Wolfram is back: Washington Post: Wolfram Alpha Getting A Public Preview On Tuesday

[Mike 03]

(This post is from April, but the same thing happened this week, and SA is supposed to be unveiled next week--Monday or Tuesday)

MG Siegler
TechCrunch.com
Friday, April 24, 2009; 3:11 PM

When it was first unveiled in March, Wolfram Alpha, a new type of search engine created by computer scientist Stephen Wolfram, got a lot of buzz. Naturally, some people threw out the "Google killer" title ¿ but it seems to be a different beast, as it's all about knowledge search. That is to say, you ask a question, and you get an answer ¿ with Google, you ask a question and you get a link to a bunch of documents. That may sound a bit bland, and simplistic, but the select few who have seen it, seem to think it works really well and could be a game changer.

The rest of us won't know for sure until May, when it's scheduled to launch. But if you want to catch a glimpse of how it will work, the Berkman Center for Internet & Society at Harvard is hosting a webcast on Tuesday of an event with Wolfram and Jonathan Zittrain, a law professor at Harvard. The sold-out event will be streamed live on the web at 3PM EST this coming Tuesday on this page.

MORE AT LINK:

http://www.washingtonpost.com/wp-dyn/content/article/2009/04/24/AR2009042403558.html

Apparently Wolfram Alpha opens for business next week, and I'm really excited. This sounds like a whole new type of search engine--almost able to cope with abstractions and generate relevant results based on virtually conceptual questions.

This will be fascinating, indeed, to see. Will it work?

I'm well aware that he has made goofs in some of his assertions over the years, and how poorly received his book "A New Kind of Science" was, generally. But I like this guy, and I'm fascinated by his ability to come up with ideas like this.

Good luck, Stephen

[xchrom]

1. Wolf, Ram and Hart. nt

[Deja Q]

2. I was at a pre-pre-screening. Here's a transcript of the test:

Me: "I need a fluorescent tube".
Computer: "Here is a picture of a dildo painted day-glo pink that glows in black light.)
Me: "No no no! I don't want anything pornographic!"
Computer: "Here is the video from the 1985 movie 'A View to a Kill', which features fluorescent gun barrels. They're tubes."
Me: "No no no no no! No violence either! Do you know what a tube is?!"
Computer: "Yes, they were used in radios until the 1970s, and in televisions until the early 2000s."
Me: *sighs* "No, I mean lighting. Do you understand? LIGHTING!!"
Computer: "Here is the link to CitiLights Retail Outlet. I will take you there, is that where you want to go?"
Me: "No! I don't want any sodding ads!! I want a fluorescent tube!!"
Computer: "What is a sodding? Why are you interested in replacement patches of turf?"
Me: "You stupid piece of junk! Don't you know any British slang?!"
Computer: "Do you have ADD? Why is junk stupid? And now you want me to take you to www.bbc.co.uk for the latest news on 'Absolutely Fabulous'?"
Me: "Go lick a wall socket."
Computer: "I thought you said you didn't want porn? And isn't it perverted of you to see two computers engaging in foreplay?"

[Canuckistanian]

3. I'm preparing myself for the awesome paradigm shift

I'm sure this will change the world as I know it. I was getting bored with the old world anyways.

[Deja Q]

5. It'll make life easier...

Ask it for a sex thread and it explodes into a thousand million burning pieces because it doesn't know where to begin to sort everything...

[The Straight Story]

4. From what I have read so far, works about 50-50

Been keeping up on this a lot. One reporter who had access had about half the things he tried not work, but then it is in it's early stages.

Looking forward to testing it out and asking "How Sexy is the Straight Story?" :rofl:

[struggle4progress]

6. A privately-maintained database won't supplant the web-crawlers

[The Straight Story]

7. There may be a google tie-in over time:

One of the google founders was an intern for Wolfram and have heard bits and pieces about them being in talks.

Could get interesting.

[struggle4progress]

8. I could be wrong, but I don't see any prospect for a good world-wide web version of this idea

in the short-term, though maybe after many decades the necessary standards might be in place so that part of the web was organized so that such a technique might apply

What I have read about Wolfram's alpha, suggests that it is based on ideas like those underlying his Mathematica package, which means that the "facts" are coded into the private database using very stringent rules, to which certain computational techniques can be applied. The private database is also carefully curated, so that the "facts" are in some sense accurate. To generalize this technique to the wider web will require: (1) making sure that enough webpage designers code some information on their pages very carefully, consistent with a general standard and (2) having some mechanism in place to ensure that the information scattered across the web in various correctly coded pages is actually accurate. The problem with (1) adoption of standards is very slow (as many Safari browser users learned from the A-record|SVS-record fiasco a while back). The problem with (2) should be familar to anyone who regularly tries to sort through the nonsense on the web in search of verifiable facts -- a problem that exists even at supposedly curated sites like Wikipedia. Either inaccurate coding or factual inaccuracy could, I think, cause substantial degradation in results from algorithms of the sort underlying Mathematica. It is possible that a volunteer project, reminiscent of Wikipedia, could greatly extend Wolfram's database, but I expect the open-edit features of Wikipedia would be lost, since the algorithms would not be terribly tolerant if the information pages were constantly revised

My conclusion: long-term, this might indeed be a good idea, but general progress will be slow