In the recent spate of abstract idea cases, the identified abstract ideas have been concepts that are old and well known such that the court can easily take judicial notice of their antiquity. The courts particularly point to the long heritage of an idea as evidence of its abstractness. However, that antiquity-limitation does not appear to be a requirement. Rather, it seems correct to say that a “newly discovered” abstract idea is also an ineligible abstract idea. My question here then is what importance or relevance should be given in the abstract-idea analysis as to whether or not the central idea is old or new? Or, is my assumption wrong and abstract ideas (in this context) are only those that are old?
To answer your question: The inability of the courts to think straight and write well results in the conflation of obviousness and ineligibility. If it is an abstract idea, a court should state that it is not subject-matter-eligible because it is an abstract idea. If it is also obvious, like implementing an old business method with the improvement of doing it on a computer, the courts should say that it is also obvious. Instead, the courts say something like “all they did was implement an old idea on a computer” so it is subject matter ineligible. From this, in our system where rule evolve from prior case law, we could see case law in which ineligibility requires lack of novelty, in addition to abstraction. That is a rule you might distill from the case law, and would make your assumption wrong.
DC, just a small point, but I do not believe that the Supreme Court is using the term “abstract” the way that term was developed in the case law, particularly in Morse and Rubber Tip Pencil – an idea without more, a disembodied concept. Many of these claims are quite detailed in the method and means for carrying them out. Virtually no one would consider them abstract in the sense announced in the original Supreme Court cases.
What I think the Supreme Court is doing is using the term abstract almost as a nonce word to describe what they really mean – that the subject matter of the claim that is new, the real invention being claimed, is not statutory. Thus a claim could include statutory subject matter such as machines like computers, articles a manufacture like index cards, or compositions of matter, but that subject matter is entirely old and conventional and rather serves to dress up the claim in order to nominally bring it within the statutory classes.
Thus I would tell you that you can reliably predict that if the novel subject matter in the claim is merely using things like computers or manufacturers or compositions of matter rather than making them, or improving them (new or improved within the meaning of section 101), then the claim effectively is claiming nonstatutory subject matter and the Supreme Court will never permit patents to extend so far.
Thus when we are talking about nonstatutory subject matter, the issue of whether it is new or old is completely irrelevant. In contrast, the issue always will be whether the statutory subject matter is new or improved – exactly what the statute requires.
Thus I would tell you that you can reliably predict that if the novel subject matter in the claim is merely using things like computers or manufacturers or compositions of matter rather than making them, or improving them (new or improved within the meaning of section 101), then the claim effectively is claiming nonstatutory subject matter and the Supreme Court will never permit patents to extend so far.
You just said that “merely using” something isn’t patent eligible, but 35 USC 101 includes “process” as one of the four statutory classes. Does the distinction between an “abstract” process and a 101-eligible one stem from whether the process produces something that is itself statutory? If not, can you address what a statutory process looks like?
Any thoughts on whether a patent directed to “Software lets parents see what their kids will look like in 50 years” would be patent eligible.
See, e.g., link to geekwire.com
It depends on the nature of the invention.
By analogy:
is a patent directed to:
“A contraption which generates 10 kWh energy per ton of substance X”
The mere statement of Software (here contraption), and the mere statement of the context or desired result parents seeing what kids look like (here 10 kWh energy per ton…) reveals nothing about what invention may or may not be present in the actual Software (or contraption) which actually achieves the result.
If a claim merely stated “A contraption … to achieve X” or “Software to let parents…x” would be invalid, that has nothing to do with whether or not something has been invented and could be claimed (and properly described).
People have been imagining what other people would look like if they were older for hundreds of years.
It’s called “theatrical make up”.
You use your brain and the persons face to figure out the solution.
What does your magical computer do differently?
Fish, the so-called software actually requires inputs and outputs and more accurately what is described in the link is a method involving images.
