Wolfram Alpha and its architecture of failure

Wolfram Alpha wants to lay claim to the output of its searches, which may ensure that virtually no one bothers to use the service at all.

Please see the response from Wolfram|Alpha at the bottom of this post.

One thing has become clear: to succeed on the Web and in the next generation of software, you need to invite, not dissuade, outside participation. Tim O'Reilly calls it an "architecture of participation," but whatever you call it, the best software strategies are those that encourage outside contributions, rather than discourage it.

This makes Wolfram Alpha's terms of service mind-boggingly backward at best, and troubling at worst. Some have pointed to the quasi-search engine's sometimes weird results as a reason to give the service a pass, but there's a far more fundamental reason to reject Wolfram Alpha , as Groklaw suggests.

Wolfram Alpha demands citation when using the results of its "searches," which is a distinct departure from Google's "use pretty much as you please" attitude, and will almost certainly curb the appeal of Wolfram Alpha, no matter how good its output becomes. Groklaw writes:

Wolfram's Terms of Use are not at all what I would expect from a search engine, probably because that isn't exactly what Wolfram Alpha is providing. It's a computational service, at least in some cases providing computational output from various sources of data that perhaps never existed until you asked your question. So, they claim copyright on the results and require attribution. That's fine with me, so long as the information provided really is uniquely theirs and not just the answer to what is meaning of life and everything, but it is different from what I'm used to from Google and other search engines, so it is counter-intuitive, something to be aware of before I include Wolfram Alpha output in a presentation on Groklaw or in a book.

In other words, Wolfram Alpha requires: "If you make results from Wolfram Alpha available to anyone else, or incorporate those results into your own documents or presentations, you must include attribution indicating that the results and/or the presentation of the results came from Wolfram Alpha." It's a fair request, but it may not be a reasonable request. Not if Wolfram Alpha wants people to actually make widespread use of the service.

Groklaw concludes that this requirement "means Wolfram Alpha will never replace Google," which is absolutely correct. Even if Wolfram Alpha delivers better "search" results, the burden of figuring out and delivering proper citation is going to keep people using Google, which doesn't make the same fetish of proclaiming its ownership of search results.

Wolfram Alpha may well prove to be the best computational search engine on the planet. But until it learns to lose the heavy hand of enforced citation, it's going to struggle to become a first-choice search tool.

UPDATE @ 12:48 PT. Wolfram|Alpha's Theodore Gray contacted me with the following commentary on my post above.

Hi, I'm the person who wrote most of the language in the Wolfram|Alpha terms of use, and I want to correct a couple of things you got from Groklaw.... There are two basic confusions they have perpetrated: First, Google does copyright its results pages just like we do in all comparable cases (e.g. images.google.com, news.google.com, etc).

And second, Wolfram|Alpha is not a search engine: We don't return "search results", we generate original content including plots and graphs. A more comparable situation would be Google Maps, which Google claims copyright on just like any other map provider. Similar, look at pretty much any site that generated financial trend charts, weather charts, you name it, they are always copyrighted. We are not doing anything more aggressive or grabby than Google or other widely used websites: Groklaw simply got it wrong.

I would be happy to answer any further questions you might have, but please, if you're going to criticize my beautiful terms of use language, please at least read it carefully first, and understand that it's not talking about search results. I'm happy to entertain criticism, but not if the starting point is that we're more restrictive than Google, because that's simply not the case.

My response? This misses the point of my post. As I wrote back to Theodore:

Thanks, Theodore. I quoted your terms of use directly - not sure how I can do more than that? Reading Groklaw's take and your note below, I stand by what I said. I absolutely think you have the *right* to require attribution, and I understand the reasons for doing so....

My argument is that by insisting on this, or at least not making it brain-dead easy for your users, you're going to a) make it difficult to enforce because you'll spend all your time chasing infringers and b) induce people to try to use an imperfect replacement for your service (like Google or other) because citation becomes too cumbersome.

You can resolve a and b by simply making citation automated in some way. I would be absolutely for that: I'd love to use results from Wolfram|Alpha in my work, but I doubt I'm going to want to chase down the citation every time. I don't think you want to burden "distribution" of your results: I think you therefore need to find a way to make it easy for users to show that the results came from your service, without making them do all the legwork. I'm not smart enough to know how to do that, but I think you/your team probably are/is.

The point, in short, is not whether Wolfram|Alpha has the right to do this, but simply that doing so may negatively impact adoption of its service. It's similar to copyright assignment in open source: David Neary recently made the compelling argument that copyright assignation can hurt community adoption of open-source software, but that companies may need to do it, anyway.

It's a trade-off, but in Wolfram|Alpha's case, I believe there are ways for it to make this trade-off less burdensome on users, thereby inviting participation and not unduly hampering it.


Follow me on Twitter @mjasay.

Tags:
Tech Culture
About the author

    Matt Asay is chief operating officer at Canonical, the company behind the Ubuntu Linux operating system. Prior to Canonical, Matt was general manager of the Americas division and vice president of business development at Alfresco, an open-source applications company. Matt brings a decade of in-the-trenches open-source business and legal experience to The Open Road, with an emphasis on emerging open-source business strategies and opportunities. He is a member of the CNET Blog Network and is not an employee of CNET. You can follow Matt on Twitter @mjasay.

     

    Join the discussion

    Conversation powered by Livefyre

    Don't Miss
    Hot Products
    Trending on CNET

    HOT ON CNET

    Want affordable gadgets for your student?

    Everyday finds that will make students' lives easier: chargers, cables, headphones, and even a bona fide gadget or two!