[Annotated] Google leads in the race to dominate artificial intelligence

Tech giants are investing billions in a transformative technology^[각주:1]

Commanding the plot lines of^[각주:2] Hollywood films, covers of magazines and reams of^[각주:3] newsprint^[각주:4], the contest between^[각주:5] artificial intelligence (AI) and mankind^[각주:6] draws much attention^[각주:7]. Doomsayers^[각주:8] warn that AI could eradicate jobs^[각주:9], break laws and start wars. But such predictions^[각주:10] concern the distant future. The competition today is not between humans and machines but along the world's technology giants, which are investing feverishly^[각주:11] to get a lead over^[각주:12] each other in AI. 

An exponential increase in^[각주:13] the availability of^[각주:14] digital data, the force of computing power^[각주:15] and the brilliance of^[각주:16] algorithms has fueled^[각주:17] excitement about^[각주:18] this formerly^[각주:19] obscure corner of^[각주:20] computer science. The West's largest tech firms, including Alphabet (Google's parent), Amazon, Apple, Facebook, IBM and Microsoft are investing huge sums to^[각주:21] develop their AI capabilities, as are their counterparts in^[각주:22] China. Although it is difficult to separate tech firms' investments in AI from^[각주:23] other kinds, so far in 2017 companies globally have completed^[각주:24] around $21.3bn in mergers and acquisitions^[각주:25] related to AI, according to PitchBook, a data provider, or around 26 times more than in 2015. 

Machine learning is the branch of AI that is most relevant to these firms. Computers sift through data to^[각주:26] recognize patterns and make predictions without^[각주:27] being explicitly^[각주:28] programmed to do so. The technique is now used in all manner of^[각주:29] applications in the tech industry, including online ad targeting, product recommendations, augmented reality and self-driving cars. Zoubin Ghahramani, who leads AI research at Uber, believes that AI will be as transformative as the rise of^[각주:30] computers. 

One way to understand AI's potential impact is to look at databases. From the 1980s these made it cheap to store information, pull out insights^[각주:31] and handle cognitive tasks^[각주:32] such as inventory management^[각주:33]. Databases powered the first generation of^[각주:34] software; AI will make the next far more predictive^[각주:35] and responsive^[각주:36], says Frank Chen of Andreessen Horowitz, venture-capital firm^[각주:37]. An application such as Google's Gmail, which scans the content of^[각주:38] e-mails and suggests quick, one-touch replies on mobile devices, is an early example of what could be coming. 

As with past waves of new technology, such as the rise of personal computers and mobile telephony^[각주:39], AI has the potential to shake up the businesses of^[각주:40] the tech giants by helping them overhaul existing operations^[각주:41] and dream up^[각주:42] new enterprises^[각주:43]. But it also comes with a sense of menace^[각주:44]. "If you're a tech company and you're not building AI as a core competence^[각주:45], then you're setting yourself up for an invention from the outside^[각주:46]," says Jeff Wilke, chief executive of "worldwide consumer" at Amazon, and adjutant to^[각주:47] Jeff Bezos. 

Fueled by rivalry^[각주:48], high hopes^[각주:49] and hype^[각주:50], the AI boom can feel like the first California gold rush. Although Chinese firms such as Baidu and Alibaba are also investing in AI, and deploying it in^[각주:51] their home market^[각주:52], the most visible^[각주:53] prospectors are^[각주:54] Western tech firms. Alphabet is widely perceived to be in the lead^[각주:55]. It has been making sizeable profits from^[각주:56] AI for years^[각주:57] and has many of the best-known^[각주:58] researchers. But it is early days^[각주:59] and the race is far from over. Over the next several years, large tech firms are going to go head-to-head in three ways^[각주:60]. They will continue to compete for^[각주:61] talent to^[각주:62] help train their corporate "brains"^[각주:63]; they will try to apply machine learning to their existing businesses more effectively than rivals; and they will try to create new profit centers with the help of AI.

Idiot savants^[각주:64] 

The most frenzied^[각주:65] rush is for human talent, which is far more scarce than either data or computing power. Demand for AI "builders" who can apply machine-learning techniques to huge sets of data in creative ways has ballooned^[각주:66], far exceeding the number of top students who have studied the techniques. 

Today AI systems are like "idiot savants," says Gurdeep Singh Pall of Microsoft. "They are great at what they do, but if you don't use them correctly^[각주:67], it's a disaster." Hiring the right people can be critical to a firm's survival (some startups fail for lack of^[각주:68] the right AI skills) which has set off a trend of^[각주:69] firms plundering academic departments to^[각주:70] hire professors and graduate students^[각주:71] before they finish their degrees. 

Job fairs^[각주:72] now resemble^[각주:73] frantic^[각주:74] "Thanksgiving Black Friday sales at Walmart", says Andrew Moore, dean of^[각주:75] Carnegie Mellon University's (CMU) school of computer science, a pioneering institution in^[각주:76] AI (whose robotics department^[각주:77] was famously^[각주:78] plundered by Uber in 2015). Academic conferences^[각주:79], such as this week's Neural Information^[각주:80] Processing Systems in^[각주:81] Long Beach, California, double up as places to^[각주:82] shop for talent^[각주:83]. The best recruiters are academia's^[각주:84] AI celebrities: people like Yann LeCun of Facebook and Geoffrey Hinton of Google - both former professors who keep a university affiliation^[각주:85] - can attract others to work alongside^[각주:86] them. Proprietary data^[각주:87] can also serve as a draw^[각주:88], if the huge salaries are not enough. 