The claims define an invention. It serves as a definition whose referents include the total sum of concretes (apparatus, acts, etc.) which fall within its metes and bounds. In essence the claims are “directed to” an invention that is a tangible thing or set of acts.
This is to be distinguished from the purpose or result, i.e. the object of the invention in use, the transformation of the information for achievement of a purpose, i.e. what the invention is “directed to”.
Does anyone else here believe there is a recurring logical error occurring, namely, of conflation between the fact that claims are directed to inventions with the fact that inventions are directed to certain ends? i.e. a conflation between the invention (what the claims are directed to) and what the invention is directed to?
The conflation you speak if is occurring at the claim drafting stage. If these new machines were described in objective structural terms that distinguish them from prior art machines, we would not be having this discussion.
MPEP 2173.05(g) Functional Limitations “There is nothing inherently wrong with defining some part of an invention in functional terms.”
“The claims define an invention.”
But the claims are not themselves the invention. Once you accept that your perspective will be freshened up.
That IS my “perspective”… claims “define” the invention.
It would be irrational to equate a claim with an invention… a claim is a string of words… not the invention itself.
“It would be irrational to equate a claim with an invention”
And yet the rest of your OP rests entirely on the claims being the invention. Gl internalizing that which you purport is already your perspective, as well as its effects on what you then want to talk about in your OP.
Useful chat about terminology here. Thanks both.
For me (and most everybody else in Europe, I suspect) the claim “defines” an “invention”. What the claim contains though is not an “invention” but rather “subject matter”. It is that “subject matter” which can exhibit the property of patentable eligibility, novelty or non-obviousness.
Drawing this distinction between “invention” and “claimed subject matter” helps focus on the claim. I never thought about it before, but now I suspect that this distinction can damp down emotion, when arguing about whether I can have a patent for my “invention”.
Question: is Europe’s “subject matter” terminology used in the same way, in the USA?
Anon2, I do believe the Supreme Court at times uses confusing language. But if you read cases like Hotel Security that helps explain the issues quite well, what they are doing is determining what in the claim is new and what in the claim is old, and whether that which is new is statutory, and if not, give it no weight from a patentability point of view.
Now they can do this at the 101 stage if and only if it is patently clear that the statutory subject matter, such as machines, manufactures or compositions of matter, or processes as in Prometheus, either are notoriously old or admitted to be old. Otherwise I can see no other choice but to analyze the claims under 102/103 to determine whether the statutory subject matter is in fact new or nonobvious. This is exactly where the problem begins if the nonstatutory subject matter is given weight in this analysis if that nonstatutory subject matter is not integrated with the statutory to produce a functional modification. Fortunately, even the Federal Circuit recognizes that nonstatutory subject matter is not be given weight in the 102/103 analysis. There are plethora recent cases on that point. (If you would like, I can cite you a some point.)
“My question here then is what importance or relevance should be given in the abstract-idea analysis as to whether or not the central idea is old or new? ”
None according to the law.
Dennis, I’ll add this one other comment. If I can write a claim in terminology that allows software engineers to understand what it covers … that allows them to know whether they would or would not be infringing it, and if it is directed to a statutory class, then there should be no problem with eligibility. If there is a question as to whether the claim is too “abstract” (a term so ill-defined it’s of no use in any principled legal analysis and is better reserved for philosophers), then leave that to a jury. Let the jury hear from experts in the field what they think the claim covers. Let proofs be offered as to the meaning of the claim. If those of ordinary skill in the art can understand what it means, then the spaghetti mess of 101 jurisprudence only complicates the issues and adds to the costs. Courts should get out of the way and apply technology-agnostic patent law as needed.
That is exactly the type of worthless eligibility test that you’re never going to see.
Why on earth would we set a jury loose on this question?
I’d sincerely like to know one reason.
Hierarchy, I agree with your post to the extent, but only to the extent, that the Supreme Court use of the word abstract not only is indefinite, and it is quite misleading.
What they are doing is applying Hotel Security without saying they are doing so.
