一個不小心，科技就成了“隱形殺手”

Musk. Hawking. Gates. The tech visionaries have redoubled their warnings about how we could destroy ourselves with technology. But theyre not talking about deliberately pushing buttons to trigger annihilation1）—Kennedy and Khrushchev could have done that back in 622）. They mean accidentally killing ourselves with something that we invent for our own benefit， that later acquires a mind of its own and a shirty3） attitude.

But theres no need to wait for the next big existential threat. Were already designing a whole new world of ways to accidentally off ourselves， and no one seems worried about it.

As biochemists concoct4） new life-extending medications， calls to Poison Control after swallowing the wrong pills or the wrong number of the right pills have recently doubled. We put a smartphone in every hand， and now more than 1，000 distraction-related crashes happen on our roads every day （also steadily rising）. Kids and pets succumb to heatstroke5） inside cars that are more environmentally sealed than ever—were on track to set a new record for hot car deaths in 2017. Falling off of ladders？ Even those numbers are climbing， and if youre wondering how that could possibly be related to technology， well， read on.

Its a little embarrassing to have to admit that accidental deaths are increasing in a world that our forebears made safer for hundreds of years. Our grandparents saw the invention of the automobile， the blender， the bulldozer， and the radial arm saw6）—and they made them all safer. Is what were inventing today really more dangerous than that stuff？ We may not have to worry about sentient A.I. any time soon， but our innovations are quietly outpacing our ability to figure out how to not get wiped out7） by them.

We rely on instincts—our common sense—to tell us whats dangerous and whats safe. Before technology came along， hazards were mostly self-explanatory. Bears， snakes， sharp sticks， cliffs—its hard to get any of this stuff wrong. And our instincts became exquisitely8） tuned to them across millennia. But the hazards we face today can be more subtle， harder to recognize， and even counter-intuitive.

We instinctively take a step back when we see something large， but technology is turning our fear of size on its head9）. Technology progresses by packing more and more power into smaller and smaller packages. One errant wave of a laser pointer and you could bring down an airliner. The number of people who visit emergency rooms following physical interactions with television sets is rising. Its not hard to see why. In the old days， TVs weighed a ton， and they sat in the corner， mostly unbothered. Now TVs are light and portable. They invite us to pick them up， sit them on stuff， or hang them on the wall. And theyre shaped like giant guillotine10） blades. Put the TV in a phone and its more dangerous still. The lighter the TV， the less we fear it， and the higher the body count.endprint

Invisible hazards lurk in the logic and code used to implement our technology. We place blind trust in complex systems that reveal little about how they work. Our medication prescriptions pass through computer systems accessed by professionals who majored in something other than computer science. Patient safety advocates loudly remind us that we are potential victims of human and machine error and that we need to think and ask questions before we swallow pills. But not many people are heeding11） their advice， and medication errors run rampant.

More insidious12） still， we create technology that assumes we have superhuman cognitive abilities， and consumers seem willing to play along. Put a phone， a latte， and a steering wheel in front of us doing 80 mph， and were proud multitaskers. Sure， other people on phones are dangerous， but you can smoothly switch your attention and notice when anything scary pops up， right？ A psychologist in a gorilla suit13） debunked14） that idea some years ago， but we pound our chests in defiance. We imagine ourselves able to accurately assess risks in complex situations after watching a news story about a 15-pound flying sausage crashing through the roof of someones house. We even think were good at seeing oncoming trains while wearing earbuds.

The hazards of yesteryear15） gave us immediate feedback when we screwed up. When we misjudged a bear， the bear instantly let us know. But technology can place the consequences of our missteps at a distance. Delayed reactions， complex chain reactions， hidden reactions—these are all part of how technology works. But adapted to a world of instant feedback， we cruise through the day on autopilot， seldom stopping to consider what could go wrong later down the road. Worse still， technology is even hacking our feedback system. We get a blast of dopamine16） when we check our phone behind the wheel. Edith Harbaugh， whose company， Launch Darkly， specializes in the controlled release of new technology into the wild， pointed out that were creating a dangerously lopsided17） system of actions and rewards： “Were not given snow cones every time we do something safe.”

Technology is even reshaping our safety culture. My grandfather， a craftsman， taught me how to use tools. If I used a ladder wrong， I was quickly corrected. But then technology moved more than half of all grandparents into office jobs. Today people are firing up tools again in record numbers because DIY is in fact cool and arguably good for your soul. But were now getting hurt more while doing it.endprint

Few stop to realize how much we have come to rely on the work of concerned， dedicated consumer product designers and the lawyers who incessantly sue them. But today were inventing stuff faster than we can design safety features. A town in Germany just installed sidewalk traffic lights to alert phone-immersed pedestrians. But what happens when wearables and VR kits come along and were looking up and through a device rather than down at it？

Even the government stepped in on our behalf to create safety standards. But our technology may now be even outpacing our ability to establish new standards. Theres a scene in Anchorman 218） where Will Ferrell flips on the cruise control19） and leaves the drivers seat of his motorhome. When Paul Rudd points out that cruise control only handles the speed， not the steering， chaos ensues. But that confusion isnt just a Hollywood comedy routine. Kelly Funkhouser at the University of Utah recently inventoried20） the names that car manufacturers are giving to speed control， lane-keeping， and blind-spot monitoring functions. She found that similar names are being given to different combinations of these functions across manufacturers. Hop in the wrong rental car， push a button， whip out your phone， and boom … youre Anchorman.