If none of that works, companies buy whole startups. The tech industry first took notice of this trend in^[각주:89] 2014, when Google spent an estimated $500m on^[각주:90] DeepMind, a startup with no revenue or marketable product^[각주:91] but a team of "deep learning" researchers; after the deal they designed a program that beat the world champion at "Go", an ancient board game. Other firms have also shelled out to^[각주:92] buy money-losing startups^[각주:93], which are typically valued not on future profits or even sales but instead receive a price for each employee that can be as much as $5m-10m.

Behind closed doors^[각주:94]

Companies have different philosophies about how to deal with staff. Some, such as Microsoft and IBM, invest heavily in AI research and publish a large number of papers, but do not require researchers to apply their findings^[각주:95] to money-making activities^[각주:96]. At the opposite end of^[각주:97] the scale^[각주:98] are Apple and Amazon, which do not have enormous research initiatives^[각주:99], expect all work to feed into products^[각주:100] and are tight-lipped about^[각주:101] their work. Google and Facebook are somewhere in between on whether researchers must toil only on^[각주:102] money-making ventures. 

The intense battle for talent may force secretive companies to^[각주:103] become more open. "If you tell them, 'come work with us but you can't tell anyone what you're working on', then they won't come because you'll be killing their career," explains Mr LeCun, who leads Facebook's AI research lab^[각주:104]. This trade-off between^[각주:105] secrecy^[각주:106] and the need to attract people also applies to the Chinese giants, which are trying to establish Western outposts^[각주:107] and hire American researchers. Baidu has opened two research labs with an AI focus in Silicon Valley, in 2013 and this year. Western AI researchers rate them highly^[각주:108] but prefer to work for the American giants, in part^[각주:109] due to their relative transparency. 

If companies can lure^[각주:110] the right people in AI, the effect is to extend their workforces exponentially^[각주:111]. AI is "like having a million interns" at one's disposal^[각주:112], says Benedict Evans of Andreessen Horowitz. That computational power is then integrated into^[각주:113] firms' existing businesses.

The advantages of AI are most visible in^[각주:114] firms' predictions of what users want. Automated recommendations and suggestions are responsible for around three-quarters of what people watch on Netflix, for example, and more than a third of what people buy on Amazon. Facebook, which owns the popular app Instagram, uses machine learning to recognize the content of posts, photos and videos and display relevant ones to users, as well as filter out spam^[각주:115]. In the past it ranked posts^[각주:116] chronologically^[각주:117], but serving up posts and ads^[각주:118] by relevance^[각주:119] keeps users more engaged. 

Without machine learning, Facebook would never have achieved its current scale, argues Joaquin Candela, head of its applied AI^[각주:120] group. Companies that did not use AI in search, or were late to do so, struggled, as in the case of Yahoo and its search engine, and also Microsoft's Bing. 

Amazon and Google have gone furthest in applying AI to a range of operations^[각주:121]. Machine learning makes Amazon's online and physical operations more efficient. It has around 80,000 robots in its fulfillment centers^[각주:122], and also uses AI to categorize inventory and decide which trucks to allocate packages to. For grocery ordering, it has applied computer vision to recognize which strawberries and other fruits are ripe and fresh enough to be delivered to customers, and is developing autonomous drones that will one day deliver orders. 

As for Google, it uses AI to categorize content on YouTube, its online-video website, and weed out^[각주:123] (some) material, and also to identify people and group them in its app, Google Photos. AI is also embedded in^[각주:124] Android, its operating system^[각주:125], helping it to work more smoothly and to predict which apps people are interested in using. Google Brain is regarded in the field of AI as one of the best research groups at applying machine-learning advances profitably, for example by improving search algorithms. As for DeepMind, the British firm may not ever generate much actual revenue for Alphabet, but it has helped its parent save money by increasing the energy efficiency of^[각주:126] its global data centers (and its Go experiment was a public-relations^[각주:127] coup^[각주:128]). 

Artificial intelligence is also being applied in the corporate world. David Kenny, the boss of Watson, IBM's AI platform, predicts that there will be "two AIs": companies that profit from offering AI-infused services to^[각주:129] consumers and others which offer them to businesses. In practice, the two worlds meet because of the tech giants' cloud-computing arms^[각주:130]. Providers are competing to use AI as a way to differentiate^[각주:131] their offerings^[각주:132] and lock in customers. The three largest - Amazon Web Services, Microsoft's Azure and Google Cloud - offer application-programming interfaces (APIs)^[각주:133] that provide machine-learning capabilities to other companies. Microsoft's cloud offering, Azure, for example, helped Uber build a verification tool that asks drivers to take a selfie to confirm their identities when they work. Google Cloud offers a "jobs API", which helps companies match jobseekers with the best positions^[각주:134]. 

AI on the brain^[각주:135]

Many firms in other industries, from retailing^[각주:136] to media, stand to benefit from^[각주:137] what those in the cloud business tout as^[각주:138] the "democratization" of AI. Providing AI to companies that do not have the skills or scale to build up sophisticated capabilities^[각주:139] independently could be a money-spinner in^[각주:140] the $250bn cloud market. But providers often must customize APIs for clients' complex needs, which is time-consuming^[각주:141]. Microsoft, with its history of selling software to clients and offering them support, seems likely to do well in this area. It is only a matter of time before AI offerings become "more and more self-help^[각주:142]", counters Diane Greene, who runs Google Cloud. 

