Projo Subterranean Homepage News

Nova Spivack rattled the blogosphere Saturday with Wolfram Alpha is Coming -- and It Could be as Important as Google. He has been allowed to peek at Wolfram Alpha, a new way to find answers to factual questions -- a brainy computer that computes the answer rather than finds it on a Web page. The goal here is the Star Trek computer, holding the sum of all knowledge, with specific answers retrievable by asking unambiguous questions. The difference: Majel Barrett's voice never says "Cannot find answer"; she says, "Does not compute."

Fast Company explains (Is Wolfram Alpha The Next Big Thing in Web Search?),

But apparently the system's not like previous efforts at this technology (ahem, Ask Jeeves), which use natural language parsing to determine your question and then simply present the web-search results. Instead Alpha is supposedly revolutionary since it actually computes the answer for you. Inside it has built-in models of how the world works in terms of science, geography, business, people and so on, and it interprets your question and uses its models to calculate an answer.

In fact what Wolfram and his team have done is, after a lot of research, broken down the problem (answering a fact-based question) into a series of small steps, each of which is reproducible and can be used for interpreting different types of question and finding the corresponding answers. And these steps have been programmed into the Mathematica computing engine, along with a natural language program and the "real world" database.

Nova Spivack's Webshaking post at Twine begins,

Stephen Wolfram is building something new -- and it is really impressive and significant. In fact it may be as important for the Web (and the world) as Google, but for a different purpose. (March 9, Spivack added the following sentence) It's not a "Google killer" -- it does something different. It's an "answer engine" rather than a search engine.

Stephen was kind enough to spend two hours with me last week to demo his new online service -- Wolfram Alpha (scheduled to open in May).

...In a nutshell, Wolfram and his team have built what he calls a "computational knowledge engine" for the Web. OK, so what does that really mean? Basically it means that you can ask it factual questions and it computes answers for you.

It doesn't simply return documents that (might) contain the answers, like Google does, and it isn't just a giant database of knowledge, like the Wikipedia. It doesn't simply parse natural language and then use that to retrieve documents, like Powerset, for example.

Instead, Wolfram Alpha actually computes the answers to a wide range of questions -- like questions that have factual answers such as "What is the location of Timbuktu?" or "How many protons are in a hydrogen atom?," "What was the average rainfall in Boston last year?," "What is the 307th digit of Pi?," "where is the ISS?" or "When was GOOG worth more than $300?"

Think about that for a minute. It computes the answers. Wolfram Alpha doesn't simply contain huge amounts of manually entered pairs of questions and answers, nor does it search for answers in a database of facts. Instead, it understands and then computes answers to certain kinds of questions.

Unfortunately, it's not possible to understand Wolfram Alpha by trying to stump it, since it hasn't launched yet.

One person looked at it and wrote about what he saw, and the kinds of questions he asked it make it sound like a homework engine.

In a followup titled Answer Engines vs. Search Engines, Spivack attempts to clarify:

The approach taken by Wolfram Alpha -- and others working on "answer engines" is not to build the world's largest database of answers but rather to build a system that can compute answers to unanticipated questions. Google has built a system that can retrieve any document on the Web. Wolfram Alpha is designed to be a system that can answer any factual question in the world.

Of course, if the Wolfram Alpha people are clever (and they are), they will probably design their system to also leverage databases of known answers whenever they can, and to also store any new answers they compute to save the trouble of re-computing them if asked again in the future. But they are fundamentally not making a database lookup oriented service. They are making a computation oriented service.

I have read the several articles, all their comments, all the tweets, and understand the difference between Google's pointers to Web pages that contain the keywords in your question -- and might have the answer -- and "Wolfram's team manually entered, and in some cases automatically pulled in, masses of raw factual data about various fields of knowledge, plus models and algorithms for doing computations with the data": There are human editors frontloading human knowledge to a computer with the ability to correlate it.

But until I can frame some questions and ask them -- play Stump the Alpha," basically -- I know nothing about the thing itself. But I don't think I can get a second opinion there, or know if its answer is wrong.

I shot an email to Shelley Powers (Burningbird) in St. Louis, my go-to guru.

Knowing your interest in the semantic Web, I'm wondering what you think about the Wolfram "computational knowledge engine."

Her reply:

First of all, it's not a new form of Google. Google doesn't answer questions. Google collects information on the web and uses search algorithms to provide the best resources given a specific search criteria.

