How Wolfram Alpha's Copyright Claims Could Change Software 258
snydeq writes "Fatal Exception's Neil McAllister suggests that Wolfram Research's claim to copyright of results returned by the Wolfram Alpha engine could have significant ramifications for the software industry. 'While software companies routinely retain sole ownership of their software and license it to users, Wolfram Research has taken the additional step of claiming ownership of the output of the software itself,' McAllister writes, pointing out that it is 'at least theoretically possible to copyright works generated by machines.' And, under current copyright law, if any Wolfram claim to authorship of the output of its engine is upheld, by extension the same rules will apply to other information services in similar cases as well. In other words, 'If unique presentations based on software-based manipulation of mundane data are copyrightable, who retains what rights to the resulting works?'"
The key word... (Score:5, Insightful)
Compiled binaries? (Score:5, Insightful)
Modern compilers do a lot of optimization. By analogy to the Wolfram claim could compiler optimized binaries be considered subject to a copyright via the compiler? That would be bad.
They better not go there... (Score:5, Insightful)
The output of Garage Band is now copyright Apple.
The document I just wrote in Word is now copyright Microsoft.
The text message I just sent is copyright Verizon.
The photo I just took is copyright Canon.
This opens Pandora's box like you wouldn't believe. We should be restraining copyright, not expanding it.
Absurd (Score:5, Insightful)
Re:Compiled binaries? (Score:5, Insightful)
That wouldn't be a creative work.
There it goes... (Score:2, Insightful)
Re:Compiled binaries? (Score:4, Insightful)
Is there a distinction or not?
Can facts be copyrighted? (Score:3, Insightful)
Sounds a lot like the retail chains claiming copyright on information from their Black Friday fliers to keep the prices from being posted too early. Granted Wolfram Alpha is a little more complicated but if it is simply processing facts and laying them out in a certain way they might be able to patent the algorithm but the results are still facts.
I doubt it (Score:3, Insightful)
Copyright requires that a human creates something. Wolfram's software is clearly not a human, and it is unlikely to be even close to artificial intelligence either. Hence, no copyright.
This is no different than the Yellow Pages (Score:5, Insightful)
A phone book publisher doesn't own the right to your phone number, nor does it own the exclusive right to print listings of phone numbers, but it does hold copyright to the unique presentation of phone numbers. That is, you can copy the phone numbers out of their book, reformat them, print it, and sell it, but you can't just photocopy each page and the sell that.
Re:Compiled binaries? (Score:5, Insightful)
It's no more or less creative than what the Alpha software is doing. Both take an input from a human, apply some transformations to that, combine it with a library of other information, and produce something new is output.
IMO the Alpha claim is totally bogus. There was creativity in writing the software, and anything it outputs that is hard coded is possibly eligible for copyright protection (i.e., a template phrasing for an answer), but claiming each output separately is ridiculous.
Think: singularity (Score:5, Insightful)
Assume for a moment that Kurzweil is right, that people will be mergeable with machines, that your mind can be dowloaded into a neural simulator and run - recreating you, thoughts, memories, etc. All of you.
So there you are, a process running on a computer, probably in some 3D game on steroids - eternal life! But if this copyright grab stands, and the software running the simulator is copyrighted, does that mean that your very thoughts are copyrighted, too?
If you assume a materialist definition of the world, that what we see is what is, and there's no spirit, no Valhalla, no flying spaghetti monster, then we humans are, in fact, a functioning material machine.
Thought police, indeed.
Re:They better not go there... (Score:4, Insightful)
in Wolfrum Alpha's case, you contribute no original content (a search string)
What?!?!
I am crafting the search string to generate output. Unless every single search string has been pre-vetted by Wolfram, it's quite obvious that it is mine. If I vary the the search string, I get different results. That's pretty obviously "original".
If anything, the search string is the *only* part that's creative (the output is just a database.)
Prior/Other Art? (Score:2, Insightful)
Hang on... this is like the "trying to copyright pictures of paintings in the public domain." You can patent a camera, but it has been unambiguously ruled that trying to copyright a photograph of something in the public domain does not add any creative value to the painting and thus does not constitute a novel creative work. Same thing here, you can patent/copyright your bit of software, but claiming that any output it generates also constitutes a creative work by the coder of the software will not fly because the user of the software is usually the one who is doing the creative work. Maybe I'm thinking more along the lines of word processors and books where this is obvious and any goon trying to claim otherwise would be laughed out of court....