So if youre really worried about summoning the demon， relax. Its already too late. The age of life- and limb-threatening technology is upon us. But our ancestral tendencies may be preventing us from changing our ways or even seeing why we should. In the midst of Silicon Valley， where some of the smartest people live and work， I spotted one driver doing a video call on one phone while texting on another phone. Meanwhile， the numbers keep rising and accidents are now the fourth leading cause of death in the U.S. （For the record， bear-related fatalities remain relatively constant.）

Theres no magic Silicon Valley gadget that will fix this problem. Instead， we need to reconsider the very way we think about being careful—to learn the nuances of a more complex world that we must now navigate and to learn the limits of what our creative， analytical， sometimes fallible21） minds can and cant get away with. The modern world is breaking everything we know about staying safe. If we harbor any hope of squaring off22） against the rise of the machines， were going to need to learn to survive the things we have today.

馬斯克、霍金、蓋茨——這些富有遠見的技術領袖已加大力度警告我們技術會怎樣摧毀人類自己。但是他們討論的并不是有人要故意按下按鈕引發(fā)人類毀滅——早在1962年，肯尼迪和赫魯曉夫原本就可以這么做了。馬斯克他們的意思是，我們?yōu)榱俗陨砝姘l(fā)明了一些東西，這些東西后來有了自己的思維和壞脾氣，一不留神就會要了我們的命。endprint

不過大可不必坐等下一個巨大的生存威脅。我們已經(jīng)正在設計意外滅掉自己的各種全新方法，而且似乎沒人為此擔憂。

生物化學家們調配出了延年益壽的新型藥物，但最近中毒控制中心接到的因服錯藥物或服錯劑量的報警電話卻翻了一番。我們每人手拿一部智能手機，而如今每天道路上有1000多起事故與分心有關（該數(shù)字還在穩(wěn)步上升）。孩子們和寵物死于車內中暑，因為車子密封得越來越好。2017年因車內暑熱死亡的案例正在創(chuàng)下新的紀錄。從梯子上跌落？甚至連這一數(shù)字也在攀升。你如果還在疑惑這怎么可能跟科技扯上關系，那就繼續(xù)往下讀。

不得不承認，雖然幾百年來我們的祖先讓世界變得更加安全，可如今意外死亡的人數(shù)正在上升，這有點令人尷尬。我們的祖輩見證了汽車、攪拌機、推土機和轉向鋸的發(fā)明——而且把它們做得更安全了。我們今天發(fā)明的真比這些東西更危險嗎？我們或許一時半會兒還不必擔心有知覺的人工智能，但是我們的創(chuàng)新正在不知不覺中超越我們揣摩如何不被它們干掉的能力。

我們憑直覺，即我們的常識，感知何謂危險、何謂安全。在科學技術出現(xiàn)之前，危險基本上是不言自明的。熊、蛇、鋒利的枝條、懸崖峭壁——我們很難把諸如此類的東西搞錯。而且經(jīng)歷了幾千年，人們的直覺已經(jīng)能敏銳地感知這些危險?？扇缃裎覀兯媾R的危險可能更隱蔽、更不易察覺，甚至是反直覺的。

我們看到龐然大物時會出于本能后退一步，但是科技卻正在把我們對物體大小的恐懼顛倒過來?？萍纪ㄟ^把越來越多的功能放入越來越小的物品中實現(xiàn)進步。隨意揮動一下激光筆，你就可以使一架大型客機墜落。與電視機進行過身體接觸后去看急診的人數(shù)也在增多。原因不難看出。以前電視機沉得要命，還擺在角落里，我們基本不去動它。而現(xiàn)在，電視機很輕便，讓我們不由想將其拿起，放在其他東西上，或者掛在墻上。電視機的形狀也像是巨大的斷頭臺刀片。電視機的功能放進手機里，這就更危險了。電視越輕，我們對它的恐懼感就越弱，死亡人數(shù)也就越多。

看不見的危險就潛藏在實現(xiàn)科技的邏輯和代碼中。我們對復雜的系統(tǒng)盲目信任，卻對其如何運行一無所知。我們的藥物處方經(jīng)由計算機系統(tǒng)傳輸，而訪問這些計算機系統(tǒng)的是其他專業(yè)人員，而不是計算機科學專業(yè)的人。倡導病患安全的人士大聲提醒我們，我們是人類失誤和機器故障的潛在受害者，因此在吞下藥片之前我們需要想一想，問一問。但是沒有多少人聽從他們的建議，開錯藥的現(xiàn)象也愈演愈烈。