What you describe is real patent law. The new Alice test is a “I know when I see it test.”
Well, actually, the court says as clearly as it can be said, “no.” The Alice test, for “long prevalent,” is entirely distinct from what we used to call “abstract.” The Alice court says several times that the characteristics we used to rely are are “beside the point” and “attached no importance to the fact.” All that matters now is that the idea is long in use.
Diehr and Alice are two distinct tests. The facts that are dispositive to one are irrelevant to the other.
David – it makes sense to me that it should be more difficult to patent the application of a long-prevalent-abstract-idea and to patent the application of a newly-imagined-abstract-idea. I also agree that we have multiple tests for eligiblity stemming from the various cases. However, I don’t read the Alice court as being so explicit about the difference.
it should be more difficult to patent the application of a long-prevalent-abstract-idea
Why should that matter under 35 USC 101? If it is an obvious implementation, then it is kicked out under 35 USC 103.
good point
It’s impossible to evaluate the obviousness of abstract labels applied to data in information processing claims.
The gambit of “do it with obviousness” is just a return to the status quo where the folks who depended on that status quo will go the wall with arguments that “available real estate is totally different from available X” or “identifying what was viewed from a list is totally different from identifying what was not viewed.”
But maybe you have some workable, practical ideas for making these evaluations. If so, let’s hear them.
“If it is an obvious implementation, then it is kicked out under 35 USC 103.”
Not necessarily.
Not necessarily.
How so?
Dennis:
Why qualify the concept “idea” with the term “abstract”. Can there be “non-abstract ideas”? i.e. what difference is there between “abstract ideas” and “non-abstract ideas”?
If there is no distinction in reality then we have just ideas… your question would reduce to “long prevalent ideas” and “new ideas”.
Again, just chipping in without reading the entire thread but on the issue of novelty of Abstract ideas, how about this.
The EPO is interested in that which is “new” or “not obvious” within the rubric of the useful arts. So an inventive step elsewhere (the plot of a thriller, how to pay less tax) doesn’t “count” towards patentability.
So too with novelty. On an old coffee mug, some new design etched on its cylindical face fails to make the mug patentably “novel”.
In the moment you discover a Law of Nature, is it any more patentably “novel” than the one Isaac Newton discovered? In the moment you have a “new” idea, does it have patentable “novelty”.
If so, is it simply an oxymoron to talk about novelty in abstract ideas?
Max, “If so, is it simply an oxymoron to talk about novelty in abstract ideas?”
Absolutely. It was a red herring argument introduced by Rader in Bilski who was then trying to derail the MOT, and Hotel Security (Dyk).
Max, if you haven’t done so already, you may actually want to read Hotel Security. Basically its “holding” is that in mixed subject matter claims, the nonstatutory part has to be ignored and invention must lie in the statutory part. If that is old the claim is unpatentable.
I put holding in quotes because the case was actually resolved because the claim subject matter actually was known. However it’s dicta was followed in Guthrie v. Curlett, and was followed in many CCPA cases including In re Russell. The gist of the holding is still followed by the Federal Circuit which in a recent case gave no weight the statutory subject matter in deciding novelty.
While there are some here who refuse to acknowledge any cases decided outside of the Federal Circuit, Hotel Security actually is prevailing law even in the Federal Circuit, but famously, or infamously, Judge Rich and State Street Bank refuse to follow that case. Judge Dyk in Bilski wanted to overrule State Street Bank. Just Stevens and three others in Bilski also want to overturn State Street Bank directly and restore Hotel Security.
There is no doubt that sense Prometheus, the Supreme Court is following the analysis of Hotel Security.
I haven’t read every word of all the thread, so please forgive if I now write something that has been already said.
Anybody who knows anything about EPO jurisprudence knows the “per se” cabining of the prohibition on patents for methods of doing business and programs for computers. Just as some subject matter that can fairly be summarised as a “business method” might still be fit to patent under the EPC, so might subject matter which can fairly be described as an abstract idea.
