Abstract vs. Concrete, Computers and Patents

When are inventions that are implemented via computer patentable? An answer is that when the inventions are not simply abstract and when the machine, the computer, is specialized in some sense. The case I am referring to is CLS Bank v. Alice Corp. How should judges think about the terms ABSTRACT and a specialized MACHINE? The reader might also be interested in the recent point-counterpoint discussion in The Wall Street Journal (May 13, 2013, p. R2).

ABSTRACT v. CONCRETE

First, let’s examine the term ABSTRACT. It means a simplification that is the essence of something. So it might be a mathematical equation. If something is ABSTRACT, it does not contain all the features that reflect reality. So an abstract model is not reality, it is an imperfect reflection of reality. An abstract idea is not itself a useful product, although it may form the core of a useful product . . . one for which a market exists or may reasonably exist in the future.

Let’s use the term CONCRETE to mean the opposite of ABSTRACT. So a concrete idea is one which is or could reasonably be useful. It is an idea that is fully fleshed out for its intended purpose. It is not a stripped-down simplification or essence of an idea.

Let’s say that a patentable invention must be CONCRETE.

MACHINE

Is a computer a machine that can be used to make a business process patentable? How do you make a computer into a specialized machine? When is it unreasonable to argue that computations that can be done on the computer cannot be done by hand, by pencil and paper?

I can only hope that we will lose the luddite gene that makes computers appear to some judges to be undifferentiated. Instead, computers are specialized machines that can make business processes patentable. What causes computers to become specialized? The answer is software. Software is the jig, the gizmo, or the widget of the information age.

If you break down software into its most elemental computation, a person can do the computation by hand. However, in many cases a person cannot do the sum total of the computations in a time that could be considered reasonable. Often, computer programs do computations in seconds that would have taken years or lifetimes before the age of computers. So speed has increased to such an extent that what we do is, in fact, different qualitatively and quantitatively than what we used to do.

What is the Point?

Why do we want business process software to be patentable? Why to we believe that it is necessary to view programmed computers as specialized machines in the modern age?

Primarily, we want business process software to be patentable, because it reduces the use of trade secrets. As a result, it eliminates reverse engineering and industrial espionage which are extremely inefficient activities. In addition, it promotes further development along similar lines by making the process completely transparent.

There are other reasons we might want patents in this arena. The traditional argument is that there are high costs associated with developing software and non-trivial business processes. Investments in these inventions are encouraged by providing a period of years in which competition is diminished. Competition is not eliminated; it is just diminished. It is expected that creative people will find ways to work around the patent and provide increasingly close substitutes as time passes. This is part of the genius of the patent system. It creates a structure in which people search for new solutions in markets that might not have existed without the initial patent. That is, a patent creates relatively well-defined lines, the scope of the patented invention, beyond which innovations are acceptable.

Computers are, of course, useless without software. So software is the sine qua non for computer usefulness. Continuing along this theme, useful computer/software applications are numerous and potentially diverse. In the same way, a steam engine is useless unless it is connected to something else. For example, a steam engine may drive a train, a car, or an electrical generator. Each of these applications and more deserve patents . . . in their proper time contexts.

I have purposely not addressed a number of issues such as novelty and obviousness in this post. However, I have covered these issues in previous posts. Briefly, most of the critiques of software patents relate to trivial software or obvious software. Of course, there is no excuse for patents of this nature: one click shopping, etc. On the other hand, novel and profound inventions that will have lives of many years deserve patents. For example, suppose one could really invent a way to insure against capital risk in housing or default risk in mortgages, now that would deserve a patent. One of the great errors in patent history is the lack of a patent for Rex Schrader’s combinatorial auction system which was later copied by the famous spectrum auctions.