IBM is another contender^[각주:143], having backed a huge marketing campaign for^[각주:144] its Watson platform. AI researchers tend to be dismissive of^[각주:145] IBM, which has a large consulting business and a reputation for valuing time billed terabytes. The firm's critics also point out that, although IBM has invested over $15bn in Watson and spent $5bn between 2010 and 2015 to buy companies, much of that with the aim of acquiring proprietary data, for the most part it does not have unique data of its own. But IBM's weaknesses may not hold it back^[각주:146]. Bosses of most businesses feel pressure to have an AI strategy, and they will pay handsomely^[각주:147] to acquire one quickly. 

To date tech giants have mostly tried to apply AI to reap profits from^[각주:148] their existing operations. In the next few years they hope that AI will let them build new businesses. One area of intense competition is virtual assistants. Smartphones know their users intimately, but AI-powered virtual assistants aim to take the relationship further, whether through phones or smartspeakers. Apple was first to explore their promise when it bought Siri, a voice assistant, in 2010. Since then Amazon, Google and Microsoft have invested heavily: their assistants' speech recognition is better as a result. Samsung, Facebook and Baidu are also competing to offer them. 

One algorithm to rule them all

It is unclear whether standalone^[각주:149] speakers will become a huge market, but it is certain that people will move beyond text to engage with the internet. "All these companies understand that whoever owns that choke point for^[각주:150] consumers will rule the market," says Pedro Domingos, author of "The Master Algorithm", a book about AI.

Further into the future, augmented-reality (AR) devices are another AI-infused opportunity. Mobile apps like Snap, a messaging app, and the game Pokemon Go are early examples of AR. But AR could more radically transform^[각주:151] people's relationship with the internet, so that they consume digital information not from a small screen but via an ambient^[각주:152], ever-present experience^[각주:153]. AR devices will offer portable AI capabilities, such as simultaneous translation and facial recognition. 

In the race for AR, big tech firms have not got much beyond the warm-up^[각주:154] phase^[각주:155]. Google and Apple have launched AR software-development kits; they both want developers to build apps that use AR on their platforms. There is also a rush to develop AR hardware. Google was early to launch a prototype for^[각주:156] AR glasses, but they flopped^[각주:157]. Microsoft has developed a headset it calls HoloLens, but with a price of between $3,000-5,000, it is a niche product^[각주:158]. Other firms, including Facebook and Apple, are thought to be planning their own offerings. Being ahead in AI could translate into^[각주:159] big leads in^[각주:160] these new fields. 

Nowhere is that truer than in the realm of^[각주:161] autonomous vehicles^[각주:162]. Tech firms are driving millions of miles to build up big, proprietary datasets^[각주:163], and are making use of^[각주:164] computer vision to train their systems to recognize objects in the real world. The potential spoils are huge^[각주:165]. Personal transportation is a vast market, worth around $10trn globally, and whoever cracks self-driving cars^[각주:166] can apply their knowledge to other AI-based projects, such as drones and robots. Unlike search engines, where people may choose to use a service that is good enough, users are more likely to favor self-driving cars with the best safety record, meaning that the companies that best employ AI to map out the physical world^[각주:167] and register the fewest crashes will enjoy outsize benefits^[각주:168]. 

Each firm is approaching the problem differently. Baidu, the Chinese giant, is trying to create a self-driving-car operating system, much like Google's Android in mobile devices (although it is unclear how it plans to make money). Alphabet has its own autonomous-car effort, as do Uber, Tesla, an electric carmaker, a herd of^[각주:169] little-known startups^[각주:170] and, increasingly, established carmakers. (Apple is rumored to have^[각주:171] scaled back^[각주:172] its car ambitions.)

Self-driving cars are just one example of how technology firms' AI strategies are pushing beyond the virtual world of software into hardware. Many companies, including Alphabet, Apple and Microsoft, are also investing to build specialized^[각주:173], powerful "AI chips" that can power their various activities. These will compete with^[각주:174] those made by NVIDIA, a tech firm that has built an empire on powerful chips used in various AI realms, such as autonomous cars and virtual reality. 

It is unclear whether the likes of^[각주:175] Alphabet and Apple will sell these chips to rival firms or keep them for themselves. They have an incentive to^[각주:176] use their innovations to improve their own services, rather than renting or selling them to^[각주:177] rivals - which could become a problem if it means a very few firms develop a meaningful advantage in brute computing power^[각주:178]. 

That begs the broader question of whether^[각주:179] AI will further concentrate power among today's digital giants. It seems likely that the incumbent tech groups will capture many of AI's gains^[각주:180], given their wealth of data, computing power, smart algorithms and human talent, not to mention a head start on^[각주:181] investing. History points to the likelihood of concentration^[각주:182]; both databases and personal computers ushered in ascendancies^[각주:183], if only for a while, of a tiny group of tech firms (Oracle and IBM in databases, Microsoft and Apple in personal computers). 

By the metrics that count^[각주:184] - talent, computing power and data - Google appears to be in the lead in^[각주:185] AI. It can afford the cleverest people^[각주:186] and has such a variety of projects, from drones to cars to smart software, that people interested in machine learning rarely leave. Other firms had to learn to take AI seriously^[각주:187], but Google's founders were early devotees of^[각주:188] machine learning and always saw it as a competitive edge^[각주:189]. 

AI's spiritual home^[각주:190]

Some in the tech industry, such as Elon Musk, the boss of Tesla and rocket firm SpaceX, worry about Alphabet and other firms monopolizing AI talent and expertise^[각주:191]. He and a handful of other prominent^[각주:192] Silicon Valley bosses funded OpenAI, a not-for-profit research outfit focused on AI with no corporate affiliation^[각주:193]. Mr Musk and others are worried about what might happen when a firm finally cracks "general intelligence", the ability of a computer to perform any human task without being explicitly^[각주:194] programmed to do so. Such a vision is probably decades away, but that does not stop Google from talking about it. "We absolutely want to" crack general AI, says Jeff Dean, the boss of Google Brain. If a firm were to manage this, it could change the competitive landscape entirely^[각주:195]. 

