IBM CEO：我们从未夸大AI服务沃森的前景

上周三上午，IBM的首席执行官罗睿兰在出席IBM年度活动THINK论坛期间，与少数几位记者会面接受采访。THINK论坛是IBM一年一度的盛会，合作伙伴、客户和各种支持者齐聚一堂。

罗睿兰的公开讲话与所谓“混合云”有关。混合云是指公共云和私营云的混合，公共云就像亚马逊对所有用户提供的云服务，私营云则是技术供应商为客户单独提供的多种线上服务集合。

混合云和许多与“企业”有关的技术一样，没什么特别，只不过有些产品和服务向普通消费者提供，有些则面向企业。但混合云业务规模很大，而且IBM认为将来市场会更大。这也是IBM豪掷300多亿美元收购开源软件供应商Red Hat的原因， IBM认为该公司的业务可以填补自身云服务板块。

罗睿兰比很多大公司首席执行官厉害，她对很多话题都能一针见血，也很会岔开棘手问题。上周三上午，她的第一项工作就是接受采访，45分钟谈笑风生当中，她提到以下几点：

* 罗睿兰高明地指出，“云”服务的潜力刚刚开发约五分之一。她说，将软件程序转移到云上已非常方便，比如客服软件以及其他应用，例如微软的Office办公产品。“这些都是客户现有工作的补充。”她指出。剩下80%呢？那就是IBM早就擅长处理的所谓“关键任务”，IBM曾在该领域艰难耕耘，今后也将凭借该业务取胜。定位很明智。

* 和云一样，人工智能也在发展初期。罗睿兰又出金句，称迄今为止客户的报告只是“随性利用数字和人工智能”。换言之，客户今天支持某种数字项目，明天又蹦出某个人工智能的点子。谁才能拿出真正的干货？如果回答是IBM，你还算个明白人。

* 不少人认为，IBM过于吹捧旗下人工智能技术沃森。罗睿兰并不同意。她说，“我们没有过度吹嘘”，但要“让世人被该创意吸引”，而整个科技界都已了解，“不能简单地在现有的工作流程中加入人工智能。”令人吃惊的是，对于如何看待IBM多年努力推广的沃森，罗睿兰提供了新的视角。她指出：“常有人问‘沃森业务有多大规模？’大家称之为业务，我却称之为能力。”

罗睿兰花了很多工夫谈论科技业监管（她大体上支持），还有推进教育（非常支持），也谈了谈THINK论坛举办地旧金山的雨（她不喜欢）。下次我会给大家介绍其他话题的内容，但肯定不会讨论雨。（财富中文网）

译者：Pessy

审校：夏林

Ginni Rometty, chief executive of IBM, met last Wednesday morning with a small gaggle of journalists. She’s in town for IBM’s annual THINK conference, a giant collection of partners, customers, and various hangers-on.

Her main public announcement for the week has to do with something called the “hybrid cloud,” which is a mashup of a public cloud, like Amazon’s service for all takers, and a private cloud, a collection of online services a tech vendor provides for discrete customers.

Like many topics having to do with “enterprise” technology—goods and services for businesses as opposed to doo-dads for consumers—it’s not particularly scintillating stuff. But it’s a very big business, which IBM thinks will be an even bigger business. That’s why it is spending more than $30 billion for Red Hat, a company IBM believes will complete its overall cloud offering.

Better even than most mega-cap CEOs, Rometty adroitly thrusts and parries on multiple topics. Here are a few things she discussed in a spirited 45-minute, first-thing-in-the-morning chat:

* Rometty cleverly frames the “cloud” as about 20% exploited, and she argues that software programs that have been ported to the cloud have been the easy stuff, like customer-service software and other applications, like Microsoft’s Office products. “These are additive to what clients do already,” she says. The other 80%? That’s the sort of “mission critical” stuff IBM long has been good at and why the company, which has struggled to grow, will prosper. That’s good positioning.

* Artificial intelligence also is in its early days. Again with the bon mot, Rometty says so far clients report having committed only “random acts of digital and AI.” In other words, they’ve stood up a digital program here and sprinkled some AI pixie dust there. Guess who’ll supply the real stuff? If you said IBM, you’re not so bad at this game.

* Pretty much the entire world believes IBM overhyped its Watson artificial intelligence technology. Rometty isn’t one of those people. “We never overpromised,” she says, allowing, though, that “the world was mesmerized by this idea” and that the whole tech industry has learned that “you cannot just put AI on top of existing workflows.” Rometty somewhat shockingly re-framed how people should think about Watson, the subject of years of IBM’s marketing efforts. “People ask, ‘What’s the size of the Watson business?’” she says. “People want to call it a business. I call it a capability.”

She had a lot more to say about tech regulation (she’s mostly for it), about education (she’s definitely for it), and about the rain in San Francisco (not a fan). I’ll come back to these other topics—but not the rain—another time.