The idea of the "utility patent", as embodied in the US Constitution, article I clause 8, is simple:
The Congress shall have Power ... To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries;(An extensive commentary on this can be found here.)
The basic idea is this: In order to promote technical innovation, we want inventors to publicly describe how their inventions work, so that others can learn from them. In order to entice them to do that, we will give them in return the exclusive right to manufacture the new device for a limited period, originally set at 20 years, but currently set at 17 years. This worked well back in the days when dissemination in printed form would take a few years, but the useful life of an invention could be close to fifty. In the meantime, however, the system has somewhat fallen apart.
A century ago, most patents dealt with mechanical machinery. Once you had an example of the machine in your possession, it would be fairly clear to see how it worked, and there was a strong incentive to get the protection of the patent. As engineering progressed from mechanical to chemical, a very large number of patents described manufacturing processes, which would not be possible to discover from examination of a manufactured sample. For these inventions, a viable alternative to patents would be trade secrets: If you don't tell anyone how it is done, they cannot match your manufacturing prowess unless they can outsmart you by independently coming up with the same invention (or an even better one). At this point, it is unclear which is the better path: To patent (and publish, and have protection for a limited time) or to keep the invention secret (and possibly have protection for a longer period before the competition catches up). When you factor into this equation that if you apply for a patent, it might be denied, and that the judicial system affords some degree of protection against disclosure of the secret by malicious employees, it might indeed be better for the corporate inventor not to disclose the invention through the patent process.
To help re-balance the equation in favor of the patent process, Congress legislated an improved patent process, where the invention is not published until after the application has been reviewed, allowing the inventor to maintain the trade secret protection if the patent is denied. Again, this was not unreasonable if the evaluation process was short, and the lifetime of the patent was shorter than the useful life of the invention.
Fast forward to our time. Today, a patent examination may take 2-5 years, while the useful life of many inventions is 2-10 years. Add also, that the various branches of technology have become so complex that it is very hard to examine whether an invention is new; this is why it takes so long to examine, but it is still not uncommon to discover that the patent office has granted a patent on something that was well known in the industry to which it applies, long before the claimed date of "invention". This means that even after a patent is issued, nobody really knows if it is valid, until it has been tested in court.
Given these absurdities, it is entirely possible that as I go about my daily business, I will develop something, about which I will assume that since I am not a genius, the very fact that I came up with it is proof that it is something which is "obvious to a persion skilled in the technical field". And it it so, it is according to the law not patentable. But if someone else believed it to be not obvious, they could have invented it at the same time, and filed a patent application, which is now being reviewed and may be granted in two years, about the same time I bring a product to market, based on my own work, with no knowledge of the work of that other person. In fact, they may decide not to pursue any work of their own to exploit their invention, but pounce on me exactly when they discover that my work is profitable. At that point, their patent surfaces like a submarine attacking a cargo ship ... and in fact this tactic is known as a submarine patent.
Large companies defend themselves against this by filing a bunch of marginal apllications of their own; then if a competitor tries to spring a submarine patent on them, they will fight back by pulling out a stack of their own patents and essentially say: "We may be infringing on your patent - which by the way we believe is invalid - but you are also infringing on several of ours. Why don't we agree not to sue each other over this?" That works for IBM and Motorola, but not for the smaller company without deep pockets.
Here are some patents with some comments on each:
I found this patent number on a bag of water softener salt I had bought at the grocery store. This is the kind of mechanical invention that the system handles very well. It is an incremental improvement over prior art, and already 4 more patents have been issued that acknowledge the prior contribution of this one.
I came across this patent in the course of my work, in preparing for a business meeting with the inventor. The application for this patent was filed in 1998, but people that I work with claim to have thought of the idea several years earlier, but not found it worthy of a patent filing because it was obvious: If you can encode a signal in spread spectrum using a "pseudo-noise spreading code" in such a way that the signal can co-exist with other signals in the same channel, it is obvious that you can transfer more data by having the same radio equipment encode two streams of data in parallel. If you use two entirely separate sets of radio equipment, programmed with different spreading codes, you have done nothing new, just used spread spectrum according to its definition. The optimization to mix the two signals at the antenna is trivial. Further optimization by mixing at the IF (Intermediate Frequency) stage is "obvious to a killed person in the field". Performing the mixing at the digital signal processor is "obvious in principle but less obvious in practice". Although the patent contains 40 claims, it appears to me that the essence of the patent is completely summarized in the first claim, and is the idea of increasing the data transfer rate of a spread spectrum radio by transferring two (or more) signals in parallel. The detail in the subsequent claims adds nothing new but just describes how a (direct sequence) spread spectum radio transceiver works in the first place.