Mindful of this, SCOTUS is perhaps trying to do its own cabining, up to now by taking judicial notice of the antiquity of the underlying idea.
At the EPO, subject matter fit for patenting is presented in a claim as a combination of features. The justification for granting a patent is not that the claimed subject matter poses a problem to be solved. Rather, that it specifies the solution through a combination of technical features. Scope of protection must be no wider than commensurate with the enabled solution.
But every so often, patentability can be recognised in the mere appreciation of the problem to be solved, with a patent being fit, even if solving the stated problem is trivial and not inventive. Perhaps SCOTUS wants to keep the way free, in the USA , for patents on such “Problem Inventions”?
You people keep using the term “algorithm” as if it were something holy…
“In mathematics and computer science, an algorithm (Listeni/ˈælɡərɪðəm/ AL-gə-ri-dhəm) is a step-by-step procedure for calculations. Algorithms are used for calculation, data processing, and automated reasoning.”
Hence it is just math, and ineligible.
computer programs are just math?
Fish, is information statutory?
[Deleted by mod]
Oh no, there is no doubt that software when applied to improve a computer, the computer art, a larger machine, or conventional process is eligible.
But the problem is that people want the claim the computer program itself sans application.
But the problem is that people want the claim the computer program itself sans application.
What’s the problem? A computer program is a tool to solve a problem.
Can I claim a screwdriver without an application? Can I claim a hammer without an application? Can a claim a 5-axis CNC milling machine without an application? Of course we can.
Of course if you, can show me where “sans application” is within 35 USC, I’ll concede the point.
Oh No,
A computer is a tool to solve a problem. A program is data provided to the tool in preparation of solving a problem.
Can you claim gripping a screwdriver? Lifting a hammer? Loading a .3dm file into a CNC mill?
Wrong. A program is an instruction set or method provided to the tool so that it CAN solve the problem.
The difference is that methods are patentable, and data is not.
Nice try though.
Bad Joke Ahead,
It depends on what “level of abstraction” from which you are viewing the computer. At the hardware level, there is no difference in retrieving an instruction operator or an instruction operand from a memory cell or file location. It is all data to the processor.
If you’re going to claim the pre-existing hardware as the structure of your invention, I think it’s a bit deceitful to claim the abstracted instructions and meanings (i.e., “method”) from the programmer’s “virtual machine” view of the computer as your “innovation”. (Not considering actual improvements to computer’s data processing capabilities, or computers as elements of a larger invention or process, etc.)
Just sharing my views as a legal lay-person.
Can you claim gripping a screwdriver? Lifting a hammer? Loading a .3dm file into a CNC mill?
As part of method claims? Sure.
A program is data provided to the tool in preparation of solving a problem
As BJA put it aptly, a program is a set of instructions (i.e., steps of a process). Moreover, not only does the program use the computer as a tool, the program when combined with the computer creates another tool.
“As part of method claims? Sure.”
*Sigh* Fine. We’ll do it your way.
Can you claim a well-gripped screwdriver as an improvement to the screwdriver? Or a raised hammer as an improvement to the hammer? Or a .3dm-file-loaded CNC mill as an improvement to the CNC mill?
“As BJA put it aptly, a program is a set of instructions (i.e., steps of a process).”
As mentioned in my response to Bad Joke Ahead, instructions and data are the same thing at the hardware level.
“Moreover, not only does the program use the computer as a tool, ”
This sounds like some of NWPA’s Skynet-like machine intelligence prognostication.
“the program when combined with the computer creates another tool.”
As you’re undoubtedly aware, there is a lot of disagreement on that point.
Can you claim a …
Why don’t you use more realistic examples?
instructions and data are the same thing at the hardware level
One skilled in the art identifies instructions and data as being different animals. Instructions are the steps by which the computer operates. Data is what is being operated on. Instructions are a process, and a process is protectable under 35 USC 101.
As you’re undoubtedly aware, there is a lot of disagreement on that point.
