“It is settled science that software transforms computer systems into new quantum states that are functionally and demonstrably different. Indeed, the physical transformation of computer systems by software is the only explanation for how computers are actually useful.”
Since the 2014 Supreme Court decision in Alice v. CLS Bank International, patent claims including software have faced a much higher barrier for receiving patents than any other field of invention. This has also infected specialized software, such as artificial intelligence (AI), which is both distressing and sad. It also explains why Chinese AI start-ups are receiving more funding than U.S. AI start-ups, a fact that should be sending a shockwave through Capitol Hill.
Since Alice, patent examiners have presumptively classified software claims that can be implemented on a general computer as covering nothing more than an abstract idea, which means they are ineligible subject matter under 35 U.S.C. § 101. To overcome this rejection, applicants must show why their claimed invention is something more than just a mere abstract idea. Ironically, what constitutes something more is itself an abstract idea, and even what is an abstract idea is itself an abstract idea. In something straight from out of the Monty Python version of patent eligibility, these key terms – something more and abstract idea – have not been defined by the Supreme Court or the Federal Circuit. As a result, most applications with software are routinely denied, which is understandable when frontline decision makers (i.e., patent examiners) are left without objective guidance. Subjectivity prevails.
This new subjective Alice paradigm has created an uphill battle for software claims. There have been a growing number of cases in which software patents have been upheld following Alice. See Software Patent Drafting Lessons from the Key Lighthouse Cases. But that does not change the fact that Alice is fatally flawed and ripe for reversal based on scientific evidence that was not previously considered. In short, it is settled science that software transforms computer systems into new quantum states that are functionally and demonstrably different. Indeed, the physical transformation of computer systems by software is the only explanation for how computers are actually useful.
In failing to consider these facts, Alice is inconsistent and unsound. See also Unintelligible and Unsound: Patent Eligibility in America.
Judicial Exceptions and the Misclassification of Software as an Abstract Idea
In case law surrounding patent eligibility under Section101, the courts have made a distinction between discoveries of natural laws and abstract ideas and the invention of something that is tangibly useful. For example, the famous formula e=mc2 was discovered, not invented, by Albert Einstein. It describes nature. Nuclear reactors and other devices dependent on an understanding of e=mc2 are patentable but not the formula itself.
In Alice, the Supreme Court expanded this judicial exception to presumptively include all software operating on general purpose computers. This new standard was the direct result of court challenges and amicus curiae briefs filed by large tech companies (including Google, Apple, Amazon, Microsoft, IBM, Adobe, and Hewlett-Packard). These industry leaders successfully argued that the Supreme Court should subsume into the definition of ineligible abstract ideas any algorithm implemented on a general-purpose digital computer in the absence of some other necessary inventive concept.
According to the syllogism governing previous precedents: (a) math equations describe natural laws, and (b) natural laws are ineligible for patent protection, therefore (c) math equations are ineligible for patent protection.
But in Alice the Court erred in assuming that all software algorithms are math equations. It is true that mathematicians study algorithms. As do logicians. And software engineers. But computer algorithms are rarely limited to describing natural laws or the relationships between numbers.
The key difference between mathematical formula describing the natural world (or relationships between numbers) and computer algorithm is that the former do not include logic functions such as “If…then…else.” These logic steps are not math. They do not describe existing natural relationships between numbers or things. These logic steps are a human creation which exist in computer coding precisely to provide a means to create new, human-defined relationships between the input and output devices used to interact with humans in useful ways. Such algorithms do not describe naturally occurring phenomena, nor relationships between abstract number sets—they exist purely to transform computer systems into useful ends.
Fact: All Active Software is Transformative
Prior to Alice, the Supreme Court explained in Bilski v. Kappos that satisfying the so-called Machine-or-Transformation test was an important clue to patent eligibility for a method claim. It is interesting to note that the transformation prong of that test never applied to software, although from a scientific standpoint software is transformative by its very nature.
Fact: every quantum physicist agrees that any and every transformation of matter can be defined in terms of a change in the quantum state of that system. Notably, this critically important fact was not addressed, or likely even recognized, by the Supreme Court in Alice. But the expert testimony of anyone trained in quantum physics would have established that there is simply no question among scientists about the fact that the quantum state of a generic computer is truly transformed by any change in the software and data contained in that computer’s operating memory. This is a physical change.
