Earlier this week, on February 23, 2010, Facebook was granted US Patent No. 7,669,123, which covers a patent on a method for dynamically providing a news feed about a user of a social network. The patent application was filed on August 11, 2006, and claims priority from a string of provisional patent applications and nonprovisional patent applications, the earliest being a provisional patent application filed on December 14, 2005.
The Background of the patent explains that social networking websites have systems for tailoring connections between various users, allowing for frequent, automatic notification of changes in the information posted by other users. Apparently as of the time the patent application was filed there were existing mechanisms that allowed a user to display information about other users, including news items, but these news items were disparate and disorganized, thereby requiring users to spend time researching a news topic by searching for, identifying, and reading individual news items that are not presented in a coherent, consolidated manner. Thus, what was needed was an automatically generated display that contains information relevant to a user about another user of a social network. Essentially, it seems the resulting “invention” was a news feed for a social network. While this may have been new to social networking sites in December 2005 through August 2006, automatically updating news feeds were hardly new even then.
If you are looking for a quick summary of what the alleged invention is I would recommend reading the first paragraph of the Detailed Description, reproduced below:
A system and method for dynamically presenting a news feed about activities of a user of a social network is provided. A user (the viewing user) of a social network may choose to view a news feed about another user (the subject user) in the social network. A list of the subject user’s activities within the social network may be drawn from various databases within the social network. The news feed is automatically generated based on the list of activities. The list of activities may be filtered, for example, according to priority settings of the viewing user and/or privacy setting of the subject user. The list of activities may be displayed as a list of news items presented in a preferred order (e.g., chronological, prioritized, alphabetical, and the like). Various news items in the news feed may include items of media content and/or links to media content illustrating the activities of the subject user. The news items may also include links enabling the viewing user to participate in the subject user’s activities. The news feed may be continuously updated by adding news items about new activities and/or removing news items about previous activities. Accordingly, the viewing user may be better able to follow the “track” of the subject user’s “footprints” through the social network, based on the news feed, without requiring the subject user to continuously post new activities.
This is shown perhaps best in Figure 4, reproduced below:
This patent is already being ridiculed by many (see for example Tech Dirt), and that is probably completely justified. It is hard to believe that the KSR obviousness requirements were meaningfully applied here, or even applied at all, which is a major problem in and of itself. Those in the industry know that some patent examiners wield KSR recklessly like a machete, claiming that most everything is obvious because it is common sense. While I vehemently disagree with the KSR standard, it is hard to rationalize why it is applied with such vigor in some areas and why it is not applied with much emphasis in other areas. It also seems, at least anecdotally, that KSR prevents independent inventors and small businesses from obtaining patents much more than it does with respect to larger, well-funded enterprises. Why is that? I mean, we can always point to patents issued to large enterprises that make you scratch your head because you know (or at least strongly believe) that had the applicant been an independent inventor or small business it would have likely been impossible (or seemingly impossible) to obtain a patent with such broad claims.
The truth is patents like this, and many other software patents no doubt, give the anti-software patent crowd all the ammunition they need to say software patents and computer related innovations should not be patentable. It is nonsense to say that they should not be patentable subject matter across the board, but when such broad patents come out that seem to cover things that are clearly common sense or would easily be appreciated within the industry there is a certain allure to the simplicity of saying ENOUGH!
I have little faith that the Supreme Court will announce the correct test in Bilski, although I do think it is extremely likely they will reverse the Federal Circuit, which is a good start. The trouble is the issue of what the test should be is not in front of the Supreme Court. It was never argued because Bilski is not a software patent or computer implemented process. It is a pure mental process, and one that I have no idea how Bilski could ever determine if anyone were infringing anyway, making the case itself of relatively little consequence, but for the enormity of the bad and over reaching decision made by the Federal Circuit.
The truth is that software patents and computer implemented processes should NOT be denied patent protection per se under some misguided belief that they do not embody patentable subject matter. Having said that, in order to obtain a patent one should be required to demonstrate that there is an innovation. Where exactly is the innovation in this Facebook patent? The fact that it has never been done before, even if true, does not mean that there is an innovation. It merely means that it hasn’t been done before — period. While it is always a shock to new inventors, the mere fact that it has not been done before does not mean your have a non-obvious invention, and it doesn’t mean you even have an invention worthy of being patented. Sure, never being done before is at least some evidence under secondary considerations, but there needs to be more.
Why has it never been done before? So often in the business method field it has never been done before because the underlying hardware and communications systems were inadequate to allow for “it”, whatever “it” is, to be done fast enough to be relevant. Hence the Priceline dutch auction patents. Dutch auctions were known for at least 400 years, and with the arrival of the Internet they could be done in real time, or at least fast enough to be relevant. So does the appreciation of the merger of a known business method with new technology developed by others result in an invention worthy of a patent? I can’t see how really, and to the extent that the KSR decision should mean anything it seems like the marrying of things done by others is little more than appreciation or discovery, not innovation.
With respect to software and computer implemented processes the test should be this… was the annunciation of the desired outcome enough to enable one of skill in the art to deliver? If it is then you don’t have an innovation, at least not unless there are some unexpected results that transform or transcend the implementation. In other words, if you merely tell a programmer that we need “this” to happen and they put pieces and parts of code together to deliver that, without having any implementation roadblocks and without the need to engage in a process of figuring out how to accomplish the task then you don’t have an invention. The invention resides in the design document, the figuring out of how to implement and deliver on a new process. Invention cannot and should not reside in being the first to implement something by merely pulling together known and readily understandable pieces of prior art.
So what is the inventive step in this Facebook patent? Who knows? Unfortunately, for reasons I have never really understood, there is no requirement to actually disclose what is believed to be inventive or unique. I realize there are perils to doing this, at least overtly, and I am not suggesting that extremely narrow and limiting disclosures are appropriate or should be required. Nevertheless, if after reading a patent you cannot ascertain what is unique then I think we have a problem, and a problem that is at the root of bad software patents.