投资银行Piper Jaffray的吉恩•蒙斯特非常严谨。大家排队抢购苹果（Apple）新产品时，他是那种会跑到现场清点人数的分析师。程序员们济济一堂，参加苹果的开发者大会时，他又会拦下几十号人，询问他们是在为Mac、iPhone，还是Android设备开发应用程序。

    所以，当明斯特想知道苹果智能语音助手Siri和谷歌搜索究竟谁更厉害时，他（和他的助手）于是向iPhone提出了1,600个问题，其中800个是在明尼阿波利斯的繁忙嘈杂街道上进行，另外800个则是在安静的室内。

    Piper Jaffray在上周四致客户的报告中公布了测试结果：

• 谷歌理解了100%的问题（这毫不意外，因为他们是打字输入的问题）

• 谷歌回答的总计正确率是86%

• 在嘈杂的环境中，Siri理解了83%的问题；在安静的室内，这一数字为89%

• 在嘈杂的街道和安静的室内，Siri的正确率分别是62%和68%。

    蒙斯特写道：“如果想成为切实可靠的另一种移动搜索工具，Siri必须达到、超过谷歌目前的正确率B+，它必须从D提升到B，甚至更高的一档。”

    虽然蒙斯特推测Siri在学习曲线上要比谷歌落后两年以上，不过他十分乐观：“随着今年秋季iOS 6的推出，我们预计Siri将会取得显著的进步，而它对谷歌的依赖程度也将从目前的60%降至48%。”

    不过，鉴于测试结果，我们现在并不清楚减少Siri对谷歌的依赖性究竟是不是好事。

    目前Siri的答案中，有60%来自谷歌，20%来自Yelp，14%来自计算数学应用软件开发商WolframAlpha，4%来自雅虎（Yahoo），另外2%则来自维基百科（Wikipedia）。

    蒙斯特表示：“如果进一步细分Siri的依赖关系，我们看到，谷歌提供了100%的导航结果、61%的信息结果、48%的商务结果和42%的本地结果。在其余搜索聚合商中，Yelp贡献的本地结果（51%）和商务结果（51%）最多，而WolframAlpha则贡献了34%的信息结果。”

    从今年秋天开始，苹果将启动自有地图服务，对谷歌地图的依赖将彻底成为历史。而随着雅虎体育、Open Table、烂番茄（Rotten Tomatoes）和Fandango等服务的整合，苹果用户将获得体育比分与统计数据、餐厅预定、电影放映预告和票务等诸多服务。

    蒙斯特将Siri所犯的错误进行了分类，通过下面他提供的一些例子，我们可以看到Siri还有多长的路要走。

    Pi per Jaffray's Gene Munster is nothing if not methodical. When customers queue up for a new Apple (AAPL) product, he's the analyst who goes to the line and counts heads. When programmers gather at an Apple developers conference, he'll stop three or four dozen to ask if they are writing apps for Mac, iPhone or Android devices.

    So when he wanted to know how Siri stacked up against a Google (GOOG) search, he (or his staff) asked an iPhone 1,600 questions, 800 on the busy streets of Minneapolis, 800 in a quiet room.

    Piper Jaffray published the results Thursday in a note to clients:

• Google understands 100% of the questions (not surprisingly, since they are keyed in)

• Google replies accurately 86% of the time

• Siri comprehends 83% of queries in noisy conditions, 89% in a quiet room

• Siri answers accurately 62% of the time on the street and 68% in a quiet room.

    "In order to become a viable mobile search alternative," Munster writes, "Siri must match or surpass Google's accuracy of B+ and move from a grade D to a B or higher."

    He estimates that Siri is more than two years behind Google in its learning curve, but he's optimistic: "With the iOS 6 release in the fall, we expect Siri to improve meaningfully while reducing its reliance on Google from 60% to 48%."

    Given the results, however, it's not clear that reducing Siri's reliance on Google is necessarily a good thing.

    Currently Siri gets 60% of its answers from Google, 20% from Yelp, 14% from WolframAlpha, 4% from Yahoo and 2% from Wikipedia.

    "Breaking down Siri's reliance further," Munster writes, "Google provides 100% of navigation results, 61% of information results, 48% of commerce results and 42% of local results. Among other result aggregators, Yelp provided the most local results (51%) and commerce results (51%), while WolframAlpha provided 34% of information results."

    Starting in the fall, Apple's in-house maps will eliminate its reliance on Google for navigation and the integration of Yahoo Sports, Open Table, Rotten Tomatoes and Fandango will provide answers for sports scores and statistics, restaurant reservations, movie show times and ticket purchases.

    Breaking down the type of errors Siri makes, Munster provides some sample questions that demonstrate how far it has to go: