Will Google Translate’s latest update end professional translations?

最新版谷歌翻译会终结人工翻译吗?

2020-05-21 20:02 newsbreak

本文共961个字,阅读需10分钟

阅读模式 切换至中文

What will Google Translate's recently announced AI-powered update mean for professional translations from human employees? Will AI really replace professional linguists? Google Translate is on the cusp of releasing a brand new feature which will allow consumers to transcribe and translate a conversation in real-time using just their smartphones. At a San Francisco press preview in January 2020, Google demonstrated how the software can record audio and create translations in text format into another language. As well as offering instant results, this live transcription feature also enables the program to continuously update its translation based on the context of what else is being said, which will offer significant improvements to the accuracy of the finished document. Though there is no fixed launch date — Google has only revealed that the feature will be released “in the future” — this is an exciting development in AI translation technology, but one which has raised serious questions over the future of professional linguistic services. If users can obtain instant language solutions without paying a penny, is there still a place for expert linguists in the translation market? Improved accuracy and instant results have mass appeal It’s not hard to see why Google Translate’s latest update is so appealing to consumers. The software remains free, easy to use, and provides users with instant translations. The key difference is that the app is poised to produce even more accurate results, and despite the fact foreign language learning in the UK is at its lowest level since 2000, more people than ever need to communicate in different languages, making the app particularly useful. When it comes to far-flung holidays, over 60% of 16-34 year-olds use their smartphones to translate the local language, and as noted by Raconteur, AI-enabled apps like Google Translate have “helped bring down the cost of translation dramatically” for businesses working on an international scale. In 2018, Google’s CEO Sundar Pichai reported that the program was translating 143 billion words daily, but the meticulous power of its new technology means Google Translate’s popularity will surely grow and attract even more users. Better technology doesn’t eliminate mistakes These impressive technological developments also mean that Google Translate is going to produce more accurate content than it ever did before. Its live transcription feature will constantly evaluate sentences, continuously correcting word choices, adding punctuation, and taking accents and regional dialects into account. This means that, unlike word-for-word translations, Google will be able to use context to inform its output. As encouraging as this may be, it’s important to remember that AI translators can still make mistakes, and machines will never understand human speech in the same way that people do. Natural language processing gives a voice to digital processes One of the greatest challenges we face in the digital generation is building the bridge between the abilities of the human thought process and the possibilities that lie within autonomous technology, particularly how we teach our human-like understanding of words in context to machines. Natural language processing (NLP) is a step towards developing this autonomous ‘thinking’ capability. Read here The odd error isn’t too problematic in menial circumstances, like a tourist trying to translate their dinner order. However, for crucial materials like conference minutes and business documents, nothing less than a perfect translation will be acceptable. There’s no guarantee that Google will pick up on the context and nuance which underlies everyday language, and the program may also fail to register tone of voice, perhaps missing hints of sarcasm, compassion or other distinctly human states and emotions. These seemingly small omissions are particularly risky within complex industries like medicine and