Computer

Funny things about computer algorithms, #1

Human or machine: the computer says "no"

As someone who has lived and breathed computer algorithms my entire life, I have been fortunate to have had an abundance of opportunities to help ensure the protection of computer algorithms, namely methods that can be put implemented by computer software or hardware. To this day, I continue to be fascinated by various perspectives from which computer-implemented methods are analyzed and applied. I feel so privileged that I would like to share some of the wonders and curiosities of working with computer algorithms.

Computers have always played a special role in the world of patents. For example, it wasn’t until 1995, a few years after the creation of the Internet, that the United States Patent and Trademark Office (USPTO) decided to accept computer science degrees for bar admission purposes. patents.[i]. Since then, computer technology has continued to innovate, and the patent community has continued to wrestle with the implications.

Like a double-edged sword, while computer technology can induce intense admiration, it can also cause deep confusion. For a patent attorney who is also a computer scientist, there is sometimes no greater joy than helping to elucidate why and how various advances in computer technology should or should not be protected.

One way to start exploring relevant issues is to remember how the world turned out before the advent of computers and the Internet. The world was dominated by physical products, and it was usually a new drug or a new mechanical device that made people’s lives better. This concept of invention apparently stuck. Many patent offices today, including the USPTO, still argue that an invention operating on digital data may not be eligible for a patent without a claim to interact with a physical object in a tangible way, such as explained in more detail below.

Based on the traditional concept of an invention, an original concern was that certain ideas that could be easily executed by a human being were so “basic” that they should not merit patent protection. Examples include converting binary-coded decimals to pure binary[ii], which is not only simple but also involves a little more than an established relationship between numbers. However, this concern has been extended to discredit many other ideas, including those that can be done by a human in theory but not in practice.

A good illustration of how computer algorithms are viewed through the prism of patents today is Example 46 of the Patent Eligibility Guide (“Guide”)[iii]. This example concerns the monitoring of livestock behavior in response to environmental and physiological conditions. This example includes four complaints and three[iv] of them which are system claims are discussed below.

Example claim 1, which sets forth (A) a “processor” programmed to (1) “obtain animal-specific information”, which may be produced by sensors placed on or near animals, and (2 ) “monitoring” the animal – specific information “regarding the animal’s “past behavior” to determine if an animal “exhibits aberrant behavior”, is considered unreadable because the indicated monitoring is deemed so “basic” that it can be performed routinely by a human.” an additional recitation in claim 2 “connects the judicial exception to a technical field” and “substantially applies the exception”.

One consequence of including this example in the Guide is that even when the claimed process no longer seems so basic but instead involves a rigorous application of computer science principles, patent examiners cling to the preconceived idea that only digital data is produced and remain unaware of any technical or practical aspects. The fact that the assay is clearly intended and only useful in a specific application, such as the improvement of animal health, has little weight in determining whether the invention has practical application where the claim does not explicitly extend to a domain outside of computers.

Additionally, claim 4 of the example, which states (A) a “processor” programmed for (1) “obtaining animal-specific information” (but not for (2) “monitoring”) and (B) a “herd monitor” comprising a “radio frequency reader” and a “transmitter” for the collection and transmission of animal-specific information, is considered eligible. Apparently, a processor performing “basic” monitoring does not meet the eligibility threshold, unlike an ordinary herd monitor. At this point, it’s hard not to laugh at the derogatory treatment between an intangible process and a physical operation. Ultimately, all ideas originate in the human mind, but apparently when they don’t directly affect external, tangible objects, they can be considered far less sophisticated and therefore far less inventive.

For me, it’s not very exciting to work with “basic” methods, and it’s even less fun to see the ingenuity of computer algorithms get lost in misconceptions or chaos.

It is not the intention here to put the USPTO in the spotlight, and I note that the series of subject eligibility guidelines provided by the USPTO are generally more valuable than the often superficial court rulings. As more and more issues around subject matter eligibility are addressed in improving these guidelines, it’s no surprise that there’s a greater emphasis on identifying truly inventive ideas. At present, however, ill-fitting litmus tests of whether ideas can be realized by a human or other somewhat arbitrary distinctions are still adopted.

As computer technology becomes more and more an integral part of people’s lives, it becomes increasingly difficult to accurately assess the state of the art, let alone truly grasp the inventive aspects of progress. computer technology using specific and meaningful claim language. This challenge was one of the main reasons many substandard computer patents emerged. Overcoming this challenge should therefore be a primary objective for improving the quality of patent applications in technical fields. I look forward to the day when so-called “software inventions” will no longer be considered of less importance and the power of computer algorithms will be universally recognized.