How AI could become an extension of your mind | Arnav Kapur

譯者: Helen Chang 審譯者: Zoe Chang

Ever since computers were invented,

自從發明電腦以來，

we've been trying to make them smarter and more powerful.

我們一直嘗試使它們 更聰明、更強大。

From the abacus, to room-sized machines,

從算盤到房間大小的電腦，

to desktops, to computers in our pockets.

再從桌上型到能放進口袋裡的電腦。

And are now designing artificial intelligence to automate tasks

如今正在發展中的人工智慧

能夠自動執行 需要人類智慧的任務。

that would require human intelligence.

If you look at the history of computing,

從計算機的演變史來看，

we've always treated computers as external devices

電腦始終被視為

代替我們計算和執行的外部設備。

that compute and act on our behalf.

What I want to do is I want to weave computing, AI and internet as part of us.

我想要把計算、人工智慧 和網路，整合成我們內在的一部分，

As part of human cognition,

成為人類認知過程的一部分，

freeing us to interact with the world around us.

讓我們自由地與周圍的世界互動。

Integrate human and machine intelligence

將人類與機器智慧結合起來，

right inside our own bodies to augment us, instead of diminishing us or replacing us.

從內在去補強我們， 而不是削弱或取代我們。

Could we combine what people do best, such as creative and intuitive thinking,

我們能不能把人類的強項， 例如創造性和直覺思維，

with what computers do best,

與電腦的強項，

such as processing information and perfectly memorizing stuff?

例如資訊處理和完美記憶， 結合在一起呢？

Could this whole be better than the sum of its parts?

一旦整合，是否比各項優勢 加總起來更好呢？

We have a device that could make that possible.

我們做了一個設備來實現這個理想。

It's called AlterEgo, and it's a wearable device

它叫做 AlterEgo （改變自我）， 是一種穿戴式設備，

that gives you the experience of a conversational AI

藉由它你可以體驗與你腦中的

that lives inside your head,

人工智慧對話，

that you could talk to in likeness to talking to yourself internally.

就像你與內在自我對話那樣容易。

We have a new prototype that we're showing here,

今天展示的是一個新的原型機，

for the first time at TED, and here's how it works.

第一次出現在 TED ， 這是它的原理。

Normally, when we speak,

通常，當我們說話時，

the brain sends neurosignals through the nerves

大腦發送神經信號，通過神經，

to your internal speech systems,

到達你內部的語音系統，

to activate them and your vocal cords to produce speech.

使你的聲帶動起來，發出話語；

One of the most complex cognitive and motor tasks

那是我們人類所做

最複雜的認知和動作之一。

that we do as human beings.

Now, imagine talking to yourself

現在想像與自己對話，

without vocalizing, without moving your mouth,

不出聲，不動嘴，

without moving your jaw,

也不動下巴，

but by simply articulating those words internally.

就只在心中默念那樣。

Thereby very subtly engaging your internal speech systems,

從而很巧妙地啟動 你內部的語音系統，

such as your tongue and back of your palate.

像是你的舌頭和軟顎。

When that happens,

當你這樣做時，

the brain sends extremely weak signals to these internal speech systems.

大腦會向內部語音系統 發送極弱的信號。

AlterEgo has sensors,

AlterEgo 的感測器嵌在

embedded in a thin plastic, flexible and transparent device

一個有彈性、透明塑膠薄片裡，

that sits on your neck just like a sticker.

像張貼紙貼在脖子上。

These sensors pick up on these internal signals

這些感測器從我們的皮膚表面

sourced deep within the mouth cavity,

接收來自口腔深處的內部信號。

right from the surface of the skin.

An AI program running in the background

在後台運行的人工智慧程式

then tries to figure out what the user's trying to say.

負責解讀使用者在表達什麼。

It then feeds back an answer to the user

然後它透過骨傳導耳機

向使用者提供答案。

by means of bone conduction,

audio conducted through the skull into the user's inner ear,

使用者聽到的聲音

透過顱骨傳到使用者內耳中，

that the user hears,

overlaid on top of the user's natural hearing of the environment,

與我們接收周遭環境的聽覺重疊，

without blocking it.

兩者並行毫無阻礙。

The combination of all these parts, the input, the output and the AI,

這一整個輸入、輸出 和人工智慧的組合，

gives a net subjective experience of an interface inside your head

讓你對這個腦中的界面 得以有一個純然的主觀體驗，

that you could talk to in likeness to talking to yourself.