Secondly, I used Mathematica years ago. It's a great tool. And I imagine that WolframAlpha will provide interesting answers for specific, directed questions, such as "what is the nearest star" and the like. But these are the simplest of all queries, so I'm not altogether impressed.

Think of a question: who originated the concept of "a room of one's own". Chances are the Alpha system will return the writing where the term originated, Virginia Woolf's "A Room of One's Own", and the author, Virginia Woolf. At least, it will if the data has been input.

But one can search on the phrase "A room of one's own" and get the Wikipedia entry on the same. So in a way, WolframAlpha is more of a Wikipedia killer than a Google killer.

Regardless, when you look via Google, then you get link to Wikipedia, but you also get links to places where you can purchase the book, links to essays about the original writing, and so on. You don't get just a specific answer, you also get context for the answer.

To me, that's power. If I wanted answers to directed questions, I could have stayed with the Britannica years ago.

Nova Spivack's writing on the Alpha is way too fannish. And too dismissive of Google, not to mention the human capacity for finding the exact right answer on our own given the necessary resources.

Again, though, all we have is hearsay. We need to try the tool out for ourselves. But other than helping lazy school kids, I'm not sure how overly useful it will be. If it's free, yeah. If it's not, it will be nothing more than a novelty.

We are all talking aobut something only Nova Spivack (and the 100 stealth knowledge workers creating it) have seen.

Yesterday, I asked Google how to cook brown rice in a crockpot, a typical human question that should have a factual answer. Spivack says cooking is one of the areas of knowledge Wolfram Alpha can serve answers about.

The number one result had no ratings, often a bad sign. The number two result: Slow Cooker Foolproof Rice, crows,

Author Notes: You'll never again have to deal with anything other than perfect rice when you make it like this in your slow cooker.

1 cup regular raw, white or brown rice
1 cup water
2 teaspoons salt
1 tablespoon butter
Instructions:

Place all ingredients together in the slow cooker. Cover and cook on low for 6 to 8 hours.
This Foolproof Rice recipe really does work, even with the small amount of water required. Covering the slow cooker and setting it on low effectively turns your slow cooker into a rice steamer. The rice will have the consistency of steamed rice and will not be the kind of "sticky Rice" that you would get from cooking it in a pot on the stove top.

There was one rating/comment below the recipe:

No, it's not "Foolproof"- this doesn't work at all! When I tried this, I was here in the house, and had it going for about 2 hours, and the rice burned. MAYBE I'll try this on the "Warmer" feature next time, if I have extra rice.

May 1, I'll hit Wolfram Alpha with my crockpot brown rice question. I hope their definitive answer is not the one that didn't work.

(I backed off the crockpot method in favor of Foolproof Oven-Baked Brown Rice, a tested Cook's Illustrated recipe at Alanna Kellogg's Veggie Venture. I'll report back after I try it. )

Update: Shelley Powers references our exchange and looks for her own crockpot brown recipe in a post titled simpley WolframAlpha. Important:

I also beg to differ with Nova, when he states that Wolfram Alpha is like plugging into a vast electronic brain. In fact, from what I can deduce about WolframAlpha, it isn't brain like at al.

The human brain is amazing in its ability to take bits and pieces of data and derive new knowledge. We are capable of learning, and extending, but we're really shite, to use the more delicate English variation of the term, when it comes to storing large amounts of data in an easily accessible form.

...Alpha will return direct answers, while Google and other search engines return resources from which we can not only derive answers, but also make new discoveries. As such, Alpha could be a useful tool, but I'm frankly skeptical whether it will become as important as Google or other search engines, as Nova claims. I don't know about you all, but I get as much from the process of discovery, as I do the result.

Background:
Nova Spivack (ironically, from Wikipedia)
Stephen Wolfram
Wofram Blog: Wolfram|Alpha Is Coming!

 Tags  future tech google Shelley Powers Wolfram Alpha

---

2 Comments

Andrea said:

Remember what happened in the 1957 Katherine Hepburn and Spencer Tracy movie, Desk Set, when EMERAC was asked a question it couldn't answer.

It went haywire.

Spencer, the computer's creator, then fixed it with a hairpin borrowed from Katherine, the librarian and human brain. Who already knew the answer.

Link | April 28, 2009 3:20 PM
Report Abuse

---

Sheila Lennon said:

I think I remember a Star Trek episode in which a computer, after being asked, "Why?," whirred itself to death.

Link | April 29, 2009 4:16 AM
Report Abuse

---

Leave a comment

Name:

Email:

URL:  

Remember personal info?

Captcha:

Type the characters you see in the picture above.