Computer scientist Stephen Wolfram gave a report today listing what the team at Wolfram Alpha, his new search engine, did this summer. They added new knowledge domains and over 2 million lines of code, classified 54,233 bugs and suggestions, and generally fixed what doesn’t work. (”Close to half the time that Wolfram|Alpha doesn’t give a result, it’s . . . because it doesn’t understand what’s being asked.”)
What he left out of his school report was the most interesting thing that actually happened to Wolfram Alpha this summer. After long talks, it finally struck a deal with Bing to license some of its data, according to sources close to Wolfram. If nothing else, this is yet another jab at Google, which has a geek rivalry with Wolfram over the growing area of using structured data to improve search results.
In fact, there was buzz among insider search circles back in May that Wolfram was going to be part of the initial Bing launch announcement. At that time, I put the question to Microsoft senior VP Yusuf Mehdi who hemmed and hawed. “We are talking to a lot of different folks,” he told me. The only thing he denied was that Wolfram would be part of the initial launch, which it wasn’t.
The original excitement around Wolfram’s launch quickly died down, whereas Bing’s continued to accelerate (see a comparison graph of both launches). It helps, of course, to have a $100 million marketing budget and be backed by Microsoft, but Bing actually seems to be striking a chord with regular search consumers.
Maybe that is because Wolfram Alpha is not as approachable as Bing. It is good at “computing answers” to arcane questions, and has some very impressive technology under the hood. But it has a long way to go before it can deliver results that really Wow you. As Paul Carr puts it, Wolfram Alpha is the “technological equivalent of a boring uncle; the method was more impressive than the effect, and so the hairs on the back of my neck remain unstood.”
The main problem seems to be one of presentation and effect, areas where Bing does very well. One of Bing’s main strategies is to present different types of information in different ways. Travel search results look different than product or image search results. Perhaps Bing’s deal with Wolfram is to license some of its data to create a specific science category search or a Q&A portion of the site.
Whatever it is, if it turns out to be popular, Bing might end up licensing more data for more categories of search. In the end, Wolfram could have more luck licensing its data to other search engines than bringing people to its site, despite the surge in “fall traffic” Stephen Wolfram is still hoping for.
6th Link is broken, ttp: not http:
thnx, fixed
Also, the report was yesterday, not today as your first sentence says.
Great move by Bing! I hope more will come!
Bing on phones, xbox, macs, surfaces, etc, etc, etc…
GO BING!
GO bing!
great!
The main point with Wolfram-alpha is, only a sepcific group of people need it. I mean scientists, engineers, etc. So it’s ignored by most internet users.
Couldn’t agree more.
Same thing wth Yauba … a great search engine, but who cares about privacy?
AAfter search is better for privacy because it does not use java script based ad systems like others. It also links to Wolfram in case you decide you need it.
true. only a septific group of people would ever use it. the future os search is about natural language location, personalization and integration. less is more, the simpler the better.
it realy is
The only thing he denied was that Bing would be part of the initial launch, which it wasn’t.
You’re talking about the launch of bing there, aren’t you?
ohh my f**** gawwd…. the game’s getting bigger n bigger…:) wow..:D
Anecdotal evidence; me and the wife discussed my BMI and how it’s calculated – I smiled and said, let’s ask Wolfram. And sure enough, on asking Wolfram BMI (weight in kilos) (height in cm) We got our BMI, after Wolfram asked us to clarify biological sex. I’m convinced that Wolfram has shown us the way.
“The only thing he denied was that Bing would be part of the initial launch, which it wasn’t.”
Someone needs their morning cup o coffee.
yes, pre-coffee slip. Thanks for the catch. Fixed now.
Wolfram Alpha sucks. No really, it does. It is terrible at the same thing that Google and Bing and Yahoo and every other search engine is terrible at: understanding the context of data during a query.
For example, they all completely fail at understanding the difference between the numerical context of a number like 3.14 (pi) and $3.14 (the price of a toothbrush). The human brain can see these differences but the search engines cannot.
So, what is needed is a structured XML language of some sort that allows data sets to be described so that the search engines can apply meaningful programmatic context to the data they are searching against.
This is a huge problem that only the complicated Business Intelligence solutions from IBM, SAP, etc. have solved — all of which require massive money and time commitments to implement. Only enterprise-level clients can afford these solutions and every instantiation is very customized to the client.
I thought that Wolfram Alpha would solve this, but from what I’ve seen so far, WA is just a documented extension of Stephen Wolfram’s brain and team of data researchers. If they have not thought of the questions and programmed it into WA, then you will not get any decent result. This does not scale.
I have found only one discussion so far that examines WA from this perspective and I feel the author is dead-on, but he does not explain how to solve WA’s problem, he only identifies it.
“Wolfram Alpha and Hubristic User Interfaces” by Mencius Moldbug.
I know of only one company, currently in stealth mode, that is solving this exact problem.
The UR blog post is great!
Let’s take a different look at this(simplified):
Information = Data in context
Context = Organized Data
Learning = Self organization of Data
Means we have all a slightly different view on things.
Context is local (you), meaning your computer doesn’t know what you have learned and when. But there is certainly overlap, shared context. Things we have learned like anybody else in our culture/ environment.
One can actually let a machine learn about PI and $3.14. But if your algorithm requires teaching it, you will most likely end up with a lot of job security since you will tweak your algorithm all the time. Teaching should be about acceleration of learning not enabling.
Google search works because people have put data into context, links (organized related data, most of the time). WA is just a really, really small subset of this, defined by an even smaller organization group.
