The AI-Powered Tools Supercharging Your Imagination | Bilawal Sidhu | TED

譯者: 麗玲 辛

All right. Good afternoon, y’all.

好的，大家下午好。

Let's talk about blending reality and imagination.

讓我們來談談如何融合現實和想像。

But first, let's take a step back in time to 2001.

但首先，讓我們回到 2001 年。

As an 11-year-old in India,

當時我 11 歲，住在印度，

I became obsessed with computer graphics and visual effects.

迷上了電腦繪圖和視覺效果。

Of course, at that age,

當然，在那個年紀，

it meant making cheesy videos kind of like this.

這意味著製作像這樣有點幼稚的影片。

But therein started a foundational theme in my life,

不過，從那時我就展開 我生命中的一個基本主題，

the quest to blend reality and imagination.

追求現實與想像的融合。

And that quest has stayed with me

這個追求一直伴隨著我，

and permeated across my decade-long career in tech,

貫穿了我長達十年的科技職涯，

working as a product manager at companies like Google

我曾在 Google 等公司 擔任產品經理，

and as a content creator on platforms like YouTube and TikTok.

在 YouTube、TikTok 等 平台上擔任內容創作者。

So today, let's deconstruct this quest to blend reality and imagination

今天，讓我們來解構 這個融合現實與想像的追求，

and explore how it’s getting supercharged -- buzzword alert --

並探索人工智慧如何增強 ——流行語警報——這一追求。

by artificial intelligence.

Let's start with the reality bit.

讓我們從現實部分開始談起。

You probably heard about photogrammetry.

你也許聽說過攝影測量學。

這是一門藝術兼科學， 使用照片和其他感測器

It's the art and science of measuring stuff in the real world

using photos and other sensors.

來測量現實世界中的東西。

What required massive data centers and teams of experts in the 2000s

在 2000 年代，這需要 大規模的數據中心和專家團隊，

became increasingly democratized by the 2010s.

但到了 2010 年代， 這變得越來越大眾化。

Then, of course, machine learning came along

接著，當然，機器學習出現了，

and took things to a whole new level

使用神經輻射場 （NeRFs）等技術，

with techniques like neural radiance fields, or NeRFs.

技術提升到了全新的水平。

這裡你看到的是一個 AI 模型，

What you're seeing here is an AI model

creating a ground-up volumetric 3D representation

僅使用 2D 圖像，就能從零開始 創建一個 3D 立體再現。

using 2D images alone.

But unlike older techniques for reality capture,

但不同於舊的現實捕捉技術，

NeRFs do a really good job of encapsulating the sheer complexity

NeRF 在封裝現實的複雜性

and nuance of reality.

和細微差別方面做得非常好。

The vibe, if you will.

如果你想，可以做出氛圍。

Twelve months later, you can do all of this stuff

十二個月後，你可以使用 口袋裡的 iPhone，

using the iPhone in your pocket,

使用 Luma 等應用程式 來完成這些事情。

using apps like Luma.

It's like 3D screenshots for the real world.

這就像現實世界的 3D 螢幕截圖。

Capture anything once and reframe it infinitely in postproduction,

捕捉任何物體的影像後， 就可以在後製中無限地重新建構，

so you can start building that collection of spaces, places and objects

你可以開始打造你的影像集，

放入你喜歡的空間、地點和物體，

that you truly care about

and conjure them up in your future creations.

並在未來的創作中將它們召喚出來。

So that's the reality bit.

以上是現實的方面。

As NeRFs were popping off last year,

隨著去年 NeRF 的興起，

the AI summer was also in full effect,

AI 夏季也全面展開，

with Midjourney, DALL-E 2, Stable Diffusion

Midjourney、DALL-E 2、 Stable Diffusion

all hitting the market around the same time.

幾乎同時上市。

但我愛上的是修復技術。

But what I fell in love with was inpainting.

This technique allows you to take existing imagery

你可用這種技術隨心所欲地 增強現有的圖像，

and augment it with whatever you like,

and the results are photorealistically fantastic.

結果非常逼真。

It blew my mind

這讓我大為驚訝，

because stuff that would have taken me like three hours in classical workflows

因為在傳統工作流程中 可能需要我三個小時的事情，

I could pull off in just three minutes.

現在只需要三分鐘就能完成。

But I wanted more.

但我想要更多。

Enter ControlNet,

介紹一下 ControlNet，

a game-changing technique by Stanford researchers

這是史丹佛大學研究人員提出的 顛覆傳統規則的技術，

that allows you to use various input conditions

你可以使用各種輸入條件

to guide and control the AI image generation process.

來引導和控制 AI 影像生成的過程。

So in my case, I could take the depth information

就我而言，我可以從 3D 掃描中

and the texture detail from my 3D scans

獲取深度資訊和紋理細節，

and use it to literally reskin reality.

並用來一絲一毫地重塑現實。

Now, this isn't just cool video.

