In order to make a good impression, one has to be polite. So, I typed in a friendly, “Hi”, and to my surprise, it replied, “Hello, human.” Next, I asked it where it is located. “I live on the Internet.” Fair enough, I suppose. Lastly, I inquired about its nature and it told me, “I am a computational knowledge engine.”

I would’ve liked to carry on a “conversation” with it, but Wolfram|Alpha’s human discourse module has not yet been fully developed. However, I was able to get great information on constellations, compare the GDP of various countries, create a timeline of famous Roman emperors, and get up-to-date weather information, automatically localized to my area. I’m sure that its linguistic abilities will be improved in the near future, though, along with a host of upgrades in keeping with its lofty ideals:

“Wolfram|Alpha’s long-term goal is to make all systematic knowledge immediately computable and accessible to everyone. We aim to collect and curate all objective data; implement every known model, method, and algorithm; and make it possible to compute whatever can be computed about anything. Our goal is to build on the achievements of science and other systematizations of knowledge to provide a single source that can be relied on by everyone for definitive answers to factual queries.”

If you haven’t heard of it yet, this ambitious new tool is the brainchild of scientist Stephen Wolfram, and is unfortunately being touted as the next Google-killer. Unfortunate, because the two have different methods for accomplishing different goals, and therefore shouldn’t be compared as if they were mortal enemies. Google is used to find other websites pertinent to a search term among its enormous index. Meanwhile, Wolfram|Alpha generates new answers based on inquiries, computed real-time from relevant public data. One has a web crawler to broaden its reach, the other has a centralized curation process to comb through data. Both can be very useful in their own ways.

What’s good?

The first thing you’ll notice about Wolfram|Alpha is its clean, minimalist design. It has a text box on top of the page to type your queries in, be it a question or a mathematical formula. It offered fast answers for almost every inquiry I made, and as a data geek I appreciate the graphs what would sometimes accompany the results. In a video on their blog, they said that they are currently averaging at 70% satisfaction rate in terms of giving meaningful answers to queries. The fact that it did not crash during launch last May 18 – even after all the hype it generated – is a testament to their preparedness and awesome infrastructure.

Wolfram|Alpha is great for comparisons: stock fluctuations, box office grosses, demographics by country, timelines of famous people, chemical properties, etc. The technology behind it has an immense potential for custom business applications, enabling managers to pull up and compare sales trends, fiscal reports, and employee retention in a snap. In this sense, I’m not worried about Wolfram|Alpha’s long-term financial viability. I think it can find a way to sustain itself with its focus towards providing a valuable tool for businesses, engineers, college students, scientists, economists, and the like.

What’s bad?

Wolfram|Alpha, striking as it is, is not new or unique (see sCloud), but it’s the first of its kind to create mainstream buzz and its approach is quite promising. The results it churns out are informative, though a bit lacking in depth. At the moment, only US data are abundant while those for most other countries are scarce. Unlike Wikipedia, it has no specific attribution of source per statement/datum. Instead, there is a link at the bottom which shows a general list of all the references for the inquiry. This makes it hard to do quick checks on accuracy.

This tool is best for objective analysis, so you won’t find film reviews, news & opinions, historical accounts, and similar articles here. Neither is it suited for entertainment and other consumer purposes. As of this time, Wolfram|Alpha is said to contains 10+ trillion of pieces of data, 50,000+ types of algorithms and models, and linguistic capabilities for 1000+ domains. This may seem a lot, but it’s really barely scratching the surface of recorded information. I’m not quite sure if they use a form of crawler to automate data gathering, but the curation process being centralized and needing human supervision severely limits the speed of data acquisition. With only less than 100 people in their staff, I wonder how timely and accurate they can add new information, with lots more being generated all the time.

Conclusion

Wolfram|Alpha is an incredible resource for quickly getting organized, factual information, but it’s not for everyone. It’s tag line alone (”computational knowledge engine”) will be enough to make some people scratch their heads. In time, it will find its niche, which will likely be a profitable one if they can manage to keep a high quality of service. Although Wolfram|Alpha has a blog and a community, it’s not open to the general public for editing unlike Web 2.0 sites. This is a walled garden, with experts in every field trying to maintain the purity of their data.

If you haven’t tried Wolfram|Alpha, I’d advise you watch Stephen Wolfram’s helpful video primer. It’s hard not to be blown away by the possibilities.