In the meantime^[각주:196], much will depend on whether tech firms are open and collaborative^[각주:197]. In addition to publishing papers^[각주:198], many companies today make their machine-learning software libraries open source, offering internal tools to rivals and independent developers. Google's library, TensorFlow, is particularly popular. Facebook has open-sourced two of its libraries, Caffe2 and Pytorch. Openness has strategic advantages. As they are used, the libraries are debugged^[각주:199], and the firms behind them get reputational benefits^[각주:200]. "Beware of^[각주:201] geeks bearing gifts^[각주:202]," quips^[각주:203] Oren Etzioni of the Allen Institute for Artificial Intelligence, another non-profit research group. 

One guru of^[각주:204] the field worries that libraries such as TensorFlow will bring in^[각주:205] talented researchers but that their owners may start charging later on^[각주:206], or use them for profit in other ways. Such caution may prove wise, but few think about the long term when a gold rush is under way^[각주:207]. So it is now in Silicon Valley. Most techies^[각주:208] are too consumed by the promise and potential profits of AI to spend too much time worrying about the future. 

1. transformative ; [형용사] 변형시키는, 변하게 하는, 변형시키는 힘이 있는. [본문으로]
2. command ; 4. VIEW | [타동사][VN] [수동태로는 안 씀.진행형으로는 쓰이지 않음] (격식) (무엇을 보거나 통제할 수 있는) 위치에 있다 ;; 5. CONTROL | [타동사][VN] [수동태로는 안 씀.진행형으로는 쓰이지 않음] (격식) …을 장악하다; …을 이용할 수 있다 [본문으로]
3. ream ; 1. [pl.] reams (비격식) (글의 양이) 많음 ;; 미국∙영국 [ri:m] [본문으로]
4. newsprint ; [U] 신문 인쇄용지 [본문으로]
5. contest ; 2. ~ (for sth) (주도권이나 권력) 다툼[경쟁] [본문으로]
6. mankind ; [U] 인류; (모든) 인간, 사람들 ;; 참고 ; humankind, womankind [본문으로]
7. draw attention ; 관심을 끌다 [본문으로]
8. doomsayer ; (특히 美) (英 또한 doom・ster / |duːmstə(r) /) 횡액[재앙]을 예언하는 사람 [본문으로]
9. eradicate ; [타동사] 뿌리째 뽑다(root up); 박멸하다, 근절하다(root out) ;; 미국∙영국 [ɪ|rӕdɪkeɪt] [본문으로]
10. prediction ; [C , U] 예측, 예견 [본문으로]
11. feverishly ; [부사] 열병에 걸린 것같이 ; 열광하여. [본문으로]
12. a lead over ; …에 대한 우위. [본문으로]
13. exponential ; 2. (격식) (증가율이) 기하급수적인 ;; 미국∙영국 [|ekspə|nenʃl] [본문으로]
14. availability ; 1. [U] 유효성, 유용성, 효용; (입수) 가능성 [본문으로]
15. computing power ; 연산 능력 [본문으로]
16. brilliance ; [U] 광휘, 광채; 슬기, 탁월, 걸출; 〈광학〉 휘도(輝度), 밝기; (음색의) 맑음. [본문으로]
17. fuel ; 2. …에 활기를 불어넣다, 부채질하다; 지지하다, 자극하다. [본문으로]
18. excitement ; 1. [U] 흥분, 신남 [본문으로]
19. formerly ; [부사] 이전에, 예전에 [본문으로]
20. obscure ; 1. 잘 알려져 있지 않은, 무명의 ;; 2. 이해하기 힘든, 모호한 [본문으로]
21. sum ; (pl. sums[-z]) 1. 합계, 총계, 총량, 총수, 총액(totality). [본문으로]
22. counterpart ; 3. (…에) 상응하는[해당되는] 사람[것], 동격의 사람[것]; 동등한 것, 등가물. [본문으로]
23. separate ... from ; ...를 ~에서 분리하다[떼어 놓다]. [본문으로]
24. complete ; 4. <계약을> 이행하다 [본문으로]
25. mergers and acquisitions ; [명사] (경영) 합병 매수, 인수 합병(약어 M&A). [본문으로]
26. sift through ; 꼼꼼하게 살펴 추려내다 ;; to carefully examine a large amount of something in order to find something important or decide what is useful [본문으로]
27. make a prediction ; 예측하다. [본문으로]
28. explicitly ; [부사] 명쾌하게, 명백하게 [본문으로]
29. in the manner of sb/sth ; ~의 방식[스타일]으로 [본문으로]