這不僅僅可以做出很酷的視頻，

There’s a lot of useful use cases, too.

還有很多好用的用例。

For example, in this case I'm taking a 3D scan of my parents' drawing room,

例如，在本例中，我對我父母的 客廳進行 3D 掃描，

（我母親喜歡這樣稱呼它）

as my mother likes to call it,

and reskinning it to different styles of Indian decor

並將其重新設計為 不同風格的印度裝飾，

and doing so while respecting the spatial context and the layout

但不改動內部空間環境和佈局。

of the interior space.

如果你仔細觀察，我相信你可以看出

If you squint, I'm sure you can see

how this is going to transform architecture

這將永遠改變建築和室內設計的方式。

and interior design forever.

You could take that 2016 scan of a Buddha statue

你可以將 2016 年掃描的佛像

and reskin it to be gloriously golden

改成華麗的金色，

while pulling off these impossible camera moves

同時完成任何其他方法 無法做到的運鏡方式。

you just couldn't do any other way.

Or you could take that vacation footage from your trip to Tokyo

或者你可以拿你在東京旅行時 拍攝的度假片段，

and bring these cherry blossoms to life in a whole new way.

以全新的方式讓櫻花煥發活力。

And let me tell you, cherry blossoms look really good during the day,

我告訴你，櫻花白天很好看，

but they look even better at night.

但晚上更漂亮。

Oh, my God. They sure are glowing.

我的天啊。 他們確實在發光。

It's almost like this dreamlike quality

就這幾乎夢幻般的特質，

where you can use AI to accentuate the best aspects of the real world.

你可以用 AI 來增強 現實世界最美的一面。

Natural landscapes look just as beautiful.

自然景觀看起來同樣美麗。

Like this waterfall that could be on another planet.

這個瀑布看起來像是在另一個星球。

But of course, you could go over the hills and far away

當然，你也可以 從另一個次元越過山丘，

to the French Alps from another dimension.

到達遙遠的法國阿爾卑斯山。

But it's not just static scenes.

但這不僅僅是靜態場景。

You can do this stuff with video, too.

你也可以用影片來做這些事。

I can't wait till this technology is running at 30 frames per second

我迫不及待地看到這項技術能夠 以每秒 30 幀的速度運作，

because it's going to transform augmented reality and 3D rendering.

因為這將徹底改變 擴增實境和 3D 渲染。

I mean, how soon until we're channel-surfing realities

我是說，要多久我們才能在現實世界上

疊加多層次的現實？

layered on top of the real world?

Of course, just like reality capture got democratized,

當然，就像現實捕捉變得大眾化一樣，

all these tools from last year are getting even easier.

每個去年上市的工具 都變得更容易上手。

So instead of me spending hours

不再需要花幾個小時 組合各種不同的工具，

weaving together a bunch of different tools,

tools like Runway and Kaiber let you do exactly the same stuff

像 Runway、Kaiber 這樣的工具，

只需要點擊幾下即可完成 完全相同的工作。

with just a couple clicks.

Want to go from day to night? No problemo.

想從白天變成夜晚嗎？沒問題。

Want to get that retro 90s aesthetic from "Full House"?

想感受《歡樂滿屋》中的 90 年代復古美學嗎？

你也可以做到。

You can do that too.

But it goes beyond reality capture.

不只現實捕捉。

許多公司， 像 Wonder Dynamics，

Companies like Wonder Dynamics

正將影片轉變為 完美的表演捕捉形式，

are turning video into this immaculate form of performance capture

so you can embody fantastical creatures using the phone in your pocket.

讓你用口袋裡的手機 就能體現奇幻生物。

This is stuff that James Cameron only dreamt about in the 2000s.

這是詹姆斯·卡梅隆 在 2000 年代做夢才有的東西。

And now you could do it with your iPhone? That’s absolutely wild to me.

現在用 iPhone 就做得到？ 對我來說，這很誇張。

So when I look back at the past two decades

因此，當回顧過去的二十年，

and this ill-tailored tapestry of tools that I've had to learn,

我得學這一大串不好用的工具，

我為下一代創作者的 未來感到樂觀。

I feel a sense of optimism for what lies ahead

for the next generation of creators.

The 11-year-olds of today

現在的 11 歲小孩

don't have to worry about all of that crap.

不必擔心那些瑣碎的東西。

All they need to do is have a creative vision

他們只需要有創造力的視野，

and a knack for working in concert with these AI models,

與 AI 模型協同工作的技巧，

these AI models that are truly a distillation

這些 AI 模型是 人類知識和創造力的真正精髓。

of human knowledge and creativity.

And that's a future I'm excited about,

這是我期待的未來，

a future where you can blend reality and imagination

一個與可靠的 AI 合作夥伴 融合現實和想像的未來。

with your trusty AI copilot.

Thank you very much.

非常感謝大家。

(Applause)

（掌聲）