Really? Let me come to your house and delete ALL the software on your computer. After that, tell me what that computer in that particular state is good for aside from being a paperweight?
Actually, to the CPU, there is no difference between instructions and data.
That is why buffer overflow errors are such a problem…. The CPU starts executing data…
Oh no,
“Why don’t you use more realistic examples?”
Yeah, you’re right, they’re really bad examples. Maybe I shouldn’t have copied them from that guy in post 11.1.1.1.1.1. I’ll try to pick better sources in the future.
“One skilled in the art identifies instructions and data as being different animals.”
But the “configured” computer does not, and that’s what’s being claimed.
“Instructions are a process, and a process is protectable under 35 USC 101.”
Only if you’re willing to stretch the meaning of “process” beyond what was contemplated by the drafters of 35 USC 101. Are we going to start patenting thought processes, legal processes, the grieving process, the trochanter major process, . . .
“Let me come to your house and delete ALL the software on your computer. After that, tell me what that computer in that particular state is good for aside from being a paperweight?”
If you delete all the saved channels and input assignments off of your TV, you’ve got a pretty nice paperweight, too. Putting the channels back in there doesn’t make it a new machine, though.
If you delete ALL of a computer’s software (OS, BIOS, init vector code, CPU microcode, etc.), I would support the argument that you no longer have a general purpose computer. Does that help?
Actually, that is only true of a PC, or any system that include more than the CPU+MEMORY.
General purpose computers don’t HAVE to have software to be general purpose.
They do have to have some form of input and output devices.
General purpose computers existed far longer than BIOS and writable microcode…
Even the PDP-8, PDP-11, IBM-360… and a host of others did not come with extra software (though the 360 did have an option for writable control store).
They could be started from the console switches…
But the “configured” computer does not, and that’s what’s being claimed.
Claims are interpreted by those skilled in the art — not a computer. A computer sees everything as data objects. Does that mean nothing is patentable since data objects are not patentable?
Only if you’re willing to stretch the meaning of “process” beyond what was contemplated by the drafters of 35 USC 101
You have insight as to what they contemplated? Please share your sources.
Putting the channels back in there doesn’t make it a new machine, though.
It is a different machine, which is the same as “new” is patent parlance. What is different than the past is patentable. You’ve changed its state, it is different.
Oh no, after your last series of posts where it has become apparent that you do not wish to engage in any serious conversation, I am putting you on ignore.
I am putting you on ignore.
My prayers have been answered. Thank you.
I guess it was my taking apart your citation to the Rubber-Tip Pencil Co. case that put you over the edge — so sorry.
Is a machine statutory?
A machine is one of the four classes; but only a new or improved, and useful machine is statutory.
And the improvement must be defined in distinct objective structural terms. Otherwise you’re just claiming your dream about a nice functionality flowing from an improvement that you haven’t described in your claim.
MM – I don’t quite understand what you are saying here. All of the elements must be distinctly claimed under 35 U.S.C. 112. However, I would expect that the Means-Plus-Function language could be used for the inventive eligible elements (the “something more”) of a claim that would otherwise be ineligible.
Dennis, However, I would expect that the Means-Plus-Function language could be used for the inventive eligible elements (the “something more”) of a claim that would otherwise be ineligible.
Frederico did not agree.
improvement to a method claim needs to be “defined in distinct objective structural terms” too???
“Fish, is information statutory?”
Information is statutory if the information is a process.
Now there is information that is not a process such as a poem, or novel, or lyrics to a song, and this type of information is not a patent eligible subject matter.
Neither is math…
On the contrary, I find that many people regard other technologies as “holy,” such as semiconductor fabrication, chemical processes, and nuclear physics. These fields are perceived with such reverence that nearly any adjustment, no matter how small or routine, is perceived as sacrosanct and worthy of patenting – if only the applicant can prove that it is different.
This reverence stands in stark contrast with computing technologies, where people can look at even impressive leaps over preceding technology and say, “it’s just software.”