如同你與自己交談那樣。

Just to be very clear, the device does not record or read your thoughts.

我要清楚指出，該設備 不會記錄或閱讀你的想法。

It records deliberate information that you want to communicate

它只會記錄你選擇要

透過內部語音系統 去溝通的特定信息。

through deliberate engagement of your internal speech systems.

People don't want to be read, they want to write.

人們不想被讀心，他們想要表達。

Which is why we designed the system

這就是為什麼我們將這系統

to deliberately record from the peripheral nervous system.

設計成只去記錄從 周邊神經系統來的訊息。

Which is why the control in all situations resides with the user.

無論何種情況，主控權 都在使用者手上。

I want to stop here for a second and show you a live demo.

我先暫停一下，來做一個現場示範。

What I'm going to do is, I'm going to ask Eric a question.

我要問艾瑞克一個問題。

And he's going to search for that information

他搜索答案時

without vocalizing, without typing, without moving his fingers,

不能出聲、不能打字、也不能動手、

without moving his mouth.

同時也不開口。

Simply by internally asking that question.

純粹就是在內心裡提問。

The AI will then figure out the answer and feed it back to Eric,

人工智慧會找出答案， 通過語音和設備，

through audio, through the device.

反饋給艾瑞克。

While you see a laptop in front of him, he's not using it.

他前面有一台筆記型電腦， 可是他不會用到。

Everything lives on the device.

所有一切都在裝置裡。

All he needs is that sticker device to interface with the AI and the internet.

只要那片貼紙裝置能與人工智慧 和網路去互動就夠了。

So, Eric, what's the weather in Vancouver like, right now?

卡普爾：艾瑞克，此刻 溫哥華的天氣怎樣？

What you see on the screen

螢幕上的這些字

are the words that Eric is speaking to himself right now.

是此刻艾瑞克與他自己的對話。

This is happening in real time.

這是即時同步的紀錄。

Eric: It's 50 degrees and rainy here in Vancouver.

艾瑞克：溫哥華這裡 50 度，下雨。

Arnav Kapur: What happened is that the AI sent the answer

卡普爾：剛剛這是人工智慧

through audio, through the device, back to Eric.

透過語音及裝置將答案發送給艾瑞克。

What could the implications of something like this be?

這樣的裝置會帶來什麼影響？

Imagine perfectly memorizing things,

想像分毫不差的記憶力，

where you perfectly record information that you silently speak,

只要無聲地默念即可完美地記錄一切，

and then hear them later when you want to,

然後在需要時得以聆聽；

internally searching for information,

還能向內搜索訊息，

crunching numbers at speeds computers do,

以電腦的速度處理數字，

silently texting other people.

以及無聲地給他人發出簡訊。

Suddenly becoming multilingual,

還有突然通曉多種語言，

so that you internally speak in one language,

你在心裡用一種語言說話，

and hear the translation in your head in another.

立即就在腦中聽到 另一種語言的翻譯。

The potential could be far-reaching.

潛力深遠。

There are millions of people around the world

全世界有數百萬人

who struggle with using natural speech.

喪失與生俱來的語言能力。

People with conditions such as ALS, or Lou Gehrig's disease,

例如肌萎縮側索硬化或稱盧·賈里格症 （俗稱漸凍人） 的患者，

stroke and oral cancer,

還有中風及口腔癌患者，

amongst many other conditions.

以及許多其他疾病的患者。

For them, communicating is a painstakingly slow and tiring process.

對他們來說，溝通變得 艱苦、緩慢而累人。

This is Doug.

這位是道格。

Doug was diagnosed with ALS about 12 years ago

大約 12 年前，道格 被診斷為漸凍人，

and has since lost the ability to speak.

從此失去了說話的能力。

Today, he uses an on-screen keyboard

如今他使用螢幕鍵盤，

where he types in individual letters using his head movements.

用頭部動作將英文字母逐一輸入。

And it takes several minutes to communicate a single sentence.

一個句子得花掉好幾分鐘。

So we went to Doug and asked him

我們去找道格，

what were the first words he'd like to use or say, using our system.

如果用我們的系統， 他要說的第一句話是什麼。

Perhaps a greeting, like, "Hello, how are you?"

或許是「哈囉，你好嗎？」的問候

Or indicate that he needed help with something.

或者表達他需要什麼樣的幫助。

What Doug said that he wanted to use our system for

道格說他可以用我們的系統

is to reboot the old system he had, because that old system kept on crashing.