It necessarily follows that software, once implemented and operating in a generic computer, is truly, physically transformative, evidenced by new properties and behaviors that are both tangible and measurable. Therefore, if Diamond v. Diehr continues to be good law as the Supreme Court repeatedly says it is, an invention that transforms or reduces an article to a different state must be patent eligible. Diehr, 450 U.S. 175, 193 (1981) (“patent laws were designed to protect (e.g. transforming or reducing an article to a different state or thing), then the claim satisfies the requirements of § 101.”)
Software is as Transformative as DNA Manipulation, a Forge, or Pharmaceuticals
Arguably, the Court erred simply because the justices are not physicists and were not adequately briefed by physicists. But for those who are not familiar with quantum physics, three simple analogies are helpful.
First, when a plant seed is reprogrammed to have new traits through DNA modification, those genetically altered seeds are eligible for patents. But why is this reprogramming a seed eligible but not the reprogramming of a computer? There is no logical consistency to classifying one form of re-programming as innovative, or “something more” and the other as nothing more than merely an abstract idea that does not encompass an innovative concept.
Second, the human body has many natural states of existence: hungry, full, sleepy, alert, anxious, depressed, and more. These states of being can be altered by ingesting sugars, fats, carbohydrates, sedatives, antidepressants, poisons and more. These different ingested compounds are analogous to different programs being inputted into the working memory of a computer. As long as these ingested compounds are active in the body, the body must react to them. These inputs necessarily transform the state of the body. Similarly, a computer that is running a program is also transformed into an altered state of being as long as that program is present in its memory.
Third, properly understood computers are really just an elaborate form of raw material, like a few pounds of copper. Given tools, time, and energy that copper can be repeatedly transformed from a pile of screws, into gears, conduit, electrical wire or any number of useful inventions. It can even be recycled into one invention, then another. The only difference between transforming copper into different devices and transforming a generic computer into a clock, a gaming system, or a spreadsheet workstation, is that it takes less time and energy for each transformation of the computer (at least, after the first thousand hours writing the software). The transformations of both copper and computers always requires tools, time and energy, but the advantage of computers is that of efficiency, requiring less energy, time, and tools between each transformation than copper.
When the programs that transform a computer into a calculator, or a clock, or a video game, or a spreadsheet, or messaging system are described in a flow chart or printout of code, that description of the code is an abstraction. But when that program is read into and actually “overlaid” into the operating memory of a computer system, each program step truly transforms the quantum state of the computer system and purposefully reconfigures the flow of information and energy into patterns that are meaningfully and intentionally designed to produce useful results.
Most notably, all of these changes are tangible and measurable. In contrast to the realm of abstract ideas that can accommodate an infinite amount of information, in a real computer system the program steps and related data face true physical limitations. Each bit occupies physical space in the form of transistors that must be held in specific electronic states as defined and controlled by the software overlay. The space occupied is measured in nanometers, but it is truly a physical space. It is not abstract. Moreover, each space in this carefully controlled physical system must be correctly transformed since a single physical error can trigger a “blue screen of death.”
Application to the Subject Matter Eligibility Test
The stumbling block for many applicants is found in Step 2A of the Alice/Mayo framework, which asks: “Is the claim directed to a law of nature, a natural phenomenon (product of nature), or an abstract idea?” Typically, examiners reflexively reject software patents under Step 2A as being directed to an abstract idea and define the abstract idea so broadly that the analysis made under Step 2A infects any analysis under 2B (i.e., the so-called hunt for the inventive concept). It is simply wrong to describe a software overlay that is in active computer program memory as non-transformative, much less, nothing more than an abstract idea. In fact, software truly changes a computer system in significant, measurable, and useful ways. That alone should make it impossible for any properly described software to be an abstract idea.
The Patent Office, Federal Circuit and Supreme Court should all thoughtfully consider the science and technology involved. There has been great mischief created by Alice, and all because no one bothered to consider what is actually occurring on a physical level.
Perhaps the software patent claims being considered by decision makers do not describe a novel innovation, or perhaps what is described would be obvious to one of skill in the art. Further still, it is certainly possible that the patent claims reviewed are not adequately supported by a written specification that describes with appropriate detail the fullness of the innovation in a way that teaches those of skill in the art. But what we know for certain is that the software, when introduced to a computer—even a general purpose computer—will make physical changes that are perceptible.
