能使用它，你們會創造出什麼？

It seems like access is going to be coming.

看來訪問即將到來。

At least it was hinted in like a few days.

至少在幾天內就有了暗示。

We're going to be able to get access to this.

我們將能夠訪問這個。

So I'd love to do like a live stream, testing it out, hanging out with you guys, doing some prompts.

所以我很想做一次直播，測試一下，和你們一起玩，做一些提示。

I don't mind paying for credits to use something like this because it's groundbreaking technology.

我不介意為使用這樣的東西支付點數，因為這是一項突破性技術。

Like I said, biggest Sora competitor that we've seen so far, at least in my mind.

就像我說的，至少在我看來，它是目前我們看到的索拉最大的競爭對手。

Anyways, that's sort of the TLDR on Gen 3.

總之，這就是第 3 代的簡要介紹。

Now let's get into some of the more interesting use cases.

現在，讓我們來了解一些更有趣的使用案例。

Some of the bigger value propositions that you might not see in other places.

一些更大的價值主張，你可能在其他地方看不到。

So Cristobal here on Twitter brings up that there's a lot of unexplored latent horror to discover.

是以，推特上的克里斯托巴爾提出，還有很多未被髮掘的潛在恐怖有待發現。

So the applications for horror movies or horror genres are very, very much already alive, as we can see by this terrifying generation.

是以，恐怖電影或恐怖流派的應用已經非常、非常活躍，我們可以從這一代人的恐怖經歷中看到這一點。

He does a few other generations.

他還做了其他幾代人的工作。

Like we've got a monkey here who is learning to play guitar on the street.

就像我們這裡有一隻在街上學彈吉他的猴子。

And again, it's just a really nice smooth shot.

同樣，這也是一個非常漂亮流暢的鏡頭。

Looks like it's in slow motion.

看起來像是慢動作。

I don't know what it is with runway and the slow-mo stuff, but it's super weird.

我不知道 "跑道 "和慢鏡頭是怎麼一回事，但真的很奇怪。

Oh yeah, he works for runway, so that's why he's got access to it.

哦，對了，他為 Runway 工作，所以才能接觸到它。

And you can see drink water over here.

你可以在這裡看到飲用水。

So the text is also really impressive.

是以，文本也確實令人印象深刻。

It's able to do like these animated things with text popping up on the screen.

它能在螢幕上彈出文字等動畫效果。

I mean, that's really darn cool.

我的意思是，這真的很酷。

Some more examples of this.

還有一些這方面的例子。

Look at the reflections on the stones, though.

看看石頭上的倒影吧。

That's not something that is easy to replicate in a 3D animation environment.

這在三維動畫環境中並不容易複製。

That's really, really crazy.

這真的是太瘋狂了

And that's one of like the impressive things about AI is that those more complex things that you would normally do in 3D animation are actually a lot easier in AI.

人工智能令人印象深刻的一點是，通常在三維動畫中完成的那些更復雜的工作，在人工智能中其實要容易得多。

And the simple stuff in animation is the most difficult part in the AI.

而動畫中最簡單的部分正是人工智能中最難的部分。

And here we go.

我們開始吧。

We've got like some animation with snow in the mountains and text comes through and says, ice, ice, baby.

我們有一些卡通片，山上下雪了，文字就會出現，說 "冰，冰，寶貝"。

And yeah, I mean, he is really, really showing off the fact that you can produce really good looking text.

是的，我的意思是，他真的是在炫耀你可以寫出非常漂亮的文字。

Can't sleep coming down there in sand.

在沙子裡睡不著。

And here we have another example where it just says, Timmy.

這裡還有一個例子，上面只寫著 "提米"。

And it's like this blasting dirt flying into view.

就像爆破後的泥土飛入視野。

It's really awesome stuff.

這真的是很棒的東西。

I would love to try some MatVidPro ones.

我很想試試 MatVidPro 的產品。

I mean, most of these are pretty simple words that he's trying in here.

我的意思是，他在這裡嘗試的大部分詞語都很簡單。

So I'd like to see how it does with more complex stuff like a username.

