瓦萊麗·赫德森

Playing with ChatGPT from Open-AI has become all the rage1 at the moment. This online correspondent powered by artificial intelligence will attempt to answer your questions with a paragraphs worth of knowledge2， write lyrics or pen stories from prompts you suggested， and more， with all responses derived from information processed by algorithms.

目前，跟OpenAI所研發(fā)的ChatGPT互動已成了潮流。這位網(wǎng)絡筆友由人工智能驅(qū)動，它會把知識組織成段落來回答你的問題，根據(jù)你給的提示詞寫歌詞或?qū)懝适拢鹊?，而它所有的回答都源自?jīng)算法處理過的信息。

ChatGPT writes on the level of sophistication of about a sixth grader， Id say， scraping the vast amount of online knowledge to do so.

依我看，ChatGPT的寫作水平與六年級學生相當，它在網(wǎng)上搜刮了大量知識來完成寫作。

Not surprisingly， people are worried about the bias of ChatGPT： online source material can be very much in line with favored narratives， especially as social media companies have been revealed to actively bury certain viewpoints in favor of others， even burying particular unwelcome facts.

人們擔心ChatGPT會懷有偏見，這并不奇怪：網(wǎng)絡原始材料和大眾偏好的敘事高度一致，尤其是社交媒體公司已遭到曝光，稱它們主動隱藏某些觀點、推崇另一些，甚至隱藏某些不受歡迎的事實。

The world is starting to react. Schools are now deliberating whether to embrace ChatGPT or view it as plagiarism. Artists are suing， as lyrics and visuals created are all by definition derivative of works by human beings. Lobbyists may find their work much easier， which may not be to the public good. Troublingly， humans are also turning to ChatGPT as a kind of ouija board3， asking AI what they should do in their personal circumstances， such as whether to divorce a spouse.

全世界都開始作出反應。學校在考慮是欣然接受ChatGPT，還是將其視為剽竊。藝術家們紛紛起訴，因為AI創(chuàng)作的歌詞和視覺資料，按定義說都衍生自人類作品。說客們也許發(fā)覺工作更輕松了，可這對公眾利益不一定是好事。令人擔心的是，還有人把ChatGPT當作某種占卦板，向AI咨詢?nèi)绾翁幚硪恍┧饺藛栴}，比如要不要與配偶離婚。

Theres some larger questions here， though. With the advent of ChatGPT， there has been increasing discussion about whether Western civilization is moving into a “post-literate” era. Were at a 40-year low in the U.S. in terms of young people reading for pleasure， according to Pew Research4. Bosses complain their younger employees boast that they dont read emails—even work emails—at all. Universities are dropping requirements for standardized test scores and even personal statements from applicants. Short TikTok videos and 280-character tweets are the limited and limiting daily fare of the rising generation.

不過，還有一些更嚴峻的問題。隨著ChatGPT的出現(xiàn)，越來越多的人討論西方文明是否正進入“后文字”時代。皮尤研究中心的數(shù)據(jù)顯示，在美國，為樂趣而閱讀的年輕人的數(shù)量處于40年來的最低點。老板們抱怨說，他們的年輕員工夸口說壓根不看電子郵件——連工作郵件都不看。大學正逐步降低對申請者標準化考試成績乃至個人陳述的要求。國際版抖音短視頻和280個字符以內(nèi)的推文是新一代人每日有限的信息源，其內(nèi)容還有局限性。

Its hard not to see a post-literate Western society as the advent of a new type of Dark Ages5. What we know occurred is that the knowledge， learning and thinking skills accrued during the golden ages of Greece， Rome and Arabia were largely lost to the rising generations through war and destruction. This time around， these things are being lost to our advances in technology—but the effect is still the same.

處于后文字時代的西方社會，很難不被視為一個新型黑暗時代的前奏。我們知道，希臘、羅馬和阿拉伯的黃金時代所積累的知識、學識和思維能力，在戰(zhàn)爭和破壞中大量失傳于后人。這一次，這些事物正是因我們的技術進步而散佚——但結果是一樣的。

What is being lost？ First， the younger generation does not commit know-ledge to memory6; why should they？ The internet serves as their memory. But without substantive knowledge banked in your own memory， you cannot think well. For example， I remember when a colleague told me that our masters students in international affairs did not need to know where Afghanistan was on a map because they could just Google it. I would argue that if you do not have a mental idea of where Afghanistan is and what nations surround it， you simply cannot think at a sophisticated level about Afghanistan.

哪些東西在消逝？第一，年輕一代不把知識記在腦子里；他們何必去記呢？互聯(lián)網(wǎng)就是他們的記憶中樞。但如果腦中沒有充實的知識儲備，人就無法進行透徹的思考。舉個例子，記得有位同事告訴我，我們國際事務專業(yè)的碩士生無需知道阿富汗在地圖上的位置，因為可以直接上谷歌搜索。我要說，如果你腦海里對阿富汗在哪兒、周邊有哪些國家沒有概念，你就不可能從深層次思考阿富汗問題。

And if you outsource your personal knowledge base to Google or AI， you may or may not receive factual information， as my little experiment demonstrated. The line between fact and opinion， always contested， is now being given up for lost. Whats true according to ChatGPT depends on what part of the internet it has happened to scrape at a particular time. And that in turn is dependent on what social media companies boost， and what they bury. The idea that we would seek a truth that stands independent of our viewpoint is considered quaint， even silly. So much for7 the Renaissance.