I believe that familiarity with computing technology, and the stunning pace of innovation, has numbed people to remarkable advances. Self-driving cars? Cameras that see around corners, and can analyze our body movements in precise detail? Transmitting live video halfway across the globe on demand? Meh. We’ve been seeing that stuff in sci-fi for decades; now that it’s here, it already seems passe.
I believe that this difference in attitude, between utmost respect for other sciences and complete disrespect for computers and specifically software, is a reflection of Parkinson’s Law of Triviality, or the “bike shed” principle:
When computation thought impossible a year ago becomes achievable today, it should be even more astounding that it is achievable by the extraordinary computational powers of a phone in our pocket. Instead, this accessibility makes such advances seem mundane. It crushes our perspective and our ability to be amazed: the most exquisite leap into the impossible holds our wonder for only a brief moment before becoming totally conventional.
I believe that familiarity with computing technology, and the stunning pace of innovation, has numbed people to remarkable advances. Self-driving cars? Cameras that see around corners, and can analyze our body movements in precise detail? Transmitting live video halfway across the globe on demand? Meh. We’ve been seeing that stuff in sci-fi for decades; now that it’s here, it already seems passe.
The concepts themselves — those you recited in particular — are beyond passe. They were passe twenty years ago, if not fifty years ago. That’s one of the reasons it’s so bizarre that certain people still believe they are entitled to claim such concepts provided they recite some shiny words like “on a computer.”
That’s one of the reasons it’s so bizarre that certain people still believe they are entitled to claim such concepts provided they recite some shiny words like “on a computer.”
Except that they don’t. Nice straw man — do you have a monthly quota of those to fill?
So you agree that merely adding “on a computer” to otherwise ineligible subject matter does not rescue the claim from ineligibity?
otherwise ineligible subject matter
Let’s assume the conclusion, shall we.
too bad.
It has been mathematically proven than any algorithm may be the input to another algorithm.
It has even been demonstrated by simulating one computer on a totally different computer.
Even the earliest analysis had shown that – in the definition of the Universal Turing Machine. A “machine” that can emulate any other Turing Machine – including the Universal Turing Machine.
It had already been shown that any computable function could be computed by a Turing machine…
Thus any computer can emulate any other computer. The only thing that affects the result is the physics of the electronics that are themselves emulating the math functions of Boolean algebra.
The piece there is that math doesn’t take account of time (physics does though). Math does, however, have the concept of sequence.
Again great comments David. I see in my practice exactly what you are saying. In fact I just interviewed an examiner who said, oh year isn’t this like what I saw in the Sci Fi movie?
And in other areas like as you say silicon manufacturing it is easy to get a patent.
I think that what we have to realize is that big corp doesn’t want software patents and that buys a lot of what we deal with. They have people that just fabricate all these arguments and it is hard to fight this. They have people paid to blog too.
Specious argument after specious argument crashed and burned for 20 years and then they came up with the abstract argument and it stuck. Not sure why.
David, did you know that once in ancient history that the United States sent a number of manned missions to the moon, drove cars all over the place there, planted flags, did experiments, and the like; and all this before the invention of the integrated circuit?
Interesting trivia: Noyce, Semiconductor device-and-lead structure, U. S. Patent 2981877, filed in July 30, 1959. There may be others. Neil Armstrong did not make it to the moon until 1969.
So what? Star Wars was a brilliant movie in it’s day, showing astounding novelty and expensive technical virtuosity.
Not.Patent.Eligible.
Brilliant, clever, expensive software should not be patent eligible either, because reasonable minds can differ about the expressions of ideas using languages, which is what software is.
Shouldn’t be patentable, because you would like to steal the ideas incorporated within it without having to pay the person(s) who invested their time and capital and creativity to bring it about.
How much “capital” does it take to write out a new computer functionality?
As far as I can tell, the answer is “none”.
That’s part of the reason why we need to put some serious restrictions down.