更隱蔽的危險是，我們創(chuàng)造的科技假定我們擁有超人的認知能力，而且消費者似乎愿意一直依照這樣行事。面前放上一部手機、一杯拿鐵，手握方向盤，以每小時80英里的速度奔馳著，我們是自豪的一心多用能手。當然，別人用手機打電話是危險的，而你則能自如轉換注意力，及時發(fā)現(xiàn)任何突發(fā)的驚險狀況，對不對？幾年前，一位身穿黑猩猩服裝的心理學家批判了這一想法，可我們卻捶擊胸口，表示輕蔑。我們在看了一包15磅重的香腸從飛機上掉落砸穿某人屋頂?shù)男侣労?，依然幻想自己能在復雜的情形下精準評估風險。我們甚至認為自己戴著耳機時也擅長看到迎面而來的火車。

以前我們如果搞砸了，危險會給出即時反饋。當我們對熊做出錯誤的判斷，熊會立刻讓我們知道。然而科技使我們過失所造成的后果沒有即刻顯現(xiàn)。延遲反應、復雜的連鎖反應、潛在反應——這些才是科技運作的部分。但是由于已經(jīng)適應了迅速反饋的世界，我們幾乎不會在使用自動駕駛儀巡航一天的中間停下來思考之后可能會出什么問題。更糟的是，科技甚至在非法入侵我們的反饋系統(tǒng)。開車看手機時，我們瞬間感覺多巴胺飆升。伊迪絲.哈博的公司Launch Darkly專門研究對新科技面向大眾的發(fā)布加以控制。她指出我們正在創(chuàng)造一種不平衡的行為和獎勵系統(tǒng)，非常危險：“我們并不是每次安全行事都能得到冰淇淋甜筒?！?/p>

科技甚至在重塑我們的安全文化。我爺爺是個手藝人，他教會我如何使用工具。以前我如果使用梯子不當，會立刻被糾正。但是后來，科技使一多半祖父母輩的人轉而做辦公室工作。如今，重新使用工具的人數(shù)創(chuàng)下了新的紀錄，因為自己動手其實很酷，同時可以說對你的內心有益處。但是現(xiàn)在我們在使用工具時卻更容易受傷了。

很少有人停下來體會我們對熱心敬業(yè)的消費產(chǎn)品設計師和不停訴訟他們的律師的工作依賴程度之深。然而今天，我們發(fā)明東西的速度比我們設計安全特征的速度要快。不久前德國的一個小鎮(zhèn)在人行道上安裝了一種交通信號燈，用于提醒埋頭看手機的行人。但是等可穿戴設備和虛擬現(xiàn)實裝備出現(xiàn)后怎么辦？屆時，我們會仰著頭通過設備看，而不是低頭看。

甚至連政府都代表我們介入去制定安全標準。可如今我們的科技發(fā)展速度甚至可能超過了我們制定新標準的速度。在電影《王牌播音員2：傳奇繼續(xù)》中有這樣一個場景：威爾.法瑞爾打開了房車的定速巡航器，然后離開駕駛員座椅。當保羅.拉德指出定速巡航器只能控制速度而非方向時，混亂接踵而至。而這種混亂可不只是好萊塢喜劇中的日常場景。美國猶他大學的凱利.芬克豪澤最近列了一份清單，把汽車制造商給速度控制器、車道保持系統(tǒng)、盲點監(jiān)控功能起的名字都列了出來。她發(fā)現(xiàn)各個生產(chǎn)廠家對于這些功能的不同組合給出了相似的名字。上錯租來的車、按下按鈕、匆匆掏出手機，然后“砰”的一聲……你就成了王牌播音員。

所以倘若你真擔心召喚到惡魔，請放松。已經(jīng)為時已晚了。因為危及生命和肢體的科技時代已經(jīng)來臨了。但是我們從先輩那里遺傳下來的那些脾性可能會妨礙我們改變行事方法，甚至讓我們看不明白為什么要改變。在硅谷——一些非常聰明的人在那里居住和工作——我看到過一個駕駛員一邊用一部手機打視頻電話，一邊用另一部手機發(fā)短信。與此同時，死亡人數(shù)持續(xù)上升，而且因此導致的交通事故是目前美國排名第四的死亡主因。（據(jù)記載，和熊相關的意外死亡人數(shù)則保持相對穩(wěn)定。）

沒有哪個神奇的硅谷小玩意兒會解決這個問題。相反，對于怎樣才算小心謹慎，我們需要重新考慮——去了解我們如今需要行進其中的一個更復雜世界的細微差別，去認識我們那有創(chuàng)意的、善于分析的、時不時犯錯的頭腦能否僥幸逃脫的極限?，F(xiàn)代世界正在打破我們有關如何才能保持安全的知識。倘若我們還心存希望，以為可以做好準備，抵御機器的崛起，那就需要學會幸存于現(xiàn)有的這些東西。endprint