The interesting twist is that it may well be in the business interest of the people I work with not to assert their claim of prior art in public, since that would disturb any cooperative ventures between the two companies.
Wearing a kneepad on the right knee allows a right-handed tennis player to return really low balls by diving down on that right knee without fear of injury.
In a way, this is harmless to issue a patent for, but what is the point? Could the inventor win a tennis tournament by using this stroke and denying his opponents the right to do the same? Would that be fair play, and how long would it take the tennis federation to change to rules to outlaw this play if he did? Note that the patent does not cover the kneepad, just the specific use of it, so the inventor cannot make money by selling the kneepads.
Bruce Dickens managed to get a patent issued for something that had been in common use for many years before he filed a patent application, but had never been patented because everyone in the field considered it obvious. Fortunately, has now been re-examined.
Thomas Higley was issued two separate patents on sending email that include a link to a web-page. His application points to descriptions of how HTML works, including Netscape's docuemntation outlining how HTML messages can be displayed in the Netscape Mail client. To me, the invention is outlined in this documentation. I believe these patents (which were filed long after the web was in common use) should never have been issued.
This patent says that a certain segment of chromosome 2 of the human genome varies a lot between individuals (like fingerprints) and this can be used to match tissue samples to determine if they are from the same person. Since it varies in a gradual manner, it can also be used to determine how closely related two people are, and how likely it is that an organ transplant will succeed.
I think this fulfills all the criteria to be a useful invention. A remarkable twist (common to many patents) is that this came out of research paid for by the US Federal Goverment, and so it may in fact be in the public domain. In any case, the owner of the patent - the state of Oregon - decided to let it lapse by not paying the required renewal fee.
The patent says "we read a sequence of nucleotides off a DNA sequencing machine, and if anyone figures out what to do with it, it is ours". It does not say that the "inventors" understand what function this section of DNA performs in the genome, nor does it reveal any ideas for how this knowledge could be used in medical diagnosis or treatment. I do not think there is an invention here - at all.
This is a disgusting abuse of the patent system.
Amazon claims to have invented - and be entitled to a monopoly on - the idea of paying a commission (percentage of sale) to those who refer a customer to a sales shop, when this referral is done by a hyperlink on the World-Wide Web.
Paying commissions on referrals is not new; doing it when the referral and the sale are on the Internet would seem obvious to those familiar with both retail sales practices and with use of the Internet.
This has been the subject of much anger in the industry, and most professionals working in the field of electronic commerce agree that it should be struck down. It is written up in 26 claims and made to sound complicated, when the whole point of it really is that it is very simple.
It seems to me that the "shopping cart" setup for online shopping was original and could have been patented, but the idea of not going through a shopping cart is much less original.
Amazon should pray that this patent (Integrated interface for vendor/product oriented internet websites) does not hold up, since it appears to claim a patent on the idea of online shopping in the first place.
As you can see, I am not against patents per se, nor even against software patents. What I am against are three things:
My signature is on an application that is currently under evaluation, so obviously I believe there is some value in the system.
In the old days, the printing press was an expensive piece of machinery, around which an entire business was built. Today, the photocopier and the personal computer allows anyone to become a publisher and a printer. We should not be surprised that laws that used to successfully regulate the interactions between create artists (including writers), publishers, printers and libraries are straining under this new reality.
Copyright laws attempt to balance the rights of writers (and artists), publishers and the public. Unfortunately, it is mostly the publishers that have been bending the ears of the legislators with suggestions on how to amend the laws. Thus, these revisions have unfairly shifted the balance in favor of the publishers. This should be stopped.
Part of the bargain of copyright included: