語音交互是未來？未來何時到來？譯

最新 07-27

Propelledby what we can legitimately call The Alexa Movement, voice is now perceived as the future of the User Interface. But we need numbers: How many of us continue to use Alexa (or Siri, or Google assistant) afterthe novelty has worn off? What do we use it for?This will help us understand the likelyactualfuture of Voice UI.

For decades, we』ve been promised voice interactions with our machines — promises that have beeniconizedin science fiction, from theobliging「Computer!」 in theStar Trekfranchise to themalevolentHAL in2001 Space Odyssey(「I』m sorry Dave, I』m afraid I can』t do that」).And, until recently, science fiction is where real-world attempts at giving our machinethe gift of gabhave remained.

多年來，我們一直被許諾將擁有配備語音交互的機器設備。這種期待在科幻小說中，已被神往了很長一段時間----從星際迷航系列中樂於助人的Computer，到2001太空漫遊中兇惡的HAL（「對不起Dave，我恐怕做不到」）。直到現在，科幻小說仍是現實生活中嘗試讓機器擁有勝於雄辯的交流能力的最佳地點。

For decades we』ve had almost yearly 『this time we』ve got it』 announcements followed by dead silence (does anyone remember Ubi)? Acheekyreporter once toured the offices of product reviewers who had touted the hands-free benefits of voice-recognition advances. We know what the reporter found: None of the visionary reviewers used the new Voice UI.

這些日子以來，我們幾乎每年都要發出「這次，我們終於要實現了！「諸如此類的聲明，而每每台下都是一片死寂（有人還記得Ubi嗎？)。一位記者曾厚著臉皮地參觀了一群產品評論人的辦公室，他們均是對語音識別技術的擁護和提倡者，認為語音識別的發展為人們帶來了雙手解放的巨大便利。然而我們已知道記者最後將會發現什麼：所有「有遠見」的產評人中根本無人使用語音交互。

Nonetheless, the desire has remained intact. The ability to speak a command is a natural and obvious evolution in the way we interact with our machines. Just as the graphical UI created at Xerox Parc freed early computers from the unforgiving Command Line Interface (CLI), the promise of a Voice UI frees us from manual touch, it』s a third hand when two are busy, letting us control thermostats, navigation systems, music selection…

但不論如何，這份願景還是被完整地保留下來了。在我們與機器交互的過程中，口述操作命令的交互形式變革本就是順應本能，顯而易見的。正如Xerox Parc發明的圖形化交互將早期電腦從命令行交互（CLI）中解放出來一樣，聲音交互也即將使我們從手觸屏幕交互的形式中解放出來。當用戶兩手無暇響應操作時，聲音將成為我們的「第三隻手」，以幫助操控溫度調節，導航系統及音樂選擇等等。

For people with neuromotor challenges, Voice UI isn』t just a convenience, it restores power and dignity. Also, lack of literacy becomes less of a roadblock. You no longer need to be able to read and write to get things done, a liberating influence in parts of the world where education is still developing.

而對患有運動神經障礙的人群來說，語音交互提供的不僅僅只是便捷，它更歸還這個群體應有的尊嚴和權利。同樣，文字通識教育缺失也漸漸不再是學習使用機器設備的障礙。你不必再通過讀或寫來處理事件，這對那些教育仍然處於發展中階段的國家將帶來解放性的變革。

Today, Voice UI is no longer just a promise. Amazon rightly gets credit for both its vision and impeccable implementation of the technology.

今天，語音交互不再是一種期待或願景。亞馬遜對該技術未來發展規劃和無懈可擊的技術實施已獲得公眾認可。

In late 2014, Amazon introduced Echo, a line of voice-activated 「smart speakers」. Prices range from less than $40, on a good day, to just under $230 for the screen-equipped version.

2014年年末，亞馬遜向社會發布了以語音交互為主導的智能音箱---Echo系列。該系列的價格從低於40美金（如果趕上好時候），到230美金左右（帶液晶屏版本）不等。

But prices tell a less important part of the story. Bezos & Co.made an early bet onArtificial Intelligence (AI),endowingAlexa, Echo』s intelligent voice responder in the Cloud,withan ever-increasing range ofSkills,i.e. response modules proposed to users. With the enthusiastic support of third party developers, Alexa counted more than 15,000 Skills by last July.

但是價格並不是這個故事的主線。Bezos & Co. 在早期便看好人工智慧領域（AI）的發展並投入了大量資源，因此Echo的雲智能語音應答Alexa的語音技能（如其用戶響應模塊）範圍得以不斷增長。截止到去年七月，Alexa就已擁有15000多個語音技能了。

To top it off, Amazon』s support for new users is considerate and actually helpful. The company provides a well-designed Skills Guide that orients users in this new world without menus, mice, and keyboards, and also helps existing Echo owners expand their enjoyment of the Alexa universe.

除此之外，亞馬遜對新用戶的幫助也十分貼心。公司設計了一套技能指南，旨在引導新用戶，能在脫離菜單，滑鼠和鍵盤的情況下，輕鬆地進入交互新世界，同時，現有Echo用戶也能在「Alexa宇宙」中體驗到更多科技帶來的愉悅。

And to top this off, Amazon publishes a weekly newsletter to Echo owners that demonstrates both new and road-tested ways to speak with Alexa. Keep in mind this isn』t your typical email peddling a new robot floor cleaner, or urging you to re-order the laundry detergent you bought last month. No, every single week, Amazon tells Echo users how to make better use of a purchase they made months or years ago. Amazon』s practice indirectly reveals Apple execs haven』t bought Echo devices. Otherwise, gettinghebdomadalemails from Jeff Bezos would have shamed them into writing weeklymash notesto their beloved Mac, iPhones and iPads users.

更有甚之，亞馬遜會每周向Echo用戶推送訂閱新聞，這證明了其一直致力於研究推廣與Alexa對話的新方式。記住，這不是你收到過的那種兜售新型掃地機器人的郵件，更不是敦促你重新訂購上個月剛剛購入的洗衣機的廣告。每周，亞馬遜都會告訴Echo用戶如何更好地去利用並優化他們數月前甚至數年前購入的產品。亞馬遜的這種行為也不經意地間接表明蘋果公司的主管們並沒有購入任何Echo設備的事實，否則，當他們每周收到Jeff Bezos推送的郵件時，他們將對自己每周向「摯愛的」Mac，iPhone和iPad用戶寫肉麻的恭維信的行為感到羞辱。

Voice UI is great progress, even if the technology feels a bit stilted and is occasionally infuriating. One challenge is that the smart device just sits there awaiting our commands, doing little or nothing to let us know what commands it understands and how precisely they ought to be formulated. Also, voice assistants generally don』t pass the Turing Test, meaning they can』t really fool us into believing we』re conversing with a human.

語音交互是一個巨大的進步。即使目前在使用上，它仍讓用戶感覺有些許生硬且偶爾讓人惱怒。該技術目前面臨的一個挑戰是智能設備通常只能被動地等待用戶指令，很少或不能向用戶傳達它們能理解什麼樣的指令或者指令組成的精確程度。語音助手之類的軟體也基本無法通過圖靈測試，這意味著這些設備還足以真正愚弄我們，進而致使我們產生一種正在與人類對話的錯覺。

A bigger frustration for those of us who are interested in the future of the technology, is that Amazon (and Google and Apple) are playing it close to the chest when it comes to numbers. How many Echo devices have been sold? How often are they used on average: Ten times a day? Five times a week? Almost never? Who (age, occupation) uses them the most and for what?

而對於密切關注未來科技發展的一部分群體來說，更大的失望在於，亞馬遜對數據採取秘而不宣的態度。Echo設備的銷量為多少？設備的平均使用頻率：十次一天？或是一周五次？或基本不用？設備的主要使用群體（年齡，職業）？主要用於完成哪些任務？

Amazon knows all this but keeps this fascinating knowledge to itself — and so we turn to the 「market analysts」. A survey from late 2016 (eons ago in tech time) found that 「the top feature tried by Echo users is the very simple act of setting a timer」. A more recent study says only that Echo users 「buy more stuff」.

亞馬遜對這些問題的所有答案瞭然於心，但它從不將這些充滿吸引力的數字公之於眾。尋果無門，我們只好轉而求助於「市場分析師」。根據2016年年末的一份調查報告說明：Echo用戶最常使用的一個功能僅僅是簡單設置時鐘。「而一份更新的調查則表明只有Echo用戶會購入更多的東西。

(As an aside: Serious investigation is, of course, complicated and expensive. You need a large sample, say 1000 people, to achieve a decent confidence interval for the results, and the participants need to reliably represent the user populationat large. I seriously doubt that most research 「reports」 caroming in the on-lineecho chambermeet the above criteria. For example, the 「top feature is setting a timer」 conclusion was based on a survey of 180 Echo owners.)

（說點題外話：嚴謹的市場調查當然會更為複雜和昂貴。你需要數量更大的樣本，比如說1000人，來獲得一個良好的置信區間，且調查參與者需要可靠地代表大多數使用群體。我嚴肅質疑大部分這些言之鑿鑿，蕩氣迴腸的研究報告是否達到了上述標準。比如在上述Echo用戶最常用的功能的調查中，設置鬧鐘這一結論僅基於180位Echo用戶的反饋而得到。）

The same lack of numbers applies to Google Assistant. Android smartphones are very successful, and there are hints that Google Assistant does wonders in countries where literacy isn』t yet universal — India comes to mind — but we』re left in the dark when it comes to quantifying uses, successes and failures…

Google助手對數據也採用了同樣的態度。安卓手機如今取得了巨大的成功，而跡象表明谷歌助手在通識教育並不普及的國家中表現尤其驚人——比如印度——但我們對使用量，成功率和失敗率這些仍然一無所知。

For Apple, we have a decent idea of the number of Siri-capable devices, it』s the number of iOS devices that are less than four-years-old (give or take). Similar to Amazon』s Skills Guide, Apple』s provides a neat Siri onboarding site, a clean and well-lighted place where users new and old can navigate the sea of possibilities… But how is Siri being used? When, where, by whom, how often? We』re not told.

對於蘋果，我們有個不錯的主意來得到能使用Siri的設備數量——通過計算機齡小於四年的iOS設備的數量。類似於亞馬遜的技能指南，蘋果同樣向廣大新老用戶提供了一個Siri入門使用的網站。該網站小巧整潔，信息整理井然有序，是個探尋Siri使用的無限可能性的絕佳地點。但究竟Siri是如何被用戶使用的呢？時間？地點？使用者？使用頻率？我們還尚且未知。

Why do we need to see these numbers? Just out of curiosity, of course, but also to get a feeling for how the Voice UI wars will play out. Today, Amazon is on top, but will it remain there? The three big contenders, Amazon, Google, and Apple, have significantly different business strategies:

我們為什麼需要這些數字？當然，可能僅僅是出於好奇，但是也同樣想通過數據感知這場語音交互的戰爭將何去何從。今天，亞馬遜的優勢遙遙領先，但它會一直保持這個有利的地位嗎？三大競爭巨頭，亞馬遜，谷歌和蘋果，有著顯然不同的商業策略。

Because of its current domination of Voice UI and e-commerce — asynergyno one saw coming — Amazon is uniquely positioned for further episodes of conquest. But in order to win theinstalled baserace against Google Assistant and Siri, Amazon needs to keep installing more Echo or Alexa-capable devices.

由於在語音交互和電商領域的統治地位——一個無人預見的協同增效作用即將到來——亞馬遜將在未來語音交互的戰場上佔據獨特優勢的主動地位。但為了贏得對陣谷歌助手和Siri的用戶基數之爭，亞馬遜需要繼續擴充安裝Echo或Alexa適配的設備。

Google doesn』t sell goods to 「civilians」, it makes its money by selling advertising tools to merchants. While Google Assistant has the reach of Android devices, it doesn』t have Amazon』s unique integration of advertising and e-commerce.

谷歌並不向終端用戶銷售實體貨品，它通過向企業出售廣告營銷工具來盈利。雖然Google助手能通過安卓設備被大眾使用，但它並不擁有如同亞馬遜廣告電商相整合的獨特資源。

Apple doesn』t sell laundry detergentoradvertising — it makes its money by selling devices. Like all other components of Apple』s ecosystem — iTunes, the App Store, the Services business with its rapidly-growing revenue — Siri』s only purpose is to help sell more iPhones, Macs, iPads, and Watches.

蘋果既不銷售諸如洗衣液的實體商品也不以廣告營銷為生——它的利潤來自於電子設備。正如蘋果生態圈中的所有其他組件——iTunes，蘋果商店，以及增收迅速的服務——Siri的唯一目的是幫助公司賣出更多的蘋果手機，電腦，平板以及手錶。

Three contenders, three different areas of domination. Which one will win? Ask again later.

三位有力競爭者，盤踞於三個不同的領域。誰將贏得最後的戰爭？這個問題還是交給未來回答。

Regardless of what happens in the future, one can』t help being impressed with Amazon』s Voice UI performance and wonder what it means for Google』s advertising business and itsforaysinto connected devices.

但不管未來如何變化，我們還是驚嘆於亞馬遜不俗的語音交互表現，同時也好奇著這對於經營廣告營銷業務並剛剛開展其互聯設備業務的谷歌又意味著什麼。

In the meantime, absent numbers, we have to pause before makingloftyclaims for the future of Voice UI. Is it, to quote a respected expert, 「THE […] critical piece of the puzzle for the next era of computing」? Perhaps. But we ought to remember the small matter of timing, best captured by Horace Dediu』s apothegm: Those who predict the future, we call futurists. Those who know when the future will happen, we call billionaires.

同時，由於準確數據的缺失，我們不得不在對語音交互的未來發表高瞻遠矚的預測前謹言慎行。引用一位受人尊敬的專家的話，這會是「下一個計算時代的拼圖中最重要的一塊兒嗎「？也許吧。但我們還應該明白，時機同樣有著不小的影響。Horace Dediu在他的格言中就有對其最準確貼切的描述：能預測未來的人，我們稱之為預言家，而知道這些未來何時發生的人，我們稱之為億萬富翁。

We will not be bored.

我們將不會感到厭倦和無聊。

譯者說：

以下發言將分為三段：

一，這是一次不小的嘗試。在未來的日子裡，這個專欄會不定時地更新一些作者個人的翻譯作品。這些作品一定程度上都為UI和數據可視化領域中的熱點問題或未來趨勢。

二，翻譯作品中難免出現缺翻漏翻，或詞義不當等問題，歡迎各位盆友在下方評論欄向我及時指正或交流討論！歡迎一切形式的批評建議！

三，文中將一些我個人認為的生詞進行了標註，希望和大家共同進步！

原文出處：

https://mondaynote.com/voice-ui-is-the-future-but-when-747fe1f74cf3

mondaynote.com

喜歡這篇文章嗎？立刻分享出去讓更多人知道吧！