是以，我想看看它如何處理用戶名等更復雜的內容。

And here you can see it does text in novel situations, not just where you would expect to see it.

在這裡，你可以看到它在新奇的環境中，而不僅僅是在你期望看到它的地方。

So we have a giant maybe on top of, you know, an exit sign on the highway or something.

所以，我們有一個巨大的，也許在頂部，你知道，在高速公路上的出口標誌或東西。

It looks like there's some interesting car accidents and really, really weird craziness going on in the corner here.

看來這裡的角落裡發生了一些有趣的車禍和非常非常奇怪的瘋狂事件。

So these models, keep in mind, are by no means perfect.

是以，請記住，這些模型絕非完美無缺。

But this absolutely competes with the best of the best, which is Sora.

但這絕對是與索拉這一佼佼者的競爭。

And the list kind of just goes on for you guys.

對你們來說，這樣的例子不勝枚舉。

I mean, you can see on top of a train, giant danger sign.

我是說，你可以在火車頂上看到巨大的危險標誌。

It obviously looks edited, but it gets the point across.

這顯然是經過剪輯的，但卻表達了意思。

This was incredible.

太不可思議了

I mean, this is clearly like a bowl of soup or something.

我的意思是，這顯然就像一碗湯什麼的。

And then you can see these little letters drop in.

然後你就能看到這些小字母掉落進來。

And the way that the physics and the letters move with like the liquid, it just looks so realistic.

物理原理和字母移動的方式就像液體一樣，看起來非常逼真。

And it looks like it's not been created in a machine or a physics environment.

它看起來不是在機器或物理環境中創建的。

It looks like it was created or actually filmed in real life.

它看起來像是在現實生活中創作或拍攝的。

And that's really one of the more interesting parts about AI generated video that you're going to notice is that since it is trained on real footage, it gets a lot more close to realism in the physics department than you'll see in like actual 3D animation, because it's something that is really hard to replicate.

你會注意到，人工智能生成視頻的一個有趣之處在於，由於它是在真實素材的基礎上訓練出來的，是以它在物理方面比實際的三維動畫更接近真實，因為這是很難複製的東西。

Again, we have more of that latent horrors type of thing.

同樣，我們還有更多潛在的恐怖類型。

We got Pytorch written on a wall here, but it's someone who is walking through the woods and then comes across something.

這裡的牆上寫著 "Pytorch"，但那是一個在樹林裡行走的人，然後遇到了什麼。

Here we've got coffee pouring into a mug.

在這裡，我們把咖啡倒進一個杯子裡。

This is by Emily Golden, another member of the runway staff.

這是另一位跑道工作人員艾米莉-戈登（Emily Golden）的作品。

So it seems like they all have access to it, but there's just so many crazy things that can go on.

是以，他們似乎都有機會接觸到它，但可能發生的瘋狂事情實在太多了。

Here, folks, is yet another impressive display of Gen 3's knowledge capabilities.

朋友們，這是對 Gen 3 知識能力的又一次令人印象深刻的展示。

We take this ice cube and we drop it on a searing hot pan, and immediately it starts to, well, melt into water, which shows that the model has a very intelligent and nuanced understanding of how physics work and interact with our natural world.

我們把這個冰塊放在灼熱的平底鍋上，它馬上就開始融化成水，這表明模型對物理學如何工作以及如何與我們的自然世界相互作用有著非常聰明和細緻入微的理解。

So the ice isn't necessarily melt, but it definitely produces the water.

是以，冰不一定會融化，但一定會產生水。

We're getting a lot closer to a model that can generalize and think about physical interactions in our world very well.

我們越來越接近一個模型，它能很好地概括和思考我們世界中的物理相互作用。

So again, coming straight from runway ML here, we do have some specs.

所以，我們還是直接從 ML 跑道過來，我們確實有一些規格。

So it takes about 90 seconds to get a 10-second video, which is not too shabby at all.

是以，拍攝一段 10 秒鐘的視頻大約需要 90 秒，這一點也不寒酸。

That's very quick generation, can generate multiple videos at once.

生成速度非常快，可以同時生成多個視頻。

They also are going to be adding motion brush, advanced camera controls, and director mode to this.

