AI的新场景应用:来“小猫钓鱼”呀

新闻

作者：微信文章


选词填空 - DeepMind 的人工智能

原文：

DeepMind's AI is getting closer to its first big real-world application.

In the six years since it was acquired by Google, DeepMind has been stringing together a long list of artificial intelligence milestones (11). It has outplayed Go champions, bested (12) professional StarCraft players and turned its attention to chess.

Aside from its work in healthcare-which in September 2019 became part of Google Health-what DeepMind hasn't been particularly vocal about is applying (13) its AI to more practical problems. There are some exceptions-DeepMind's AI has already helped to make Google's data centres more energy efficient (14) and improve the firm's text-to-speech systems-but most of its headline-grabbing work has focused on using games as proving grounds (15) for AI systems.

But now DeepMind is starting to tackle one of science's trickiest problems: protein folding (16). A paper published in the journal Nature details (17) how DeepMind's AI system was able to beat all of its opponents in a competition where algorithms predicted the structure of a protein based on its genetic makeup (18). Being able to predict the structure of proteins could make it much easier for us to develop new drugs, understand how genetic mutations lead to disease and develop genetically engineered (19) proteins.

“They did blow the field apart," says Paul Bates... "We were all a bit surprised that they did quite as well as that, since it was their first attempt (20) in the field.”

中文翻译：

DeepMind的人工智能正接近其首个重要的实际应用

自被谷歌收购的六年来，DeepMind 已经在人工智能领域取得了一长串的里程碑式成就。它击败了围棋冠军，战胜了《星际争霸》职业选手，并将注意力转向了国际象棋。

除了在医疗保健领域的工作（这部分于2019年9月并入谷歌健康部门），DeepMind 对其将人工智能应用于更实际的问题上并不特别高调。当然也有例外——DeepMind 的AI已经帮助谷歌的数据中心提高了能效，并改善了公司的文本转语音系统——但其大部分引人注目的工作都集中在将游戏作为AI系统的试验场。

但现在，DeepMind 开始着手解决科学界最棘手的问题之一：蛋白质折叠。发表在《自然》杂志上的一篇论文详细说明了 DeepMind 的AI系统如何在一场竞赛中击败所有对手，该竞赛要求算法根据蛋白质的基因组成来预测其结构。能够预测蛋白质结构将使我们更容易开发新药，理解基因突变如何导致疾病，并开发基因工程蛋白质。

“他们确实颠覆了这个领域，”保罗·贝茨说……“我们都对它们表现如此出色感到有些惊讶，因为这是他们在该领域的首次尝试。”

二、关键词汇与核心句型

1. 核心科技词汇

英文词汇 中文释义 学习提示

⭕️milestone 里程碑 比喻重大的成就或转折点。

⭕️best (v.) 击败，胜过 常以过去分词

"bested" 出现，是 "beat" 的更正式替代词。

⭕️apply A to B 将A应用于B 科技文中超高频短语。

⭕️energy-efficient 节能的 合成形容词，

"-efficient" 表示“效率高的”。

⭕️proving ground 试验场 比喻测试理论或技术的场所。

⭕️tackle a problem 解决问题

"tackle" 比 "solve" 更生动，强调“着手处理”难题。

⭕️detail (v.) 详细说明 名词变动词，意为“在论文/报告中详述”。

⭕️genetic makeup 基因组成 替换

"structure" 的高级表达。

⭕️genetically engineered 基因工程的 生物技术核心术语。

⭕️attempt 尝试，企图 既可作名词也可作动词。

2. 核心句型与语法点

1. 现在完成进行时：表示持续至今的动作

* 原句： DeepMind has been stringing together a long list of... achievements.

* 句型：

"has/have been doing" (现在完成进行时)

* 讲解： 强调从过去某时开始一直持续到现在的动作，并且可能仍在进行。此处说明DeepMind持续取得成就的过程。

* 仿写： Scientists have been working on this project for a decade. (科学家们研究这个项目已经十年了。)

2. 名词性从句（what引导的主语从句）

* 原句： What DeepMind hasn't been vocal about is applying its AI to practical problems.

* 句型：

"What ... is ..." (什么...是...)

* 讲解：

"What" 引导一个主语从句，在整个句子中作主语。这是使句子结构更复杂的高级写法。

* 仿写： What we need most now is time and patience. (我们现在最需要的是时间和耐心。)

3. 现在分词作状语

* 原句： ...a competition where algorithms predicted the structure of a protein based on its genetic makeup.

* 句型：

"V-ing / V-ed phrase" (分词短语) 作状语，表方式、原因等。

* 讲解： 此处

"based on..." 是过去分词短语，作方式状语，修饰

"predicted"，表示“根据...来预测”。

* 仿写： He answered the question using the new formula. (他运用新公式回答了问题。)

4. 比较状语从句（not... but... 结构）

* 原句： Processing takes place not in centralized servers, but on the "edge" of the network.

* 句型：

"not A but B" (不是A而是B)

* 讲解： 用于对比，否定前者，肯定后者。

* 仿写： Success depends not on talent, but on effort. (成功不靠天赋，而靠努力。)