The problem with software patents isn’t that they are granted on obvious innovations, but rather that those who spend so much time complaining about them are just about completely clueless, at least with respect to patent law. It borders on the comical to observe some of the apoplectic rants against software patents, which almost universally conclusively prove that the person writing (or ranting) has not read past the title of the software patent in question. That is, of course, assuming they have even looked at the patent and are not merely mimicking what they have read from some other equally clueless and irresponsible critic.
For example, recently an article intended to point out the absurdity of software patents – Start-ups in the maze of software patents – ranted about how there is “even a patent for automatic e-mail answers…” At lease a link to the allegedly ridiculous patent was provided, so I figured I would check it out and see if the Patent Office did actually grant a patent on an automatic e-mail answering system. Guess what I found? What a shock it was to discover the gross exaggeration and oversimplification of the innovation in question. Did this guy even read the patent? Did he even understand that the claims are what defines the exclusive right?
The patent in question is U.S. Patent No. 6,411,947, titled “Automatic message interpretation and routing system.” This patent was filed on April 2, 1998 and claimed priority from a provisional patent application filed on April 4, 1997. In other words, in computer terms it is from many generations ago in terms of technology. Furthermore, in the ordinary case (i.e., almost all cases) the patent applicant never has to declare when the invention was made, so there is no way to know when the earliest date of conception was here, which would be the date that one would need to get behind with prior art that would be useful to demonstrate the inventor was not the first inventor. Similarly, with evidence of a prior user, sale or publication the critical date for the statutory bar would be April 3, 1996. So realistically to demonstrate the claims awarded are invalid one would have to come up with prior art from on or before April 3, 1996. Of course, no of this was mentioned by the ranting article, in all likelihood because none of this was known or appreciated if known.
If you look at the first claim, which is the broadest issued in this patent, it covers:
1. A method for automatically processing a non-interactive electronic message using a computer, comprising the steps of:
(a) receiving the electronic message from a source;
(b) interpreting the electronic message using a rule base and case base knowledge engine; and
(c) classifying the electronic message as at least one of (i) being able to be responded to automatically; and (ii) requiring assistance from a human operator.
Does the characterization of “a patent for automatic e-mail answers” seem to fit after reading the claim? Of course not. There is at a minimum the requirement of some kind of machine intelligence to sift e-mail messages and determine which ones can be responded to without a human operator and which ones require a human operator. A cool and quite useful invention really.
Unfortunately for the patent owner, claim 1 and a majority of the 66 claims obtained were invalidated during reexamination. However, some 22 claims survived reexamination and 25 additional claims were added to the patent as it exited from reexamination recently, on March 22, 2011 (see last page of ‘947 patent pdf). More specifically, the patentability of claims 9, 10, 14, 30-37, 50-51, 57-61 and 63-66 was confirmed. In fact, the patentability of these claims was not confirmed once, but was confirmed twice by the U.S. Patent and Trademark Office. In legal circles that means to a 99.9% certainty that those claims are patentable and will remain patentable henceforth.
So you can poke all the fun at this patent that you want, but at least two things are clear relative to the ‘947 patent: (1) the system worked to invalidate claims that were not patentable; and (2) there is still a patented innovation here that employs computer intelligence to shift through e-mails, allowing some to be responded without human intervention and triaging others and putting them in a queue for human response. So the innovation is hardly just automated e-mail responses, and despite this patent being on the the Electronic Frontier Foundation (EFF) hit-list, the patent still exists with many valid claims, many of which remain unchanged from the date it issued.
The trouble I have is that to tell the story of the ‘947 patent takes time and understanding to get it right. Truthfully, I have only scratched the surface of the story here, but have quite clearly demonstrated that the off-hand characterization of this patent as being ridiculous because it covers automated e-mail response is wholly irresponsible. It is untrue, and such demagoguery does nothing to forward the position of those who dislike software patents because those who are familiar with the law understand it to be the nonsensical, hysterical rants of those who don’t understand what they are talking about and who obviously don’t care enough to educate themselves. Simply stated, they never let facts get in the way of forming an obviously incorrect opinion.
Software patents cover the computer implementation of a process. To listen to those who are programmers talk they think that every single computer implemented process is not new or is obvious. That is, obviously, not correct and a terrible simplification of both patent law and computers, but I suspect pretty much everyone who is a computer user, save those anti-patent computer programmers, already inherently knows what I just said to be true.
If computer implemented processes are not new and are obvious then why can’t they ever work smoothly? Seriously, if computer implemented processes are not worthy of patent protection then isn’t that an admission from the anti-patent computer programmers that what they are paid to do is trivial and merely ministerial? Of course it is, yet they almost to a person treat their craft as some kind of black magic that mere mortals couldn’t ever hope to comprehend, and their projects routinely crash, burn, have bugs, need patches and simply can’t be made to work on some platforms or under certain conditions.
Either computer programmers are perpetrating one of the better scams of all time, they haven’t a clue what they are doing, or it really is harder than they admit whenever they talk about patents. So which do you think it is? Regardless, the reality that we all know to be true undercuts their protestations about the patent system and the patentability of software or computer implemented processes.
Truthfully, what makes software and other computer implemented processes patentable is the human element. Patentability and the fact that innovations are new and are nonobvious are confirmed by the very fact that things don’t go as they are predicted, that there is not cross platform operability, and that there are truly challenged users who deviate from the norm and find new and ever more creative ways to accidentally crash whole systems.
Software and computer implemented process don’t protect the code, anyone can write code, some of which works and much of which has the added feature of unintended bugs. The code is not the magic, the code is the translation of the innovation into terms capable of being processed by a machine. The innovation is the overall system from a computer engineering perspective that takes into consideration anything that can and will go wrong and addresses those possible occurrences, whether likely or not. That is why you don’t need code to receive a patent on software or a computer implemented process. What you need is a design document that would direct those who would be coding so that they are not interjecting any creativity. You want them to just code and not get creative, the vision is the inventors and the coders are the means to the end.
Listening to those who code complain about patents is nearly hysterical. They still haven’t figured out that by and large they are not innovators, but rather merely translators. Perhaps that is why they so frequently think that whatever they could have come up with themselves is hardly worthy of being patented. Maybe they are correct, but that doesn’t mean that an appropriately engineered system isn’t patentable, it just means that those who code are not nearly as likely to come up with such a system in the first place because they rarely, if ever, seem to approach a project as an engineer would. Rather, they jump right in and start coding. In the engineering world that is a recipe for disaster, and probably explains why so much software that we pay so much money for today is hardly worthy of being called a beta, much less a finished product.