正像我的小實驗所證明的那樣，如果你把個人知識庫外包給谷歌或者AI，你可能會得到事實性的信息，也可能不會。事實和觀點之間的界限始終有爭議，現(xiàn)在人們正在放棄劃清這條界線的希望。ChatGPT認定的真實，取決于它在某時刻碰巧搜羅了互聯(lián)網(wǎng)的哪一部分。而這又取決于社交媒體公司宣傳什么，又遮掩什么。尋求不受個人看法影響的真理成了老古板，甚至愚蠢之舉。文藝復興就別指望了。

Second， allusion is lost. The cultural touchstones that enriched our society are just gone. If I used the phrase， “Ask not what your country can do for you，” many of my students would not know it was JFK who immortalized that phrase. If I were to say， “Blessed are the peacemakers，” many of my students would not know it was Jesus Christ who said this. And dont even think about a Shakespearean8 allusion.

第二，典故在消亡。賦社會以豐沛的文化試金石不復存在。如果我引用“不要問國家能為你們做些什么”這句話，我的許多學生不會知道是肯尼迪讓它流芳百世。如果我說，“締造和平的人是有福的”，許多學生不會知道這是耶穌基督說的。更遑論莎士比亞作品中的典故了。

Third， deep reading and critical thinking skills are lost. The most important arguments cannot be distilled into 280 characters. These arguments are not only deep， but they are wide and nuanced， and have many dimensions. And it is in that richness that critical thinking becomes possible， as one dives beneath the thin surface of an argument to its complex roots to understand where the weaknesses are. In our time， there is but shallow understanding， for all nuance has been lost. Indeed， even debates on controversial issues， where one could hear critical thinkers engage each others viewpoints， are now anathema because if there is only surface， what could there be to debate？

第三，深度閱讀能力和批判性思維能力在喪失。最重要的論證沒法凝練成280個字符。這些論證不但高深莫測，也寬廣無際、細致入微，有諸多維度。正是在此種詳實的論證中，批判性思維成為可能，因為人可以從某條論點薄薄的表面層深潛下去，遁入其復雜的根源，以了解這一論題的弱點所在。我們的時代空有膚淺的理解，因為一切細微的差別業(yè)已喪失。事實上，爭議性問題的辯論中本可耳聞批判性思想家們的觀點交鋒，如今卻也令人極度反感，因為空談表象，還有什么可辯論的呢？

Last， of course， is that all these things combine to create not only extremist ideologies divorced from reality， but these in turn create mobs who are prepared to literally or figuratively rend dissenters limb from limb in the quest for purity of thought.

最后一點當然是，以上所有因素結合在一起，不僅制造出脫離現(xiàn)實的極端主義意識形態(tài)，而且這些意識形態(tài)進而制造了群氓，他們?yōu)榍笏枷氲募兇庑裕瑴蕚鋸淖置嫔匣蛳笳饕饬x上，將異見者撕成碎片。

All of this also means that those who continue to read and to practice critical thinking skills will increasingly be unable to effectively communicate with those who do not.

這一切還意味著，繼續(xù)閱讀、繼續(xù)訓練批判性思維技能的人，將越來越無法和不讀不練的人有效溝通。

Some have suggested that the only way forward is to emulate those histor-ical oases from the Dark Ages： the monasteries—more particularly， an offline， hard-copy repository of the worlds knowledge that can be resurrected by its keepers once the current Dark Ages have passed. I do my small bit by insisting my students memorize a map of the world and important historical facts， figures and dates.

有些人認為，唯一的出路是效仿黑暗時代的那些歷史綠洲：修道院——更確切地說，一個離線的、紙質(zhì)的世界知識寶庫，待當前的黑暗時代過去，它的保存者就可以使其復活。我堅持讓學生識記世界地圖和重要的歷史事實、數(shù)字、日期，以盡我綿薄之力。

Ill end with an allusion—see if you recognize it. “Do not go gentle into that good night ... Rage， rage， against the dying of the light.”9 In your own way， and however you can， keep the light alive during these new Dark Ages.

我將以一個典故收尾——看看你認不認得出。“不要溫和地走進那個良夜……怒斥，怒斥光明的消逝。”用你們自己的方式，不管是什么樣的方式，在新黑暗時代讓光長明不滅。

（譯者為“《英語世界》杯”翻譯大賽獲獎者）

1 be all the rage風行一時，十分流行。? 2 a… worth of sth像……那么多的東西。

3 ouija board通靈板，占卜板。? 4美國的一間獨立性民調(diào)機構。? 5黑暗時代（歐洲歷史上從羅馬帝國衰亡至公元 10 世紀的時期）。

6 commit to memory牢記。

7 so much for（表示行不通或沒用）作罷好了。? 8 Shakespearean由莎士比亞寫的；與莎士比亞有關的；典型的莎士比亞式的。

9該詩句出自《不要溫和地走進那個良夜》，由英國詩人狄蘭·托馬斯創(chuàng)作于20世紀中期。此處采巫寧坤的譯本。