Tangible physical changes and rearrangement or an article into a different state or different thing has always been the hallmark of a patent eligible invention. The fact that the alteration in state is something that cannot be seen or isn’t appreciated by the masses should not justify a per se rule that innovation of a certain category is simply not eligible for patent protection. As Judge Pauline Newman has explained repeatedly in her opinions, § 101 is intended to be a low threshold so that it does not prevent innovation from flourishing before it is known what may result from a particular field of endeavor.
It is recommended that these arguments (and if practical, the expert opinion of physicists regarding the change in computers’ quantum states due to active software) should be presented in every prosecution of software claims, until such time as USPTO guidance acknowledges this fact.
Image Source: Deposit Photos
Join the Discussion
12 comments so far.
chriswMarch 29, 2019 07:19 am
Isn’t the theft of one’s software-implemented invention by others evidence that the invention as a whole cannot possibly be a mere abstract idea ? Perhaps some mileage can be had from considering what exactly about your invention, another person can steal. i.e., thieves cannot steal mere ideas. Theft involves re-venuing some tangible property of value. No person in any market would ever pay to acquire an abstract idea. So, use it as a test, if what you are claiming is something that can be sold, it can’t possibly be abstract. A computer having your code embedded in memory which is a machine capable of a new and non-obvious, useful function, should hands-down qualify. Otherwise, the only means for restoring the system appears to be a boycott of every product and service offered by the entities that lobbied to put AIA and the PTAB in place. They are waaayyyy too powerful, where’s the anti-trust regulators ?? On the take also. Quitting the system and letting it implode might not sound like an option, but the sooner it does, the sooner it can be rebuilt.
anonyMarch 25, 2019 12:01 pm
Great article and great fodder for an in depth practical example. E.g., which of the following claims should be patentable under 101 and why? (See suggested answers below; assume E = m * c * c is novel and nonobvious).
1. The formula E = m * c * c.
2. An apparatus comprising a processor configured for: receiving a mass value (m) and a speed of light value (c); and calculating an energy value (E).
3. The apparatus of claim 2, further configured for: calculating the energy value (E) using the formula E = m * c * c.
4. The apparatus of claim 2, further configured for: calculating the energy value (E) by multiplying the mass value (m) by a square of the speed of light value (c).
5. The apparatus of claim 2 [[or 3 or 4]], further configured for: storing the calculated the energy value (E) in a register of the processor.
6. The apparatus of claim 5, further configured for: displaying the calculated energy value (E) on a display to a user.
7. The apparatus of claim 2 [[or of claims 3 through 6]], further configured for: calculating a length of a bar of uranium, wherein energy stored in the bar of uranium after being cut is equal to the energy value (E).
8. The apparatus of claim 7, further configured for: displaying the length of the bar of uranium after being cut.
9. The apparatus of claim 7 [[or 8]], further comprising: a cutter; and further configured for: cutting the bar of uranium to the length determined from the energy value (E).
One can argue (1) that claim 1 is not patentable because it is not to a machine or improvement thereof and (2) that claims 2 through 9 “should” be patentable because the machine (the claimed apparatus) is improved by performing new functionality (i.e., a machine configured to perform E = m * c * c is improved over an otherwise equivalent machine that is not configured to perform E = m * c * c).
Based on Alice, all 9 claims would be rejected because (1) claim 1 is not to a machine, (2) claims 2 through 8 are draftsman’s tricks that merely add functions of generic computers, and (3) claim 9 is a draftsman’s trick to try and make claim 8 patentable and cutting uranium is old and well-known. (The argument against claim 9 may be the weakest, but the Court has thrown out the machine or transformation test, and this example shows how Alice can reject any claim under the sun).
Joachim MartilloMarch 23, 2019 12:12 pm
[email protected] 22, 2019 10:31 am
The book entitled The Employed Inventor in the United States by Fredrik Neumeyer and John C. Stedman is supposed to survey the common law of invention. I have never read it.
BMarch 22, 2019 12:31 pm
@ non-sequitur “The issue is the abstractness of software claims.”
