As the recipient of the first software patent in 1968, I have been following (and written about) the software patent controversy in the United States for almost 50 years. This issue remains fiercely controversial throughout the world, including in the US. But in the US, unlike New Zealand, software-related patents, including pure software patents, have been issued by the thousands for the last 30 plus years.
The amended NZ Patent Bill distinguished what is patentable and what is not patentable.
Patentable: a washing machine which performs its task more effectively because of a program encoded onto a chip is patentable because “the actual contribution [toward better washing] lies in the way in which the washing machine works (rather than in the computer program per se)”;
Not Patentable: a method of automatically completing the legal documents necessary to register an entity, via a computer-delivered questionnaire.
Here “the hardware used is conventional. The only novel aspect is the computer program,” so the innovation is not patentable.
This article provides two examples why such an New Zealand ban on pure software inventions makes no sense because there is no difference between true inventions that are pure software and those that are not.
In November 2012 Microsoft stated in a press release “Microsoft researchers have demonstrated software that translates spoken English into spoken Chinese almost instantly, while preserving the unique cadence of the speaker’s voice — a trick that could make conversation more effective and personal.”
Certainly the way Microsoft’s researchers accomplished this complex translation is not obvious. And I have little doubt that Microsoft will try to protect its research investment with a patent application. But its patent claims could not be that it invented “speech recognition” or “language translation”, or even “voice to voice translations”. Patents for Voice and Speech Recognition patents go back well over 30 years with IBM having over 200 speech recognition patents. And there have been language translation patents for many years.
But Microsoft could invent a new way to do voice recognition and language translations and at the same time retain the cadence of the speaker’s voice. And, let’s imagine, the invention would use a standard headphone. Such a patent on that invention would be “pure software” description in its application disclosure. But what if the same invention was described as a computer chip in a special modified headphone in the patent application. Is a speech recognition/translation/cadence computer chip and special headphone invention any more innovative than a “speech recognition/translation/cadence” described as a pure software invention?
In 2000, a renowned inventor, Ray Kurzweil received a patent named Reading System which Reads Aloud from an Image Representation of a Document. The patent disclosure shows a diagram of a monitor, scanner, speakers, and a PC computer composed of a processor, storage and a keyboard. The essence of the disclosure and the invention is a logic chart describing a machine system which interacts with a speech synchroniser and the various devices.
At that time, Ray Kurzweil’s company, Kurzweil Educational Systems marketed a special purpose hardware/software system called the Kurzweil 3000 Reading Machine which was marketed to the blind and poor readers. They received a patent on this invention in 2000.
Today the company sells a “pure software” system called Kurzweil 3000 and continues to have the protection of the patent system through his original 2000 patent and with additional patents eg, Reducing processing latency in optical character recognition for portable reading machine - another pure software invention.
Few would argue that the Kurzweil 3000 Reading Machine was not an invention and not deserving of a patent.
It is also a fact that software and chips (hardware circuitry) are interchangeable. As noted in Hardware/Software Tradeoffs: A General Design Principle? the decision to put certain features in hardware and others in pure software is based on such factors as cost, speed, reliability and frequency of change.
In the US, a patent application must contain a detailed description and specification of the invention in a manner sufficiently clear and complete for the invention to be carried out by a person skilled in the art. In the US that description can be hardware and/or software.
I conclude with the following question: Should a true invention described as a computer chip (hardware) be any more patentable than the same true invention described in pure software?
*This opinion piece is part of a series looking back at major issues covered by Computerworld to mark the final print edition, published Monday July 1, 2013.