There’s been so much hype about the would-be, Google-killer, computational knowledge engine Wolfram Alpha, that I’ve been reluctant to blog about it both because I’m not eager to contribute to the hype tsunami but also because I’m so overwhelmed by the number of articles I’ve bookmarked that it would take hours to sort them out into a coherent post.
To give you an idea of how much Stephen Wolfram thinks his work, the “quick introduction” video is 13 minutes and 23 seconds.
In short, I think Wolfram Alpha is “Powerset meets semantic web.” What do I mean by that?
- Powerset was all about natural language for query formulation and the non-elimination of stopwords. They accused search engines of making you grunt in “pidgin English,” eventually leading to the creation of grunting pidgeons t-shirts.
- Semantic web is, in my opinion, all about the web as queryable database.
Wolfram Alpha is about putting those two ideas together. Wolfram Alpha will use the contents of the web to get you an answer — as opposed to a link that might contain an answer — to your question. (It is like MarkLogic in returning answers, not links.)
But with all the hype and pre-positioning, I think this thing is way more likely to be the next Cuil (total crater) or the next Powerset (expedient early $100M exit to Microsoft) than the next Google. Ironically, Google itself didn’t come with massive pre-hype. They just built up a great business.
So, rather than take a deep, heavy approach to Wolfram Alpha, I thought up a fun example instead. I asked Wolfram Alpha: how much wood could a woodchuck chuck if a woodchuck could chuck wood?
Here’s what Wolfram Alpha said, which wasn’t bad:
More information:
- Wake us when Wolfram Alpha can solve an actual problem (Valleywag)
- Wolfram Alpha: The next Google or the next Cuil (Mashable)
- Harvard posts the Wolfram Alpha video (TechCrunch)
Interestingly when I asked following on Wolfrom Alpha, “how much was Dow Jones industrial average on 21st April, 2009” it just said “Wolfram|Alpha isn’t sure what to do with your input”. But when I asked it on google first link took me to hubdub.com.
Totally agree on the hype of Wolfram Alpha. Reminds me of the hype that surrounded the Segway.
Agreed on all counts – I really want to find a reason to like Wolfram|Alpha, but am unable to grok its sweet spot. Wikipedia has a community, and some form of wiki-slicing will no doubt become indespensible for teachers at all levels. Google delivers, and builds innovation onto success – I am completely enamoured with Google’s “Wonder Wheel” as a search refinement tool which is appropriate for my 5 yr-old.Not sure if WA has applications for learners. I would personally appreciate a semantic search engine that could fact check for me, for instance. From a young learner’s perspective, wikipedia and google search tempt students towards the slippery slope of plagiarism. WA does not.
Not sure I’m with the hypers yet Dave, there are so many inaccuracies I fear it needs an element of Wikipedia behaviour to help out. Which rather negates the purpose.
The query “How many roads must a man walk down before you can call him a man?” also returns an intelligible answer.The technology might not yet be there for the sorts of semantic searching we would all like to see (to say nothing of the tagging and standards necessary to really pull it off) but this is a good start.
Mike:The answer, my friend, is blowin' in the wind, the answer that Wolfram Alpha gave.It does a bit less impressively on "how many seas must a white dove sail before she sleeps in the sand."(Remapping it to the same question about roads and men.)Thanks,Dave