2014年大数据一定不会发生的4件事和一定会发生的一件事

大数据环境下，对大数据进行揭露和预测可谓是呈现出百家争鸣的情形。但Jordan Novet为此却给出了不同的见解，主要从这些角度进行阐述:

1)大数据来临之际，是不是每一家公司都需要聘请一位首席数据官（CDO）？

2）对于（上市）公司来说，将如何去面对大数据？

3) 数据科学应用软件将会取代数据科学家吗？

4）借着大数据的东风，google将会成为新的美国黑水保安公司（Blackwater）吗?

5）另外，他还希望未来有关大数据的应用将会给消费者带来更多前瞻性的体验和吸引力。

该文作者是Jordan Novet，Jordan Novet是美国科技新闻博客网站作者，同时也是商业分析机构VentureBeat 特约撰稿人，主要撰写与大数据、云计算和其他商业技术相关的文章。

今年，许多高管，分析师，还有许多作者已经在对大数据进行揭露和预测，但是我们认为其中一些人的想法看起来却有点稀奇古怪、不切实际。仅仅为了验证我们的猜疑合乎情理， VentureBeat公司和LinkedIn公司的前首席数据科学家Pete Skomoroch进行了一次对话。

在谈话中，Skomoroch证实了我们的猜疑，同时对大数据预测提出了他自己的一些相反的观点。这些观点如下：

不，并不是每一家公司都将会聘请一位首席数据官（CDO）。上周，商业智能（BI）初创公司Birst的首席执行官布拉德·彼得斯认为这个职位的“光辉时刻终于到来了”。 但Skomoroch并不认为CDO这个职称在今年将会得到广泛采纳接受。即使有些公司在今年真的决定填补这个职位空缺，而Skomoroch就仅仅只是想知道增设这个工作岗位将会对公司带来什么样的影响。如果CDO向公司首席执行官汇报工作时，那么这家公司或许会受到一定影响。而Skomoroch却说道，“如果CEO把所有精力都倾注在某一个旁支机构上，那么设立CDO这个工作岗位是否会大获成功还尚无明朗”。

不，公司并不会在盈利报表上列出他们到底拥有多少数据。虽然在上个月的VentureBeat公司 2013年 DataBeat/Data Science③ 峰会活动中，Skomoroch说道人们在未来将会 “开始去查看资产负债表中那些【上市】公司的数据资产是什么”，但是他却还说道，今年在实际情况中很可能很难去做到这一点。Skomoroch也确实把数据当成是一种资产。而且数据越多，越好；数据越详细，越好；数据越稀缺或越受欢迎，越好。虽然如此，可是计算一家公司保存在许多数据仓库中的数据总量仍旧不是一件轻而易举的事。并且还可能很难将数据报告标准化、规范化。最后，就是有些公司可能不想透露他们所拥有的宝贵数据。

不，数据科学应用软件无法取代真正的实时的数据科学家。虽然初创公司可能会努力把在大量数据当中发现趋势和异常的这一过程使其自动化，但是软件最有可能还是无法复制经过数据科学原理和最佳实践训练的人的能力。“自动化”这整个概念让Skomoroch想起了多年前软件外包的这一主张。虽然初创型公司如Trifacta为了进行数据分析会缩短数据准备的时间，但是以后的工作仍然会用到（具有专门技能的）专业人士的意见和建议。而且对Skomoroch来说，创造力，直觉，这些“人的因素”仍旧显得极其重要。他说道，“（因为）这是一种知识工作（knowledge work）”。

不，google并不会成为一家国防承包商。尽管google参与了美国国防部高级研究计划局举办的机器人挑战赛（简称DRC大赛），并最近还完成了对机器人工程公司波士顿动力（Boston Dynamics）的收购，但是Skomoroch却说道，“google将并不会成为新的黑水保安公司（Blackwater）⑧”。除此之外，Skomoroch还说道，“我认为google所做的这一切都是和吸引人才有关，和吸引致力于研发自动驾驶汽车的人才有关，以及和他们正在研发中的其他项目有关”。有人可能会说从一开始google就一直在做涉及到大量数据的工作，并以此改进公司业务和工作流程。想一想google搜索（Google search）、google邮箱（Google mail）、google收件箱过滤器（Gmail inbox filters），就可知道都是与大数据有关的项目。而目前google正准备更多地是去研发与硬件相关的项目。