30. rise ; 2. 입신, 출세; 향상, 진보; 번영 ((of, to)) [본문으로]
31. pull out insights ; 흐름상 "전망하다, 통찰력을 얻다" 정도의 의미로 이해 [본문으로]
32. cognitive ; 인식의, 인식력 있는; (지각·기억·판단·추리 등의) 지적·정신적 작용의[에 관한] [본문으로]
33. inventory management ; 재고 관리 : 보급기관에서 장차의 수요에 신속히 그리고 경제적으로 적응할 수 있도록 재고를 최적정 수준으로 유지하고 과학적으로 관리하는 것. [본문으로]
34. power ; 3. …의 (정신적인) 힘이 되다, 고무시키다, 분기시키다 [본문으로]
35. predictive ; 1. 예언[예보]하는, 예언적인. ;; 2. (…의) 전조가 되는[of]. [본문으로]
36. responsive ; [~ (to sb/sth)] 1. [대개 명사 앞에는 안 씀] 즉각 반응[대응]하는 ;; 2. 관심[열의]을 보이는, 호응하는 [본문으로]
37. venture capital ; Venture capital (VC) is a type of private equity,[1] a form of financing that is provided by firms or funds to small, early-stage, emerging firms that are deemed to have high growth potential, or which have demonstrated high growth (in terms of number of employees, annual revenue, or both). Venture capital firms or funds invest in these early-stage companies in exchange for equity, or an ownership stake, in the companies they invest in. Venture capitalists take on the risk of financing risky start-ups in the hopes that some of the firms they support will become successful. The start-ups are usually based on an innovative technology or business model and they are usually from the high technology industries, such as information technology (IT), clean technology or biotechnology. [본문으로]
38. scan ; 5. (컴퓨터) <데이터를> 주사하다, 훑다 [본문으로]
39. mobile telephony ; 이동전화서비스업 [본문으로]
40. shake up ; [동사] 흔들어 섞다; 개편하다, 개혁하다 [본문으로]
41. overhaul ;[vn] 1. (기계・시스템을) 점검[정비]하다 [본문으로]
42. dream up ; [VERB] to invent by ingenuity and imagination [본문으로]
43. enterprise ; 1. [C] 기업, 회사 ;; 2. [C] (특히 모험성) 대규모 사업 [본문으로]
44. a sense of menace ; 위기감, 위협적인 느낌 [본문으로]
45. core competence ; 핵심 역량 [본문으로]
46. set oneself up for ; ~에 이르게 되다 [본문으로]
47. adjutant ; 2. 조수, 보좌. ;; 참고 ; helper, assistant ;; 미국∙영국 [|ӕdƷʊtənt] [본문으로]
48. rivalry ; [U, C] (pl. -ries) ~ (with sb/sth) (for sth) | ~ (between A and B) (for sth) 경쟁 (의식) ;; 경쟁, 대항; 경쟁[적대] 행위, 대항 (의식), 대립 (관계). ;; 미국∙영국 [|raɪvlri] [본문으로]
49. high hopes ; 큰 기대. [본문으로]
50. hype ; 《美구어》 1. 과대[과장] 선전, 떠들썩한 판촉 활동; 과장 보도; 선동. ;; 2. 속임(수), 사기; 거짓, 허위. [본문으로]
51. deploy ; 2. (격식) 효율적으로 사용하다 [본문으로]
52. home market ; 국내시장 [본문으로]
53. visible ; 2. 명백한, 보아 알 수 있는(perceptible) ;; 3. <사람·사건 등이> 빈번하게 뉴스에 나오는, 활동이 눈에 띄는 [본문으로]
54. prospector ; (금・광물 등을 찾는) 탐사[탐광]자 [본문으로]
55. be in the lead ; 앞장서[선두에] 있다; 리드하고 있다 [본문으로]
56. sizeable ; [형용사] 꽤 큰[많은], 상당한 [본문으로]
57. for years ; 수년간, 몇 해 동안 [본문으로]
58. best-known ; [형용사] [WELL-KNOWN의 최상급] 가장 잘 알려진 [본문으로]
59. it is early days (yet) ; 아직 때가 이르다(그러니 좀 더 두고 봐야 안다는 뜻) [본문으로]
60. go head-to-head (with sb) ; ~와 정면으로 맞서다 [본문으로]
61. compete (with sb) for ; …을 위해 (…와) 싸우다. [본문으로]
62. talent ; 1. [C , U] ~ (for sth) 재주, (타고난) 재능, 장기 ;; 2. [U , C] 재능[재주] 있는 사람[사람들] [본문으로]
63. train ; [타동사] 3. (재능 등을) 연마[훈련]하다 [본문으로]
64. savant ; (격식) 1. 학자, 석학 ;; 2. 서번트(전반적으로는 정상인보다 지적 능력이 떨어지나 특정 분야에 대해서만은 비범한 능력을 보이는 사람) ;; 미국식 [sӕ|vɑ:nt] 영국식 [|sӕvənt] [본문으로]
65. frenzy ; [수동형으로] …을 몹시 흥분[격분]시키다, 미쳐 날뛰게 하다[with]. (또는 phrensy) -zi·ly 부사 [본문으로]
