Wolfram Alpha, the computational search engine, had a soft launch on Friday. While some are calling this a Google killer, I personally think the two engines will complement each other and allow for information to be served depending on your needs. Wolfram Alpha provides me with some capabilities that Google did not, i.e. the ability to ask the engine a question and have it compute an answer (think the computer on Star Trek).
Depending on my context, these search engines can provide me with valuable different perspectives on the question: What is the population of Metro Atlanta?
Google’s answer: Links to Wikipedia and other sources which is useful if I want to cite a source in a blog or a paper.
Wolfram Alpha’s Answer: 4.565 million people, which is extremely useful when my father calls me and expects me to recite this number from memory.
It becomes even more useful when you need to compare the financial data of two or more companies, say Coca-Cola (KO) and Pepsi (PEP). Simply typing the two stock tickers brings up all sorts of useful financial data on recent trades, market caps, volatility, revenue per employee, etc. You can even download the data as a PDF.
6 thoughts on “Test Driving Wolfram Alpha”
A review of Bing is in order and will be forthcoming…
Looks like microsoft just launched yet another search engine. Bing!
@Jon @Mikal Lewis,
I definitely see this driven as via corporate sales. Take a look at this article from Forbes pointing out how they might do it, but note that it also agrees with Jon’s point.
I also don’t foresee ever actually clicking an ad after getting the computational answer to a question.
Each page has a source information listed, its subtle. But I think that tells you something about where wolfram gets it from
But to your point seth is awesome, its cited data is now just a list of links.
Source information that I can’t copy and paste else where, cause I can’t even highlight it.
This is a problem because in their sea of links they don’t say ‘who’ specifically is on record as claiming its 4.565 million.
Thats a problem for me.
Because the way the source information is stated, its as if they don’t want us to be able to explore this for ourselves.
@Jon agree with you they aren’t likely to be able to sell ads. I do see an enterprise model here. Though maintaining a search engine is EXPENSIVE and I’m not sure how much companies would be willing to pay for a better search for their data repositories.
Google & WA have to different results for that atlanta question which raises teh question for me. “how does WA know this?” google has 5,278,904, and WA as you said, says 4.565 million. Difference is google is linking me to something to verify its numbers.
Agreed that Wolfram Alpha’s potentially a valuable complement to Google. I’m somewhat skeptical how broadly the computational approach can go, and whether the system will be able to overcome the biases and selective information from the (almost-all-male, almost-all-white, older) designers. From a consumer perspective, the areas where it’s likely to excel seem the hardest to monetize: when people are looking for facts like this, they’re probably not going to be likely to click on ads … what’s their business model, corporate sales?