去重新開機，因為他那個 舊系統老是當機。

(Laughter)

（笑聲）

We never could have predicted that.

這完全出乎我們的意料。

I'm going to show you a short clip of Doug using our system for the first time.

我要給你看道格第一次 使用我們系統的短片。

(Voice) Reboot computer.

（語音）重新啟動電腦。

AK: What you just saw there

卡普爾：你剛才看到的

was Doug communicating or speaking in real time for the first time

是道格失去說話的能力後

第一次能夠即時地溝通和說話。

since he lost the ability to speak.

There are millions of people

有數百萬人或許可以像道格這樣，

who might be able to communicate in real time like Doug,

即時地與其他人、朋友和家人溝通。

with other people, with their friends and with their families.

My hope is to be able to help them express their thoughts and ideas.

我希望能幫助他們表達想法和意見。

I believe computing, AI and the internet

我相信計算機，人工智慧和網路

would disappear into us as extensions of our cognition,

能夠內化成我們認知的延伸，

instead of being external entities or adversaries,

不再是外部裝置或競爭對手，

amplifying human ingenuity,

它們可以擴大人類的聰明才智，

giving us unimaginable abilities and unlocking our true potential.

賦予我們難以想像的能力， 而且釋放我們的真正潛力。

And perhaps even freeing us to becoming better at being human.

甚至可能還解放我們， 去善盡作為人類的本分。

Thank you so much.

非常感謝。

(Applause)

（掌聲）

Shoham Arad: Come over here.

秀涵 · 阿瑞德：過來這裡。

OK.

好。

I want to ask you a couple of questions, they're going to clear the stage.

我想問你幾個問題， 他們要清理那邊的舞台。

I feel like this is amazing, it's innovative,

我覺得這很棒、很有創意，

it's creepy, it's terrifying.

也有點毛骨悚然、令人恐懼。

Can you tell us what I think ...

你能告訴我們為什麼——

I think there are some uncomfortable feelings around this.

我覺得這個發明引發 了一種令人不自在的感覺。

Tell us, is this reading your thoughts,

請告訴我們，這個裝置是像讀心術嗎？

will it in five years,

是否在五年之內，

is there a weaponized version of this, what does it look like?

會發展出一個武器化的版本？ 屆時它又會是什麼樣子？

AK: So our first design principle, before we started working on this,

卡普爾：我們開始研究前的 第一個設計原則就是

was to not render ethics as an afterthought.

不要等到最後才考慮道德問題。

So we wanted to bake ethics right into the design.

因此我們一開始就把 道德規範納入其中。

We flipped the design.

我們翻轉了設計。

Instead of reading from the brain directly,

與其直接從腦部讀取資料，

we're reading from the voluntary nervous system

改成從自主神經系統去讀取，

that you deliberately have to engage to communicate with the device,

由你決定與裝置分享溝通那些資訊，

while still bringing the benefits of a thinking or a thought device.

仍能享有思考裝置的好處。

將人工與人類智慧結合得兩全其美。

The best of both worlds in a way.

SA: OK, I think people are going to have a lot more questions for you.

阿瑞德：嗯，我想大家一定 有更多問題要問你。

Also, you said that it's a sticker.

此外，你說它是貼紙。

So right now it sits just right here?

現在就貼這裡嗎？

Is that the final iteration,

這是目前最新的版本嗎？

what the final design you hope looks like?

你希望最終的設計是什麼樣的？

AK: Our goal is for the technology to disappear completely.

卡普爾：我們的目標是讓技術隱身。

SA: What does that mean?

阿瑞德：什麼意思？

AK: If you're wearing it, I shouldn't be able to see it.

卡普爾：意思是， 你穿戴著它，我卻看不出來。

You don't want technology on your face, you want it in the background,

因為沒有人希望它清楚秀在你的臉上，

而是要它隱身於後，增強你的能力。

to augment you in the background.

So we have a sticker version that conforms to the skin,

所以我們做了一個貼紙版本，

that looks like the skin,

看起來像皮膚。

but we're trying to make an even smaller version

但我們正努力製作更小的版本，

that would sit right here.

可以貼在這裡。

SA: OK.

阿瑞德：好的。

I feel like if anyone has any questions they want to ask Arnav,

我覺得如果每個人都能提問的話，

he'll be here all week.

阿納夫可能得在這裡待上一個星期。

OK, thank you so much, Arnav.

非常感謝你，阿納夫。

AK: Thanks, Shoham.

卡普爾：謝謝秀涵。