2019年经济学人 约翰逊语言专栏--机器写手无法完全取代人类(3)

 

This is possible at the moment only on a very limited scale. Ask Apple’s Siri or Amazon’s Alexa for a single fact—say, what year “Top Gun” came out—and you will get the answer. But ask them to assemble the facts to prove a case, even at a straightforward¹ level— “Do gun laws reduce gun crime?”—and they will flounder. An advance in integrating knowledge would then have to be married to another breakthrough: teaching text-generation systems to go beyond sentences to structures. Mr Seabrook found that the longer the text he solicited² from GPT-2, the more obvious it was that the work it produced was gibberish. Each sentence was fine on its own; remarkably³, three or four back to back could stay on topic, apparently⁴ cohering⁵. But machines are aeons away from being able to recreate rhetorical and argumentative flow across paragraphs and pages. Not only can today’s journalists expect to finish their careers without competition from the Writernator— today’s parents can tell their children that they still need to learn to write, too. Aside from making scribblers redundant⁶, a common worry is that such systems will be able to flood social media and online comment sections with semi-coherent but angry ramblings that are designed to divide and enrage⁷. In reality, that may not be much of a departure from the tenor⁸ of such websites now, nor much of a disaster. Perhaps a flood of furious auto-babble will force future readers to distinguish between the illusion of coherence⁹ and the genuine article. If so, the Writernator, much like the Terminator, would even come to do the world some good.

目前，这仅在非常有限的范围内是可能的。如果你问苹果的Siri或亚马逊的Alexa一个简单的事实——比如《壮志凌云》在哪一年上映——你就会得到答案。 但要求电脑收集事实来证明一个案件，即使是在一个简单的层面上——“枪支法能减少枪支犯罪吗?”——电脑就会解释不清楚。在整合知识方面的进步必须与另一个突破相结合：训练文本生成系统超越句子层面注重结构层面。西布鲁克先生发现，他向GPT-2索取的文本越长，就越明显地看出，GPT-2所做的文章是胡言乱语。每个句子单拿出来都很好；值得注意的是，三四个连续的句子还可以保持话题的连贯性。但机器要想在段落和页面中重现修辞和论证流，还需要很长时间。如今的记者不仅可以在没有作家竞争的情况下完成职业生涯，如今的父母也可以告诉孩子，仍然需要学习写作。除了让乱写乱画的人变得多余外，人们普遍担心的是，这样的系统将能够以半连贯但愤怒的漫谈淹没社交媒体和在线评论区，达到分裂和激怒他人的效果。

1 straightforward   

adj.正直的，坦率的；易懂的，简单的

参考例句：

• A straightforward talk is better than a flowery speech.巧言不如直说。
• I must insist on your giving me a straightforward answer.我一定要你给我一个直截了当的回答。

2 solicited   

v.恳求( solicit的过去式和过去分词 )；（指娼妇）拉客；索求；征求

参考例句：

• He's already solicited their support on health care reform. 他已就医疗改革问题请求他们的支持。 来自辞典例句
• We solicited ideas from Princeton University graduates and under graduates. 我们从普林斯顿大学的毕业生与大学生中征求意见。 来自辞典例句

3 remarkably   

ad.不同寻常地，相当地

参考例句：

• I thought she was remarkably restrained in the circumstances. 我认为她在那种情况下非常克制。
• He made a remarkably swift recovery. 他康复得相当快。

4 apparently   

adv.显然地；表面上，似乎

参考例句：

• An apparently blind alley leads suddenly into an open space.山穷水尽，豁然开朗。
• He was apparently much surprised at the news.他对那个消息显然感到十分惊异。

5 cohering   

v.黏合( cohere的现在分词 )；联合；结合；（指看法、推理等）前后一致

参考例句：

• It is necessary for cohering public feeling and inspiring fighting will to cultivate and publicize models. 摘要积极培养与宣传典型对凝聚人心、鼓舞斗志、推动工作很有必要。 来自互联网
• Objective To explore the relation between meridians cohering on the organism and channel and point-entrails effect. 目的探讨经脉线的附着组织与经穴—脏腑效应之间的关系。 来自互联网

6 redundant   

adj.多余的，过剩的；(食物)丰富的；被解雇的

参考例句：

• There are too many redundant words in this book.这本书里多余的词太多。
• Nearly all the redundant worker have been absorbed into other departments.几乎所有冗员，都已调往其他部门任职。

7 enrage   

v.触怒，激怒

参考例句：

• She chose a quotation that she knew would enrage him.她选用了一句明知会激怒他的引语。
• He started another matter to enrage me,but I didn't care.他又提出另一问题，想以此激怒我，可我并没在意。

8 tenor   

n.男高音（歌手），次中音（乐器），要旨，大意

参考例句：

• The tenor of his speech was that war would come.他讲话的大意是战争将要发生。
• The four parts in singing are soprano,alto,tenor and bass.唱歌的四个声部是女高音、女低音、男高音和男低音。

9 coherence   

n.紧凑；连贯；一致性

参考例句：

• There was no coherence between the first and the second half of the film.这部电影的前半部和后半部没有连贯性。
• Environmental education is intended to give these topics more coherence.环境教育的目的是使这些课题更加息息相关。