"Im sorry Dan Brown, but Bill Gates has the rights to your new book since you use MSWord2008, should have used emacs."
Also, fuck Wolfram, I was given a copy of his big fat book "New Science" or whatever, I'm not going to read it, and I can get Mathematica for free through my Uni, but I think I'll stick with my TI83 emulator since TI doesnt have a God complex.
Comment removed (Score:4, Insightful)
Here we go again (Score:4, Insightful)
The copyright of machine generated work has been a matter of law for more than a hundred years.
If you think this is in any way open to debate, ask yourself who drew Toy Story.
Re:Compiled binaries? (Score:2, Insightful)
Well by that logic, nothing is "creative", because we're just machines that are producing works based on computation made by our brains (unless you want to claim that there's something mystical about humans).
So there are two possibilities: either we say that creativity is something that can never apply to software (even if someone writes a human level AI that is indistinguishable from a creative human), or alternatively we judge "creativity" in the same way we judge humans.
US courts have already ruled that hard work is not sufficient for copyright, there needs to be some originality, so I'm not sure that a process such as compiling would ever be considered "creative", even if it was done by hand. Should what Alpha does be considered creative too? I don't know - probably not. However, the point is that even if a court ruled that it should be copyrightable, this does not imply that all other software results would be copyrightable. Just as a court ruling that a painting made by a human is copyrightable, doesn't mean that taking a photo of a public domain artwork is copyrightable.
One question that I have trouble answering: supposing a company took requests for research, then had its employees compile its results, perhaps using computers along the way, and then present them back to the person who requested it. Clearly in that case, copyright belongs by default to the company. But what if they then gradually switched to using a fully computerised method, but didn't advertise this fact? Should this suddenly mean they lose copyright? Would a court have to investigate the internal processes in order to judge whether it should be copyrightable?
There is another point by which a distinction could be made: the copyright belongs to the person controlling the software. So in my example above, the employees are entering the data to their software, and so the company owns the copyright. But a person using a compiler, or perhaps an automatic music generator, is the one inputting the data, and so should own the copyright. In both cases, the software is irrelevant. Alpha gets tricky, because it's online - they could make the argument that the software is running on servers owned by them. OTOH, there's the argument that the user is still using the software, and should own the copyright. If online software was treated differently, this would be a cause for concern, though online software already has dangers regarding licencing, in that software companies can charge for it as a service, unlike offline client-side software.
Re:The key word... (Score:5, Insightful)
Maybe they realized their position would mean the people making their compiler own their software?
Re:The key word... (Score:3, Insightful)
Personally I don't even think the "testing in court" is as far as we need to go. By even suggesting that the output is copyrighted is merely harming his own work. I refuse to use Wolfram Alpha, and have ensured my friends (many of them in the science and mathematics community) not to use it.
I thought WA was amusing but totally useless to start with, when I found out it was all copyrighted, I almost ended up on the floor in laughter.
Wolfram has no idea how monumentally retarded this whole idea is, I suppose until he gets a slap in the face by no visitors he won't learn. I don't know about others, but I see WA as a gimmick more than a real resource, there's nothing meaningful you can get from WA that you can't get from formulating a decent search or even, shock horror, reading up to date books.
Re:This is not new - NMAP is the same. (Score:5, Insightful)
I can state that the sky is green, but it don't make it so.
Yes, the NMap authors claim that a program which "Executes Nmap and parses the results" is a derivative work, but that doesn't make it so. They don't actually claim the output is copyrighted.
Re:The key word... (Score:3, Insightful)
Your argument is flawed for two reasons. First, most translations are not perfectly reversible. Second, you analogize a case where no identifiable creative content from the original work survives in the derived work to a case where much does.
If I take a DVD of The Phantom Menace and compress it to an AVI file, the compression is irreversible and done by machine. However, the AVI file is still precisely the same work -- The Phantom Menace. Why? Because nearly all of the creative content survives in the AVI file, and it is only that creative content that is copyrightable.
Compilation eliminates some creative content (such as variable names) but leaves much of it. It's roughly akin to translating a play into another language and changing the character's names.