66. balloon ; [자동사] 3. 급증[급상승]하다 [본문으로]
67. correctly ; [부사] 바르게, 정확하게; [문장을 수식하여] 정확히 말하면 [본문으로]
68. fail for ; ~의 이유로 실패하다 [본문으로]
69. set off ; 흐름상 "~하기 시작하다" 정도의 의미 [본문으로]
70. plunder ; 1. (군대 등이) …을 약탈하다; 〔사람·장소〕에게서 (…을) 강탈하다, 훔치다[of]. [본문으로]
71. graduate student ; [명사] 대학원생. [본문으로]
72. job fair ; [명사] (사원을 모집하는 회사가 개최하는) 공개 취직 설명회. [본문으로]
73. resemble ; [타동사][VN] [수동태로는 안 씀,진행형으로는 쓰이지 않음] 닮다, 비슷[유사]하다 [본문으로]
74. frantic ; 1. 정신없이[미친 듯이] 서두는[하는] ;; 2. (두려움・걱정으로) 제정신이 아닌 [본문으로]
75. dean ; 2. (대학의) 학장; (미국의 대학의) 학생처장; (Oxford, Cambridge 대학의) 학생감 [본문으로]
76. pioneering ; [형용사] (주로 명사 앞에 씀) 개척[선구]적인 [본문으로]
77. robotics ; [명사] 로봇 공학 [본문으로]
78. famously ; [부사] 1. 유명하게, 이름 높게 ;; 2. 뛰어나게, 훌륭하게 ;; (구어) 굉장히 잘. [본문으로]
79. academic conference ; 학술 대회 [본문으로]
80. neural ; (전문 용어) 신경(계통)의 ;; 미국식 [|nʊrəl] 영국식 [|njʊərəl] [본문으로]
81. information processing ; (컴퓨터) (컴퓨터 등에 의한) 정보 처리(cf. DATA PROCESSING) [본문으로]
82. double (up) as ; to be also used as something else [본문으로]
83. shop for ; [싼 물건·투자 대상 따위]를 물색하다, 찾아다니다 [본문으로]
84. academia ; [U] (또한 격식 또는 유머 aca・deme / |ækədiːm /) 학계 [본문으로]
85. affiliation ; 6. 제휴; 《美》 협력[친선] 관계; 제휴[협력] 기관[with, to]. [본문으로]
86. work alongside ; (+사람) ~와 함께 일하다. (+일,작업) ~에 대해 함께 일하다. ;; 참고 ; work with [본문으로]
87. proprietary ; 3. 독점의; 독점적인 ;; 4. 사유의, 전유의 ;; 미국식 [prə|praɪəteri] 영국식 [prə|praɪətri] [본문으로]
88. draw ; [명사] 2. 이목[인기]을 끄는 것; (사람을) 끄는 것 [본문으로]
89. take notice of ; …을 알아차리다, 주의하다; …을 후대하다; (신문 따위가) …을 들어 논평하다 [본문으로]
90. estimated ; [형용사] 견적한, 어림잡은, 추측한 [본문으로]
91. marketable ; 시장 판매에 알맞은, 잘 팔리는; 시장에서 현재 매매되고 있는 ;; 시장성이 있는 [본문으로]
92. shell out ; (비격식) (~에 거금을) 들이다[쏟아 붓다] ;; 참고 ; spend ;; 동의어 ; shell something out (for something) [본문으로]
93. money-losing ; [형용사] 적자를 내는 [본문으로]
94. behind closed doors ; 비공개로[비밀리에] [본문으로]
95. findings ; [명사] 조사[연구] 결과들 [본문으로]
96. money-making ; (법률) 영리 [본문으로]
97. the opposite end ; 반대쪽, 정반대 쪽 끝. [본문으로]
98. scale ; 3. RANGE OF LEVELS | [C] [주로 단수로] 계층 구조 ;; 4. MARKS FOR MEASURING | [C] (측정기의) 눈금 ;; 5. WEIGHING INSTRUMENT | [pl.] (美 또한 scale) scales 저울 [본문으로]
99. initiative ; 1. [C] (특정한 문제 해결・목적 달성을 위한 새로운) 계획 [본문으로]
100. feed into ; ~에 반영되다, …에 넣다, 투자하다, 입력하다 [본문으로]
101. tight-lipped ; 입을 굳게 다문 ; 좀처럼 입을 열지 않는, 입이 뜬, 말이 적은. [본문으로]
102. toil ; [자동사] (toils[-z]) 1. 애쓰다, 꾸준히[열심히] 일하다(away)[at, on, over, through]. [본문으로]
103. secretive ; 숨기는 경향이 있는 <사람·성질 등>, 비밀주의의; 말 없는 [본문으로]
104. research lab ; a workplace for the conduct of scientific research [본문으로]
105. trade-off ; [명사] ~ (between sth and sth) (서로 대립되는 요소 사이의) 균형 [본문으로]
106. secrecy ; [U] 비밀 유지[엄수]; 비밀(인 상태) [본문으로]
107. outpost ; 1. [군사] 전초; 전초 부대[기지]; 주둔 기지 ;; 2. 변경(邊境)의 식민지[거류지] [본문으로]
108. rate ; [진행형으로는 쓰이지 않음] 1. ~ sb/sth (as) sth | ~ as sth (특정한 수준으로) 평가하다[여기다]; 평가되다[여겨지다] [본문으로]
109. in part ; 부분적으로는; 어느 정도는 [본문으로]
110. lure ; [타동사][VN + adv. / prep.] (못마땅함) 꾀다, 유혹하다 [본문으로]
111. exponentially ; [부사] 전형적으로, 기하급수적으로. [본문으로]
112. at one's disposal ; [부사] …의 마음대로 이용[사용]할 수 있게. ;; 동의어 ; available for one's use. [본문으로]
113. be integrated into ; ~ 에 융화되다, 통합되다 [본문으로]
114. visible ; 3. <사람·사건 등이> 빈번하게 뉴스에 나오는, 활동이 눈에 띄는 [본문으로]
115. filter out ; 2. (원치 않는 사람·물건을) 걸러 내다 [본문으로]
116. rank ; 2. PUT IN LINE/ROW | [타동사][VN] [주로 수동태로] 정렬시키다, 늘어서게 하다 [본문으로]
117. chronologically ; [부사] 연대순으로 [본문으로]
118. serve (sth) up ; 2. ~을 내놓다[제공하다] [본문으로]
119. relevance ; 1. (표현 등의) 적절, 타당성; (당면 문제와의) 관련(성) ((to)) [본문으로]
120. applied ; 3. 실용적 기능을 가진 기술의[에 관한]. [본문으로]
121. a range of ; 다양한 [본문으로]
122. fulfillment ; [U] 1. (의무·직무 등의) 이행, 수행, 완수; 실천; 실현, 달성; (예언의) 성취 ;; 2. 고객의 주문 처리[과정] [본문으로]
123. weed out ; 바람직하지 않은 것을 제거하다; (원하지 않는 사람 또는 물건을) 그룹이나 수집에서 제외하다, (불필요하거나 부족한 대상 등을) 제거하다[뽑아 버리다], 잡초를 뽑다, 제거하다 [본문으로]
124. embed ; (-dd-), [주로 수동태로] 1. ~ sth (in sth) (단단히) 박다[끼워 넣다] [본문으로]
125. operating system ; [명사] (컴퓨터) 운영 체제 [본문으로]
126. energy efficiency ; [명사] 연료[에너지] 효율. [본문으로]
127. public-relations ; 1. [U] (약어:PR) 홍보[선전] (활동) [본문으로]
128. coup ; 2. (힘든 일의) 성공, 대단한 성취 [본문으로]
129. infuse ; 2. [타동사][VN] (격식) (속속들이) 스미다[영향을 미치다] [본문으로]
130. arm ; 6. OF ORGANIZATION | [주로 단수로] ~ (of sth) (조직의) 부문 [본문으로]
131. differentiate ; 2. (특징 따위에) 차이가 나다, (다른 것과) 구별되게 하다. [본문으로]
132. offering ; (참고: burnt offering , peace offering) 1. (사람들이 사용하거나 즐기도록) 제공된[내놓은] 것 [본문으로]
133. application programming ; (컴퓨터/통신) 응용 프로그래밍 [본문으로]
134. match ... with ; …와 짝을 맞추다, 어울리는 것을 찾다 (=find the equivalent of). [본문으로]
135. on the brain ; 마음속에 [본문으로]
136. retailing ; [U] 소매업 ;; 참고 ; wholesaling [본문으로]
137. stand to ; 2. [약속·주의·입장 따위]를 고수하다 ;; 3. [의견 따위]를 고집하다, …을 주장하다 [본문으로]
138. tout as ; …라고 칭찬하다, 장점을 내새우다 [본문으로]
139. build sth up ; 1. ~을 창조[개발]하다 ;; 2. ~을 더 높이다[더 강력하게 만들다] [본문으로]
140. money-spinner ; [명사] (英 비격식) 돈을 아주 많이 벌어들이는 것 [본문으로]
141. time-consuming ; [형용사] (많은) 시간이 걸리는 [본문으로]
142. self-help ; [U] 자립, 자조(自助) [본문으로]
143. contender ; [명사] (어떤 것을 두고 겨루는) 도전자[경쟁자] [본문으로]
144. marketing campaign ; [명사] 마케팅 캠페인, 마케팅 전략 [본문으로]
145. dismissive ; ~ (of sb/sth) 무시[멸시]하는 [본문으로]
146. hold sb/sth back ; 2. (진전·발전을) 방해[저해]하다 [본문으로]
147. handsomely ; [부사] 훌륭하게, 멋지게; 당당하게; 후하게, 관대하게 [본문으로]
148. reap ; 1. [타동사][VN] (특히 좋은 결과 등을) 거두다[수확하다] [본문으로]
149. standalone ; self-contained; especially : operating or capable of operating independently of a computer system [본문으로]
150. choke point ; (교통의) 애로(bottleneck), 요충, 관문(關門) [본문으로]
151. transform ; [vn], [~ sth/sb (from sth) (into sth)] 1. 변형시키다 ;; 2. (모습・성격을, 특히 더 좋게) 완전히 바꿔 놓다[탈바꿈시키다] [본문으로]
152. ambient ; 1. [명사 앞에만 씀] (전문 용어) 주위[주변]의 [본문으로]
153. ever-present ; [형용사] 항상 존재하는 [본문으로]
154. warm-up ; 2. 일의 시초, 시작, 사전 연습 [본문으로]
155. phase ; 1. (변화・발달 과정상의 한) 단계[시기/국면] [본문으로]
156. prototype ; 4. 시작품(試作品), 시제품. [본문으로]
157. flop ; 3. (비격식) 완전히 실패하다 [본문으로]
158. niche ; 1. 아주 편한[꼭 맞는] 자리[역할/일 등] ;; 2. (상업) (시장의) 틈새 ;; 미국식 [nɪtʃ;ni:ʃ] 영국식 [ni:ʃ;nɪtʃ] [본문으로]
159. translate into ; ~로 바뀌다, ~로 이어지다, 변환되다 ;; "번역하다" 라는 의미 외에도 사용가능함을 확인 [본문으로]
160. lead ; 1. FIRST PLACE | [sing.] the lead (경주・경쟁에서) 선두[우세] [본문으로]
161. in the realm of ; ~의 영역 내에서, 부문, 분야에서 [본문으로]
162. autonomous vehicle ; 자율주행차 [본문으로]
163. proprietary ; 2. 독점의, 전매의. ;; 3. 사유의, 전유(專有)의; 사립[사설]의, 개인 경영의. [본문으로]