There is really no “smarts” to it, as far as I can see. Smarts meaning the machine has self organized the data.
.
.
.
“I know of only one company, currently in stealth mode, that is solving this exact problem.”
I bet that company is yours. Nice try but it’s obvious.
Here to make a narky comment on the broken link and the confusing statement on Bing not announcing Bing as part of Bing’s launch.
Unfortunately I failed in mentioning it first.
Oh well, it’s Friday night (at least here in Hong Kong) and at least I have the excuse of some few (OK 5) pints of Tsing Tao Beer as an excuse for my sloppy self editorial skills.
I think a fall surge is a real possibility, especially with a little careful marketing (like getting mentioned in TechCrunch). Look at the drop in Wolfram use on the chart, sure, it makes the launch just look like a spike, but it also happens in summer. Wolfram will be very useful to returning students, if they catch on to it. In fact, this suggests the initial launch was poorly timed, or else they wanted to get some early buzz to percolate a bit until the new school year.
Remember what happened to Cuil when they launched at the right time with nothing to offer the masses?
WA’s strategy might have been to get a little buzz and enough users to do one hell of an alpha test to see what they needed to do before the real launch with a complete product in the summer.
Wolfram serves a completely different purpose… I can see how Bing is trying to hit on the more Wolfram-like side of Google, but comparing Google and Bing is always silly.
It’d be more reasonable to compare Wolfram to Google Calc.
Needless to say, I’m a Wolfram fanboy. Big time. I live on this thing.
From a marketing point of view, Wolfram Alpha is a difficult name to position against the likes of Yahoo, Bing and Google.
Working behind the scenes seems to be more in their DNA. With the rise of analytics in marketing (all hail the Quants), there’s a natural path from science to business.
The trend so far suggests Bing moving towards being a broad search engine integrated with specialized verticals from other search providers and Google being the single all in one search provider.
Couldn’t have said it better. I find myself using Bing for image, travel and video searches but sticking with google for general searches from the browser searchbar. When I do upgrade my laptop this fall I’ll probably try out their cashback service with bing shopping.
Once Google rebranded Froogle into some Microsoftie(Google Product Search Beta) they let it languish.
Speaking of licensing Wolfram data, that reminds me… I applied for an API key a couple of months ago and haven’t heard back… Let’s see:
“Where’s my API Key?”
“Wolfram|Alpha isn’t sure what to do with your input.”
these guys failed to learn the lesson of Cuil. they should have kept their mouths shut until the product was good. The product, even today, is not that good.
I wasn’t under the impression Wolfram is aiming for the same crowd as Bing. If I need to calculate some complicated maths, graph something or find a fact, Wolfram is my tool of choice (a stunningly impressive tool at what it does) . If I’m looking for more general information, and my result will be in the form of a review or blog post, then I’ll google/bing it.
Don’t underestimated the powers that be you said it yourself the buzz was there but it did not quiet catch up and now it does you still not seeing the big picture it’s could be more then a jab to google ask yourself this searching for something is one thing but giving search result with a little more then information better yet a good conducted story on knowledge and facts eh. Think about that when will the madness end.
quite frankly Wolfram ALPHA SUCKS.. in so many ways.. start with the name and it just continues from there.
Too much hype, here is an idea for a story Techcrunch.
WHY DONT YOU GO BACK IN TIME AND LOOK AT THE VAPORWARE COMMENTS ARTICLES YOUR GUEST WRITERS PRODUCED, remember Nova Spivak, orgasmic enjaculation about how wolfram alpha will reshape the universe. If you re read that article all it does is expose how much of an idiot spivak is… do some soul searching… write about it how wrong they were.. and I can bet you next time someone tries to send you some PR advertising article, they gonna think about it twice.
Fact is, WA sucks, it’s unusable, and traffic is the best indicator. Huge hype curve and then it just sucked… all the way to irrelevance.
Click on search the web at your right on any query in Wolfram, it goes to Google.
I hope W|A can pull through and become relevant. There’s a lot of confusion out there about what W|A actually is. I made a little ditty to help people understand it. Check it out: http://www.nuti...a-in-a-nutshell
“Google Squared” which is projected to be a killer for wolfram alpha appears to be similar to my patent application:
Frankly, I am getting a Déjà vu effect while going through the “Google Squared” application because it appears to be very similar in function to my United States patent application which was filed on April 12, 2007 and as publicly disclosed by the United States Patent and Trademark Office on October 16, 2008, when the patent application was published.
My patent application is titled as “Method And System For Research Using Computer Based Simultaneous Comparison And Contrasting Of A Multiplicity Of Subjects Having Specific Attributes Within Specific Contexts” bearing Document Number “20080256023” and Inventor name “Nair Satheesh” which may be viewed at http://patft.uspto.gov/ upon Patent Applications: Quick Search.
Google Squared appears to be using at least some if not many of the same methods and systems as set forth by me more than two years ago in my patent application. In fact there are many more methods and systems disclosed in my patent application which I believe will help resolve certain inaccuracies found in current Google Squared application.
I have issued legal notices to Google through my Patent Attorney in the US but Google has not responded yet to any of my notices.
Ganging up on Google, eh?
I’m curious as to how long it will take Microsoft to backstab Wolfram, steal their tech, and then leave them bleeding in the gutter? They’ve got a long history of doing that to “technology partners” after all.
Bing’s image search has been giving me better results than Google’s. I’ve now made Bing my primary image search engine, Google is now number 2.
Looks good