Actually, this is what Investpic and Electric Power Group hold. Taranto’s war on patents is front and center at the SCOTUS now
BMarch 22, 2019 10:31 am
There is now a record number of s101 cases b/f the SCOTUS. All make EXCELLENT points representative of the arguments made above. Berkheimer, Investpic, Kamrani, Villena (https://www.supremecourt.gov/Search.aspx?FileName=/docket/docketfiles/html/public\18-1223.html), and Vanda (all cert.) and soon Bhagat (mandamus).
One question I wanted to ask but didn’t: Is “invention” a technical term or a legal term? I mean, how does one determine if a new process or other thingy has an “inventive concept?” Judges seem to just KNOW whether a device is inventive or not, and typically they’re the most technically-ignorant people on the planet.
So is “invention” a legal conclusion with factual underpinnings? No, Berkheimer is already dead-letter jurisprudence. Things like “evidence” are just idiotic in the view of the CAFC.
So is “invention” a pure legal term? If so, how is it determined? Certainly not based on the claims as a whole, ordered combination. Villena proves this. Certainly not based on preemption. Berkheimer, Investpic, Villena, Kamrani et al. prove that.
“Invention” is a term morons in black robes (as well as the EFF) use in order to pretend they know of what they speak.
TernaryMarch 22, 2019 12:19 am
There is no need to use quantum states to explain that a computer running a program is a physical apparatus. An oscilloscope or even a voltmeter is sufficient to demonstrate the variations in physical states of the circuits.
While the Supreme Court declares computer implemented inventions such as in cryptography as being directed to an abstract idea, hackers use the physical properties of running computers in “side channel attacks” to hack the alleged “abstract ideas” to access some of our most valuable data.
AnonMarch 21, 2019 10:56 pm
Sorry NSII, but such Pollyannaism ignores the reality (and in no small part, one but need look at the massive amount of propaganda against software to realize that a “it’s the claim, not the innovation” is just B$.
Non Sequitur IIMarch 21, 2019 08:31 pm
The issue is not the abstractness of software. The issue is the abstractness of software claims. Unless the things that make software tangible are claim limitations, the quantum state of a particular implementation is beside the point.
Joachim MartilloMarch 21, 2019 03:17 pm
I think I made the same point in my response to the USPTO’s request for comments on the new 2019 § 101 guidance.
If you are interested, just start reading on page 16.
I am such an idiot. Until two months ago I thought that all patent practitioners, hardware engineers, and software engineers viewed programmable devices and program-executing devices (processors) as I do.
BTW, I agree with Anon about In re Nuitjen.
Maybe, I don’t understand the decision, but it says to me the following.
If I can make a soliton projectile, I can claim the method for creating it as well as the device that shoots it. Yet I can’t claim the soliton projectile even though it has a well-defined structure.
EGMarch 21, 2019 02:54 pm
Software is equivalent to firmware which is equivalent to hardware in terms of its functionality in computers. SCOTUS’ Neanderthal decision in Alice is on par with believing that the Sun revolves around the Earth or that our Moon is made of green cheese.
AnonMarch 21, 2019 02:25 pm
Another case that deserves to be over-turned on the pure reality of the matter:
In re Nuitjen
Pro SayMarch 21, 2019 01:48 pm
The year is 2030.
As a result of misguided court decisions including but not limited to Mayo, Alice, and eBay; in combination with Congressional inaction; American companies, research institutions, and independent inventors have been unable to obtain constitutionally guaranteed patent protection for their inventions for more than a quarter century.
At the same time, countries including China, European Union members, and others have greatly strengthened their own patent protections.
As a result, American companies continue to receive stronger patent protection for their cutting edge innovations in foreign lands than they do in their own home country. While at the same time being unable to stop foreign companies from legally stealing these same, unprotectable-in-the-US innovations of American companies, research institutions, and independent inventors.
Including due to the dearth of new, patent-based company formation in America, China is now the undisputed leader in myriad cutting edge innovations including but not limited to artificial intelligence, machine learning, software, and medical diagnostics.
Indeed, many of today’s most important, life-saving diagnostic technologies come stamped, “Made in China.”
Foreign media regularly report on apparently rampant intellectual property theft by American companies; possibly with the tacit if not actual support of and at least partial funding by the U.S. Government; from these new, foreign leaders in patent protection.