engineering, particularly if Google’s software is unable to fully grasp the meaning of sector-specific terminology. For example, in the case of financial translations, professional linguists Global Voices insist that accuracy is about far more than just numbers: “Financial translators should understand the industry inside out, and be aware of new developments in the sector, as well as being fluent in the target language.” Google Translate will not be well-versed in advanced financial concepts, nor be aware of the industry laws and practices of every country for which it offers translations. Only a professional translator with a learned combination of language fluency, cultural awareness and sector expertise will be able to produce an impeccable translation in these circumstances. Context and purpose continue to drive translation needs Google Translate’s latest update is clearly a major breakthrough in the world of AI translation software and it’s reasonable to predict that this will probably be the first port of call for linguistic queries. However, this certainly doesn’t mean we’re heading towards a future where professional translations no longer serve a purpose. Artificial intelligence in marketing: when tech converges with the traditional Artificial intelligence in marketing is more common than you would think and today, the CMO and CTO are working more closely to make it a success. Read here Like many other automated services, certain needs will be adequately met by a computer program, while others will require a human touch. For casual use, Google Translate will remain a massively convenient tool, and this new update will make real-time communication across different tongues easier than ever. The instant removal of language barriers offers a valuable and accessible means of enabling connection and boosting cultural understanding. Nevertheless, for professional purposes, authorised human translation services still come out on top. In any project where mistakes and misunderstandings could wreak havoc, expert linguists can offer the skill, talent and insight to produce faithful and flawless translations, no matter how complicated the content. Global Voices is a translation and interpretation provider that works with over 3,000 professionally qualified linguists globally.
谷歌翻译最新版加入了人工智能,这对人工专业翻译来说意味着什么? AI真的会取代专业语言学家吗? 谷歌翻译即将发布一项全新功能,让消费者只需使用智能手机就能实时转录和翻译对话。 在2020年1月的旧金山媒体预览会上,谷歌演示了该软件如何录制音频,并以文本格式将其翻译成另一种语言。 除了提供即时结果外,这种实时转录功能还使翻译程序能够根据所说内容的上下文不断更新其翻译,这将大大提高完成文档的准确性。 虽然该产品推出的日期还不确定--谷歌只是透露该功能将在“未来”发布--这个发展对人工智能翻译技术来说,是十分振奋的,但也引发了对专业语言服务未来的严重质疑。 如果用户无需支付一分钱就能获得即时语言解决方案,那么翻译市场上还有专家语言学家的一席之地吗? 高精度和即时结果具有广泛的吸引力 不难看出为什么谷歌翻译的最新更新如此吸引消费者。 该软件免费且易于使用,并为用户提供即时翻译。 关键的区别在于,这款应用将会产生更精确的结果,尽管英国的外语学习水平是2000年以来的最低水平。但需要用不同语言交流的人比以往任何时候都多,这使得这款应用特别有用。 在未来,超过60%的16-34岁的年轻人使用智能手机翻译当地语言,正如Raconteur所指出的,像谷歌翻译这样的人工智能应用程序已经“帮助大幅降低了翻译成本”。 2018年,谷歌首席执行官桑达尔•皮查伊报告称,该程序每天翻译1430亿字,但其新技术的细致力量意味着谷歌翻译的人气必将增长,并吸引更多用户。 更好的技术并不能消除错误 这些技术发展令人印象深刻,也意味着谷歌翻译的精确度在不断刷新历史。 它的实时转写功能会不断评估句子,不断纠正选词,添加标点符号,并将口音和地区方言考虑在内。 这意味着,与逐字翻译不同,谷歌将能够使用上下文来通知其输出。 尽管这可能是令人鼓舞的,但重要的是要记住,人工智能翻译仍然会犯错误,而且机器永远不会像人类一样理解人类的语言。 自然语言处理为数字处理提供了发言权 在数字时代,我们面临的最大挑战之一是在人类思维过程和自主技术的可能性之间建立桥梁,特别是我们如何向机器传授我们对上下文中单词的类似人类的理解。 自然语言处理(Natural language processing,NLP)是发展这种自主“思考”能力的一个步骤。 奇怪的错误在不太重要的情况下并不太成问题,比如一个游客试图翻译他们的晚餐订单。 然而,对于像会议记录和商务文件这样的重要材料,只有完美的翻译才是可以接受的。 无法保证谷歌会掌握日常语言的背景和细微差别,而且这个程序也可能无法记录声音的语调,可能会漏掉讽刺,同情或其他明显的人类状态和情感的暗示。 在医药和工程等复杂行业,其当谷歌的软件无法完全理解特定行业术语的含义时,这些看似微小的疏漏却尤其危险。 例如,就金融翻译而言,专业语言学家“全球之声”坚持认为,准确程度远不止是数字:“金融翻译人员应全面了解金融行业,了解该行业的新发展,并流利地使用目标语言。”谷歌翻译不精通先进的金融概念,也不了解其提供翻译服务的每个国家的行业法律和惯例。 在这种情况下,只有具备语言流利性,文化意识和行业专门知识的专业翻译人才能够做出无可挑剔的翻译。 语境和目的继续驱动翻译需求 谷歌翻译的最新版本显然是人工智能翻译软件领域的一个重大突破,有理由预测,这很可能是语言发展探索的第一个窗口。 然而,这并不意味着我们将走向一个专业翻译不再发挥作用的未来。 营销中的人工智能:当科技与传统融合 人工智能在市场营销中的应用比你想象的要普遍得多,如今,CMO和CTO正在更加紧密地合作,以使其取得成功。 就像许多其他的自动化服务一样,某些需求将由计算机程序充分满足,而其他的则需要人情味。 对于随意使用,谷歌翻译仍然是一个非常方便的工具,这个新的更新将使不同语言之间的实时交流变得比以往任何时候都更容易。 语言障碍的即时消除提供了一个宝贵和可利用的手段,使联系和促进文化理解。 然而,出于专业目的,认证的人工翻译服务仍然占据了上风。 在任何一个错误和误解可能造成破坏的项目中,无论内容多么复杂,专业语言学家的技巧,天赋和洞察力,可以产生忠实和完美无缺的译文。 “全球之声”是一家提供翻译和口译服务的公司,在全球范围内与超过3,000名具有专业资格的语言学家合作。

以上中文文本为机器翻译,存在不同程度偏差和错误,请理解并参考英文原文阅读。

阅读原文