… and based on what I read, I am not likely to use it.
I know the hype around it. The very first review (By Talbot of Technology Review) I read about it did tell me why people gets excited, but it also tells me why I am not going to use it.
One rightly gets excited when one’s search engine apears to understand what you are looking for and give you the results you want. So far, no search engines does that. If you read the first page of the Technology Review article, one will get excited. The search engine appears to try to interpret your intention. This can be useful at times. I am still constantly frustrated with Google insisting on serving me unrelated links because I cannot refine my search term to be specific enough to achieve my goal. My last attempt was trying to find the unix command that will list all the functions available in a Library file (answer: nm). It took me 15 minutes brainstorming possible search term.
Did I get excited? No. First, I immediately realized I am looking at a design flaw seen in most Apple products. If the software interprets your intention correctly, everything is just amazing. If not, it’s beyond frustrating. In Apple’s case, if you want to do something that is not the “norm”, e.g. changing your default file format for “ripped” music to mp3 for iTune, you will find it a frustrating exercise. The normally intutitive iTune design suddenly works against you. With Wolfram Alpha, take the Talbot’s first search term “Microsoft Apple”. If you are not interested in the business performance of Microsoft and Apple, the page is badly presented. In Wolfram Alpha’s defense, Google did very poorly if you are interested in the business performance of the two company, and it did try to please as many people as possible by populating the page with diverse range of information.
Second, I also immediately realized that the data is curated. The technologist in me told me so once i saw slick search results presentation. I am that one person that do not like curated data. Curated data is good when you can trust the source, and i can trust Wolfram to get it correct. However, most of the time I find myself wondering whether the data is up-to-date, especially in absence of supporting information such as the year the data is collected etc. Also, curation takes time, and this immediately tells me that the scope of the WolframAlpha is going to be limited, which it is.
I was just about to write Wolfram Alpha off when I read this Economist’s article, which makes me realize the potential of the software. WolframAlpha is not going to topple Google anytime. It lacks the breath and scope of Google, and cannot hope to catch up. For me, being a curated work means it lack objectivity. However, if breath, scope, objectivity is not a problem and you can reasonably guess the context of the search, then I see its value. For example, as a corporate search engine shifting through corporate data, and answering the same questions from sales department, IT department and management differently. This will requires an automatic way of shifting through the data, rather than curation by hand as WolframAlpha now do. However, let’s not forget it is only the early days of the application and an automated curator is just a matter of time. It is also easier to create curators for corporate data, as the scope and the context is more limited. Watch out Autonomy!
From a business perspective Wolfram Research is probably on to something that can be turned into $$$.
I agree with Economist that at present, it is more a competition for Wikipedia instead of Google. However, while I can see it takes some eyeballs away from wikipedia, again it lacks the breath and scope of Wikipedia (but this time, it might catch up), and whatever wikipedia’s flaw, I believe it is likely to be less biased than WolframAlpha.