164. make use of ; [동사] …을 이용하다; …로 덕보다. ;; 동의어 employ; benefit from. [본문으로]
165. spoil ; 3. (약탈 따위의) 목적물 ;; 4. [pl.] (노력 따위의) 성과; (수집가의) 수집물, 횡재물 [본문으로]
166. crack ; 4. [구어] 〔어려운 것 따위〕를 풀다, 해결하다, 해독하다. [본문으로]
167. map out ; [동사] (계획을) 입안하다, 계획하다, 배치하다. ;; 동의어 plan, organize, draft. [본문으로]
168. outsize ; [주로 명사 앞에 씀], (또한 out・sized / 'aʊtsaɪzd /) 1. (보통 것보다) 대형의 [본문으로]
169. herd ; 2. (보통 못마땅함) (같은 종류의 한 무리의) 사람들[대중] [본문으로]
170. little-known ; [형용사] 거의 알려지지 않은 [본문으로]
171. be rumored to... ; ~가 ... 한다는 소문이 있다 [본문으로]
172. scale back ; [VERB] to reduce or make a reduction in the level of activity, extent, numbers, etc [본문으로]
173. specialized ; [형용사] 전문적인, 전문화된 [본문으로]
174. compete with ; ~와 겨루다 [본문으로]
175. the like(s) of ; (구어) …과 같은 사람[것] [본문으로]
176. incentive ; [U, C] 1. 자극, 유인(誘因), 동기[to, for]. ;; 동의어 ; STIMULUS [본문으로]
177. rent ; 2. ~ sth (out) (to sb) (집세・사용료 등을 받고) 세 놓다[임대하다] [본문으로]
178. brute ; 2. 짐승의[같은]; 야만적인(savage), 난폭한, 잔인한 ;; 3. 적나라한, 가감하지 않은 [본문으로]
179. beg the question ; 1. 질문을 하게 만들다 ;; 2. (사실이 아닐지도 모르는 것을) 단정짓다 [본문으로]
180. capture ; 3. 〔상품 따위〕를 획득하다 [본문으로]
181. head start ; (경기·사업 따위에서의) 유리한 스타트, 시발점에서의 우위, 순조로운 출발[over, on]. [본문으로]
182. likelihood ; [U] 1. 있음직함, 가망, 가능성. ;; 2. [英] 유망함, 장래성. [본문으로]
183. ascendancy ; [U] ~ (over sb/sth) (격식) 지배력[영향력]을 행사할 수 있는 위치[지위] ;; 주로 단수로 쓰거나 불가산으로 활용하는듯하지만 본문에서는 복수로 활용한 것을 확인 [본문으로]
184. by the metrics that count ; 흐름상 "측정, 평가 될 수 있는 척도를 보자면" 정도의 의미 [본문으로]
185. in 을 반복해서 쓰는게 어색하게 느껴질 수 있지만 사용해도 무방한 것을 확인 [본문으로]
186. afford ; 1. [수동태로는 안 씀.특히 부정문이나 의문문에서 보통 can, could, be able to와 함께 쓰여] (…을 살・할・금전적・시간적) 여유[형편]가 되다 ;; 3. (격식) 제공하다 [본문으로]
187. take ; 22. ACCEPT/RECEIVE | [타동사][VN + adv. / prep.] (어떤 반응을 보이며) 받아들이다 [본문으로]
188. devotee ; [~ (of sb/sth)] 1. 헌신적인 추종자, 열성적인 애호가 ;; 2. 열성 신자 [본문으로]
189. competitive edge ; [명사] 경쟁 우위. [본문으로]
190. spiritual home ; [명사] (one’s ~) 정신적인[마음의, 영혼의] 고향. [본문으로]
191. expertise ; [U] ~ (in sth/in doing sth) 전문 지식[기술] ;; 미국식 [|ekspɜ:r|ti:z] 영국식 [|ekspɜ:|ti:z] ;; 참고 ; expertize ; 미국·영국 [ékspərtàiz] [본문으로]
192. prominent ; 1. 중요한; 유명한 [본문으로]
193. affiliation ; [U] 1. 입회, 가입; 합병, 합동, 제휴 [본문으로]
194. explicitly ; [부사] 명쾌하게 [본문으로]
195. landscape ; 3. (전체적으로 본) 지형, 지표(地表). ;; 4. 분야, …계. [본문으로]
196. in the meantime ; (두 가지 시점·사건들) 그 동안[사이]에 [본문으로]
197. collaborative ; [형용사] 협력[협조]적인, 합작하는; 공동 제작[연구]의 [본문으로]
198. in addition to ; …에 더하여, …일 뿐 아니라 [본문으로]
199. debug ; [타동사][VN] (-gg-) (컴퓨터) (컴퓨터 프로그램에서) 오류를 검출하여 제거하다, 디버그하다 [본문으로]
200. reputational ; [형용사] 평판의; 명성이 있는. [본문으로]
201. beware of ; …에 주의하라. [본문으로]
202. bear gifts to[for] ; …에게 줄 선물을 가지고 가다. [본문으로]
203. quip ; [타동사][V speech] (-pp-) 재담을 하다 ;; 미국∙영국 [kwɪp] [본문으로]
204. guru ; 2. (비격식) 전문가, 권위자 [본문으로]
205. bring in ; 들여오다; <이익·이자를> 가져오다; <새로운 것을> 받아들이다, 수입하다; <의제 등을> 제출하다; <협력자 등의> 참가를 의뢰하다; <배심원이> <평결(評決)을> 답신(答申)하다; 야구 생환시키다; 경찰에 연행하다 [본문으로]
206. later on ; (비격식) 나중에; (지금 이야기 중인 시간보다) 후[뒤]에 [본문으로]
207. be under way ; have started and be now progressing or taking place [본문으로]
208. techie ; [명사] pl. -ies (비격식) 기술[컴퓨터] 전문가, 컴퓨터에 열광하는 사람 [본문으로]