The ethical dilemma of self-driving cars - Patrick Lin

譯者: Ann Chen 審譯者: Max Chern

This is a thought experiment.

這是一個假想的實驗

Let's say at some point in the not so distant future,

假設在不久將來的某時

you're barreling down the highway in your self-driving car,

你坐在無人駕駛車內 高速行駛於高速公路上

and you find yourself boxed in on all sides by other cars.

你發現你的四周都有車

Suddenly, a large, heavy object falls off the truck in front of you.

突然前面卡車有大型重物掉落

Your car can't stop in time to avoid the collision,

你無法及時煞車以避免衝撞

so it needs to make a decision:

所以必須做個決定：

go straight and hit the object,

直行撞擊那重物、

swerve left into an SUV,

轉向左邊撞運動休旅車、

or swerve right into a motorcycle.

或轉向右邊撞摩托車

Should it prioritize your safety by hitting the motorcycle,

它應該以你的安全 為優先考慮去撞摩托車

minimize danger to others by not swerving,

還是應該不要轉向，以減少傷害他人

even if it means hitting the large object and sacrificing your life,

即使這意謂著直撞那重物 犧牲你的生命

or take the middle ground by hitting the SUV,

或是取折中方式去撞運動休旅車

which has a high passenger safety rating?

因為它有較高的乘客安全等級？

So what should the self-driving car do?

無人駕駛車應該怎麼做呢？

If we were driving that boxed in car in manual mode,

如果我們是徒手駕駛那輛 被夾在車陣中的車子

whichever way we'd react would be understood as just that,

無論我們做出那種反應 都將被視為

a reaction,

一種生理反應

not a deliberate decision.

而非經深思熟慮後的決定

It would be an instinctual panicked move with no forethought or malice.

那是一種本能的驚慌舉動 不是預先策劃或蓄意的

But if a programmer were to instruct the car to make the same move,

但如果是程式設計師下指令 讓車子做出同樣的動作 ─

given conditions it may sense in the future,

針對它未來可能感應到的狀況 ─

well, that looks more like premeditated homicide.

這就比較像是預謀殺人了

Now, to be fair,

平心而論

self-driving cars are predicted to dramatically reduce traffic accidents

無人駕駛車是預計來 大幅降低意外車禍

and fatalities

及死亡事故

by removing human error from the driving equation.

因可經由駕駛程式來去除人為疏失

Plus, there may be all sorts of other benefits:

況且尚有其他許多好處

eased road congestion,

疏解道路壅塞

decreased harmful emissions,

減少排放有害廢氣

and minimized unproductive and stressful driving time.

及減少無效率及緊張的開車時間

But accidents can and will still happen,

但意外還是仍然會發生

and when they do,

當發生時

their outcomes may be determined months or years in advance

它們的結果可能早在數月 或數年前就已經預先

by programmers or policy makers.

被程式設計師或決策者所決定了

And they'll have some difficult decisions to make.

他們將要做出一些困難的決定

It's tempting to offer up general decision-making principles,

試圖提出一般性決策原則

like minimize harm,

例如「傷害最小化」

but even that quickly leads to morally murky decisions.

即使如此，它仍很快變成 道德上含糊不清的決定

For example,

例如

let's say we have the same initial set up,

比方說我們遇到和原先相同的狀況

but now there's a motorcyclist wearing a helmet to your left

但現在有個戴安全帽的 摩托車騎士在你左邊

and another one without a helmet to your right.

而另有一位沒戴安全帽的在你右邊

Which one should your robot car crash into?

你的自動車應該撞向那邊？

If you say the biker with the helmet because she's more likely to survive,

如果你說是那位戴安全帽的騎士 因為她存活率比較高

then aren't you penalizing the responsible motorist?

那麼你不就是在懲罰 一位守法的摩托車騎士呢？

If, instead, you save the biker without the helmet

反之，你說撞向那位沒戴安全的騎士

because he's acting irresponsibly,

因為他不負責任

then you've gone way beyond the initial design principle about minimizing harm,

那麼你這做法已經遠離 「傷害最小化」的最初設計原則了

and the robot car is now meting out street justice.

而那自動車現在做出懲罰 執行道路正義了

The ethical considerations get more complicated here.

道德的考量在此變得更複雜

In both of our scenarios,

這兩種情況

the underlying design is functioning as a targeting algorithm of sorts.

其基本設計是一種 「標靶運算法則」的運作

In other words,

換言之

it's systematically favoring or discriminating

它蓄意地厚此薄彼

against a certain type of object to crash into.

撞向一個特定物體

And the owners of the target vehicles

那位被選中的車主

will suffer the negative consequences of this algorithm

因這運算的負面結果而遭殃了

through no fault of their own.

而這並不是他們自己的錯

Our new technologies are opening up many other novel ethical dilemmas.

我們的新技術出現了 許多其他新的道德難題

For instance, if you had to choose between

例如，如果你必須二選一時

a car that would always save as many lives as possible in an accident,

一輛意外時會有最多存活率的車

or one that would save you at any cost,

或是一輛不計一切代價先救你的車

which would you buy?

你會買那一輛？

What happens if the cars start analyzing and factoring in

如果車子開始分析並考慮

the passengers of the cars and the particulars of their lives?

乘客及他們的身份地位 你想會發生什麼事？

Could it be the case that a random decision

會不會一個隨機的決定

is still better than a predetermined one designed to minimize harm?

仍優於一個「傷害最小化」 的預定程式呢？

And who should be making all of these decisions anyhow?

到底應該由誰來做所有的決定？

Programmers? Companies? Governments?

程式設計師？

公司？

政府？

Reality may not play out exactly like our thought experiments,

實際情形也許不會和我們 假想實驗的情況完全相同

but that's not the point.

但重點不在這兒

They're designed to isolate and stress test our intuitions on ethics,

它們被設計來單獨在壓力下 測試我們對道德的直覺

just like science experiments do for the physical world.

就如同真實世界的科學實驗一樣

Spotting these moral hairpin turns now

發現這些道德急轉彎

will help us maneuver the unfamiliar road of technology ethics,

將協助我們駕馭在這條 陌生的 “技術倫理學” 路上

and allow us to cruise confidently and conscientiously

讓我們有信心地、有良知地

into our brave new future.

駛向我們新的美好未來

翻譯：Yun An Chen