所以我们应该期望看到未来将会是什么样的呢？Skomoroch认为大概在智能代理（intelligent agents）在人机对话技术中应运而生的20年之后，智能代理对消费者来说将会变得更具吸引力。

1216比如，Google Now 和苹果公司的Siri应用已经吸引了许多消费者，而且在未来还应该会继续吸引更多的用户。

Skomoroch说道，“我认为下个阶段将会更具前瞻性”。“这实际上将会更多地帮助我们委托事情或处理事情，以致于我们其实真的并不需要亲自去做”。为此，他还举出了华尔街（Wall Street）高速度的、算法驱动的交易模型作为例证。

他，和试图要把这个模型做出来的许多初创公司一样希望今年能够再深入具体些。同时他还说道，“我并不认为这个模型在2014年将会取得怎样巨大的飞跃，但是我相信你将会对它产生足够的兴趣”。

英语原文：

Executives, analysts, and, yes, writers have been spilling out predictions for big data this year, but some of them seem a bit outlandish. Just to be sure we weren’t crazy, VentureBeat checked in with Pete Skomoroch, former principal data scientist at LinkedIn.

Skomoroch confirmed our suspicions and added a few anti-predictions of his own. Here they are:

No, not every company will hire a chief data officer. Last week Brad Peters, chief executive of business-intelligence startup Birst, argued that the position’s “time to shine has finally come.” Skomoroch doesn’t think the adoption of the CDO title will be widespread this year. But even if some companies do decide to fill the position this year, Skomoroch wonders just what difference such a move would make. If a CDO reports to the company’s chief executive, perhaps the company could have an impact. “If you’re buried in, like, a side unit, it’s not clear that you’re set up for success,” he said.

No, companies won’t list how much data they have on earnings statements. Skomoroch himself said at VentureBeat’s 2013 DataBeat/Data Science Summit event last month that down the line people would “start to see on balance sheets what are the data assets of those [public] companies.” But that practice probably won’t take hold this year, Skomoroch said. Skomoroch does view data as an asset. The more of it, the better. The more detailed, the better. The more rare or sought after it is, the better. Still, calculating the total amount of data a company retains inside its many storehouses isn’t a breeze. And it might be hard to standardize the reporting of data. Finally, some companies might not want to reveal what jewels of data they possess.

No, data-science applications won’t replace real, live data scientists. Startups might try hard to automate the process of discovering trends and anomalies in large volumes of data, but software most likely will not be able to replicate the capabilities of humans who are trained in data science principles and best practices. That whole notion reminds Skomoroch of an idea years ago that software would be outsourced. Startups such as Trifacta could slim down the time it takes to get data ready for analysis, but later work still could use a skilled person’s input. Creativity, intuition — “the human element” — still matter a great deal for Skomoroch. “This is knowledge work,” he said.

No, Google won’t become a defense contractor. “Google will not be the new Blackwater,” Skomoroch said. This despite the company’s involvement in DARPA robotics competitions and its recent acquisition of Boston Dynamics. “I think it’s about talent and acquiring the market on robotics and the type of people who can work on self-driving cars and other things that they’re building,” Skomoroch said. One could argue that from the beginning Google has been about taking in lots of data and modifying operations as a result. Think of Google search, think of Google mail, think of Gmail inbox filters. It’s a big data thing. Now it’s just becoming more about hardware.

So what should we expect to see, then? Skomoroch believes intelligent agents will become more compelling to consumers, about two decades after they emerged in technology conversations.

Google Now and Siri have caught on with consumers, and they should take hold further. But there’s more.

“I think the next phase will be more proactive,” he said. “It will actually be more about helping you delegate things or do things so that you don’t actually have to do them.” Skomoroch pointed to high-speed, algorithm-driven trading on Wall Street as an example.

He expects the model to trickle down a bit this year, with startups trying to put it into action. “I don’t think it will be a huge leap in 2014, but I think you’ll see enough interest in this,” he said.