Comments on: 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.

[wperry1]

Call it what you will... a "computational service", a search engine or a widget. If Wolfram Alpha is used to find information on other websites it will have a very difficult time laying copyright claim to the results.

[pentest]

You obviously don't know what wolfram alpha is.

[waynerifish]

I don't think Wolfram Alpha was ever intended to replace or even directly compete with Google. They are totally different services. The Terms of Service are different because the service is different.

[sdf0013]

I've wondered this myself. Like you, I see them as very different services. I can understand their desire, especially from a marketings perspective, the desire to be sited as the source. But, certainly there cannot be a copyright issue since the results are not their own - merely regurgitated. Also, if anyone were to use these results in a presentation, you need to quote the source, not the delivery -- meaning if it came from government survey data your audience needs to know that rather than you found it on Google or Wolfram, etc.

Still. I do like the concept of what they are trying to do. It's very Star Trek-ish in simply asking the computer a question and having the answer pop-out. If you think about it, as we continue to gather more and more information, data, etc. finding it and being able to produce an answer is going to get harder and harder. We simply have to have better tools to be able to do this. We've come a very long way from the card catalog and the encyclopedia. Our research tools have to be able to keep up with the data that we're able to collect and maintain.

[aMUSICsite]

It's only version 1 and it's not supposed to be a Google killer, Wiki would be a better comparison anyway. As it's trying to give answers to answerable questions without the fear the some kid might have just changed the whole article (Wiki) or getting noting but sales pitch/latest new stories (Google).

I think it will grow over time and be a valuable tool like Wiki, but it's not going to replace either Google or Wiki it will be something in between that get's used when you are after the kind of facts it turns up well.

[divisionbyzero]

Uh, Wolfram Alpha could just auto-generate a proper citation when it does a computation that results in a unique result rather than a collation of disparate results. It should be pretty easy.

[Amyaz]

it's actually fairly stupid to post this type of information without a citation. if, say, i post info based on a calculation in a report, and you later want to know "what the number really means," you'll waste quite a bit of time if i haven't noted the original source....

[vamman]

Its a search engine designed to keep idiots out.

I have no issue with citing them in a report.

What the hell is your problem CNET?

[J. Blow]

How do you know your not considered an idiot? Pompus at best.

[Hunnter2k3]

The instant the first person gets in trouble for not citing them, the instant it dies.

[sroussey]

Are you serious? This is the "failure"? They create, essentially, a research report for a someone that wants to claim it as their own, and can't because they have to cite it? Sheesh...

Gosh, soon sites like this will put something into their TOS that forbids site scrapers from copying articles without license (or citation!)... What WILL the world come to...

[orlandorr]

What is really terrible about Wolfram Alpha is it's NAME.

Too long and strange. They've basically alienated 90% of the world audience with it.

[Maximus_000]

Are you serious? Means if I am give input 2+2 then result 4 is copyright property?
I am going to get copy rights of 10 so every time they display 10 they should attribute it to me.....
Lolx Great Idea Wolfram.... I assume you were sleeping when God was distributing common sense....

[Magallanes]

The main trouble with Wolfram Alpha is the lack of market study. Wolfram Alpha is pointed for some academic and scientific level of user, but those people have better tools to calculate a single 1+1 or to do a search of Yahoo Finance to check about some share.

[ueno54]

You say:
"the burden of figuring out and delivering proper citation is going to keep people using Google"

But it's pretty clear from the terms of use:
"Whenever possible, such attribution should take the form of a link to Wolfram|Alpha"

You're complaining about linking.

[albertsoler]

"Terms of Use" issues seem to be popping up all over these days. I never bothered to read the terms for WolframAlpha. Apparently, only journalists and lawyers are the only ones who do anyway. Copyrighting the result does seem to be a strange concept regardless on how it was determined.

I wanted to know which element had the highest melting point and was impressed on the speed and representation of the result. But, do they now have a copyright on the answer (which was carbon at 6422 degrees)? Is every single article, web page, periodical, text book, science report/term paper, (*or even this post?), etc., with this bit of information suddenly infringing on this copyright? Very strange indeed. But, I don't think so. Citing the source of information is not unreasonable and a very old practice; except for plagiarists.

Still, we now have one more complication to add to old, out-dated, copyright, trademark, and patent laws.

*If this post does get DMCAed by WolframAlpha -- please let us all know. It would be high-larious!

[vikinzer]

I've taken a look at Wolfram Alpha after reading this and I have to say I don't see a lot of legitimacy in your argument. Here is why. The service bears little to no resemblance to Google. It is not a search engine of other people's content. It is a search engine of a database of information they hold. You cannot copyright information. So if you input 2+2 and get 4 legally they can't actually ask you to cite that. If however, you get a graph, or a chart, or some other presentation of data that the site generated, then they do hold copyright to that content. It isn't a link to external content someone else created, it's unique content.

Asking for credit isn't really a big deal. This is basically the same as an electronic searchable encyclopedia that makes use of modern dynamic generation technologies. While I think that their legal ability to lay claim to say the population of Bloomington, IN (my hometown) is questionable, I also don't think they are going to prosecute anyone who just lists the number.

I also think this mentality of citation comes from an academic culture. Someone who would invest in a research resource like this likely comes from an academic background. At the very least even if the founders aren't academics they involved a lot of academics in collecting the information used and deciding what was worth including. The net is a place where people read and accept, a lot of stupid rumors have circulated as a result of that aspect of net culture. I personally have no problem with an organization that wants to promote a culture of citation and making sure your facts are indeed correct. It lacks the instant gratification of net culture, and takes a little more time out of our day, but maybe we should all slow down and realize that might not be such a bad thing.

[ikjadoon]

EBSCOHost has a fantastic citation system.

If you find an article, you basically click the button and a smooth citation dialog comes up with all the available types of citation are able to be copied/pasted. It has MLA, APA, AMA, Chicago, Vancouver,etc.

It works fantastically well. :D

[jmfb_k7]

Wolfram Alpha is not a search engine and has never claimed to be one. Only CNET and Wired call it a search engine (because they didnt read the FAQ's posted on the site. They just saw the interface and made an assumption).
Wolfram Alpha does not have aspirations to replace Google. But it would be wise for Google to licence the service (providing that they are willing to properly citate the results).

[monkeyfun14]

"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."

How does anyone perceive its trying to compete with Google in the first place? They are totally different search engines...

[Sir_Sid]

Your comparison is very flawed. Wolfram Alpha should have a different policy because it takes commonly known data and puts it in a presentable form. If you use that data you should cite it. What is so hard about citation anyway

[chaseneb]

How hard can it be to add a auto-cite linK? Wikipedia already does it for you. Wikipedia even lets you choose the formatting style you'd like to use for your citation. It can't be that hard. The only concern would be that depending on your search parameters the search results would change for the person who checked the citation. If you search "here" or "today" you get results based off of where or when you search. Ideally the autocitation would lead you back to the EXACT results that are cited instead of a link that would change daily.

[mnemopolous]

Would the Wolfram Alpha terms of service make more intuitive sense if the service were offered on a subscription-based (rather than ad-based) revenue model?

I haven't played with Wolfram Alpha too much; does it rely on crowdsourcing (like Google or Wikipedia) for its data? Does Wolfram Alpha cite its sources? Google does (and that's really the point of the thing)...