他們還將在其中添加動作畫筆、高級相機控制和導演模式。

And they're also going to have more fine-grained control over structure, style, and motion.

他們還將對結構、風格和動作進行更精細的控制。

I mean, again, we didn't really get this kind of announcement with Sora.

我的意思是，再說一次，我們在索拉身上並沒有得到這樣的消息。

Sora is very much kept under wraps as an AI video generator, and we're already getting all of these crazy other generators cropping up, and Gen 3 seems to be the most fully-fledged of the bunch.

作為一款人工智能視頻生成器，索拉一直處於保密狀態，而我們已經看到了其他各種瘋狂的生成器，3 代似乎是其中最成熟的一個。

Here is a nice example of some 3D animation of an anime girl.

下面是一個動漫女孩三維動畫的精彩示例。

This one's a little bit creepy.

這個有點令人毛骨悚然。

The eyes and the face are very consistent, and the hair and everything, it looks good, but it's creepy.

眼睛和臉部的造型非常一致，頭髮和其他一切看起來都不錯，但卻讓人毛骨悚然。

It looks like a doll or something.

看起來像個洋娃娃什麼的。

Here we've got some footage of some fire burning in the distance.

在這裡，我們看到了一些遠處火光沖天的畫面。

Again, this looks super real, almost scary real.

同樣，這看起來超級真實，幾乎真實得嚇人。

People might think that this is a genuine actual video that was recorded.

人們可能會認為這是一段真實錄制的視頻。

Lots of really great cinematic stuff as well.

還有很多非常棒的電影內容。

Again, some of it does delve into the more creepy side of things, like we've got giant ears on someone.

同樣，其中有些內容確實讓人毛骨悚然，比如我們在某人身上發現了巨大的耳朵。

This just looks weird.

這看起來很奇怪。

I don't know.

我不知道。

It's just a strange generation.

這真是奇怪的一代。

It freaks me out a little bit.

這讓我有點害怕。

Here they kept also showing off this example.

在這裡，他們也一直在展示這個例子。

You might have seen this floating around, where we've got slow-motion video of a wig and glasses dropping perfectly on someone's head, and they did this with multiple different people.

你可能已經看到過這樣的視頻，在慢動作視頻中，假髮和眼鏡完美地戴在一個人的頭上，而且是在多個不同的人身上完成的。

It's clear that they have a pretty precise ability to at least manipulate a good seed, which is nice to see.

很明顯，他們有相當精確的能力，至少能操縱好種子，這一點很值得欣慰。

Here, that is the one that kind of went viral here, where it's this really surprised face of a guy, and the wig and glasses drops right on his head.

在這裡，這就是那張引起病毒式傳播的照片，照片中的人一臉驚訝，假髮和眼鏡正好掉在他的頭上。

This one's really cool.

這個真的很酷。

So a hidden civilization pops out in the clouds, and it looks so legit.

於是，一個隱藏在雲層中的文明突然出現了，它看起來是那麼的合法。

I mean, again, scenes from movies.

我是說，還是電影裡的場景。

That's what a lot of this looks like.

很多事情看起來就是這樣。

I have no doubt in my mind that it was trained on cinematic prompts.

我毫不懷疑它是根據電影提示訓練出來的。

The training data has got to be really good, just like Sora's was.

訓練數據必須非常好，就像索拉的訓練數據一樣。

Their runway is like right behind OpenAI.

他們的跑道就在 OpenAI 的後面。

Right behind OpenAI, I would say.

我想說，它僅次於 OpenAI。

This is by always editing on Twitter here.

這是在 Twitter 上經常編輯的內容。

This shows a little bit of sound design to go along with a video and really show how you can make this AI-generated content come to life in a way that is no BS.

這展示瞭如何將人工智能生成的內容活靈活現地呈現在視頻中的聲音設計。

You know, it's going to be usable much sooner than you think.

要知道，它比你想象的更快就能使用了。

When you're ready, gently open your eyes.

準備好後，輕輕睜開眼睛。

Returning to your surroundings.

回到你的周圍