Enco Systems, Inc. v. DaVincia, LLC (Fed. Cir. 2021) [ENCO]
Judge Stephen Limbaugh (E.D. Mo.) sided with the accused infringer DaVincia– holding that the claims of ENCO’s U.S. Patent No. 7,047,191 are invalid under 35 U.S.C. § 101 as directed to an abstract idea. On appeal, the Federal Circuit has affirmed.
U.S. Patent No. 7,047,191 claims a method of providing captioning in an audio-visual signal using speech-to-text processing. Claim 1 includes a number of limitations:
- Selecting the number of lines of caption data to be displayed (I have an image below showing how my phone does this).
- Determining the caption encoder system being used
- Training the system to on new words;
- Using AV cues to time the captioning so that it displays at the appropriate time.
On motion to dismiss, the district court found the claims directed to the abstract idea of “automated stenography implemented on a computer.” The court looked particularly to the claim limitations and found them written at a “high-level of generality” and using “broad form functional terminology.” With regard to Alice step two, the court found the claim limitations lacked any particular or concrete configuration that could serve to ground the abstract idea.
To know whether a patent claim is improperly “directed to” an abstract idea, the court have been looking to the claims and specification in a search for objective suggestions of what the inventor thinks is the advance provided by the invention. What does the patent document assert as the “focus of the claimed advance over the prior art.” Slip Op, quoting Affinity Labs of Tex., LLC v. DIRECTV, LLC, 838 F.3d 1253 (Fed. Cir. 2016). Here, the court looked to the claims and the specification and concluded that the focus “is simply the abstract idea of automating the AV-captioning process.” In this process, the court is typically looking for a “technical solution to a technical problem,” although that is not always required. Here, the court noted that, although the invention involves computers it is not directed toward “any specific improved computer techniques for performing those functions—functions intrinsic to the concept of AV captioning.” Rather, the benefit from the invention is simply automation of work previously done by humans.
Although the abstract idea test is not a novelty test, the courts repeatedly fall back on novelty in their explanation. Here, the court writes:
The advance is only at the abstract level of computerization because claim 1 fails to set forth specific techniques for processing the data, instead reciting known computer techniques for automation of known processes.
Slip Op. The court also found no help for the patentee under Alice step two. “The claims do not incorporate anything more beyond conventional computing hardware and software, which do not transform the subject matter into an eligible application of the abstract idea.”
In a recent “informative” opinion, the PTAB found speech-to-text patent claims eligible under Alice. Ex parte Hannun, No. 2018-003323, 2019 WL 7407450 (P.T.A.B. Apr. 1, 2019). On appeal the court gave no deference to the PTAB’s interpretation of eligibility doctrine and also distinguished the case — noting that Hannun recited a particular algorithm for measuring tailored parameters.
= = = = =
I love figure 1 below with the bulky video camera and boxy computer equipment. The patent claims priority back to a 2000 filing date — before Apple rounded all the corners.
I appreciate the discussion following my post #5 below.
What does “directed to” mean if it is not a reference to the claimed invention?
Once we depart from using the claims as the measure of the invention, the whole analysis becomes subjective, unpredictable, and unjust.
It is just bizarre and has no meaning other than I am going to say your claims are abstract and here is the incantation.
I once read everything under the Sun that had to do with patent law going back into England. I know in some case I read that “abstract” as an exception was meant to mean something that wasn’t enabled. So, I remember the example that I am pretty sure was in a legal opinion from England in the 1700’s was “a machine with fewer parts that operates faster.’ And with no enabling disclosure. So this is an abstraction as it is not enabled. It makes sense that claims directed to an abstraction like this would not be patentable.
What I think the Scotus did was cleverly figure out that abstraction had this meaning but that 112 took care of that problem so they just added a new meaning to it. The new meaning is just yap out “directed to” that is broader than the claim and then say the claim provides nothing more and that all computers are just generic tools so using them can’t add anything more. Bingo. There goes any claim under the Sun.
NB. I have quoted engineering books used at MIT that state explicitly that functional language is necessary for discussing modern technology as there are so many solutions to perform the functions. The functional language is meant to encompass all known solutions.
That is how functional language is used by modern innovators.
What does this have to do with patent claims?
Patents and engineering books serve different purposes, which is why they are not interchangeable.
Leave it to NSII to choose to blunder in on this point.
Just as Greg “I Use My Real Name” DeLassus has done in the past — and squelched by me personally — any actual patent attorney** full well understands the legal analog to Night Writer’s engineering point (Ladders of Abstraction, see at least Cluskey).
** not in the “true Scotsman” sense, but rather in the sense of a registered attorney with actual prosecution experience – the type of experience in which one is active in protecting innovation.
NS II:
No. You have no idea what you are talking about. Read Phillips. Read about the presumption in the Patent Act that the patentee knows how to present their invention.
And just add this as it is part of this nonsense.
J. Breyer and Stevens have said the way you do something on a computer is you write it down and then give it to a clerk to program.
Nothing could be further from the truth.
And, they miss that whether a claim is enabled is based on whether with the specification the problem can even be solved.
But this is another aspect of this that is at the core of the false assertions made by the Alice people.
RandomGuy>So what the CLAIM SCOPE is is nothing more than “we should have a processor do what the human traditionally did.” It may be that the specification discloses an eligible invention (for example, an algorithm that actually accomplishes that end). But that is not what was claimed. What was claimed was the abstract functionality of a computer doing what a human used to do.
This is total nonsense. Why? Because the claims are covering are the scope of solutions that are enabled by the specification. The abstraction is the language used to convey the meaning of here we claim all the solutions that are enabled by the specification to perform these functions. That is the only way modern technology in almost every area can be claimed.
Reality. Please.
In all of these cases from Alice to this one, what is missing is actually solutions that are enabled by the specification for a person skilled in the art.
When you are real person that does real stuff and work with people that create real technology, all of these opinions seem absurd. What are they talking about? There is just this set of solutions to the problem and you claim it in functional language and the CAFC says it is abstract? How is that?
It is just nonsense that is created to invalidate claims the CAFC doesn’t like. And all this gibberish about “directed to” is absurd. The claims are functional statements that capture a set of solutions to the functional problem claimed. That is pretty much how every technology has to be claimed now.
“Because the claims are covering are the scope of solutions that are enabled by the specification.”
How can the above be true with the enablement requirement?
Say someone receives a patent claiming “A fast Fourier tranform algorithm with a complexity less than O(n log n).” Their specification discloses a O(n) FFT algorithm.
Say the accused infringer argues that the claim scope includes a O(log log n) FFT algorithm (since O(log log n)<O(n log n) ), and that the claim is not enabled because the spec doesn't enable the O(log log n) algorithm.
Your claim interpretation rule would limit the scope of the claim to O(n) FFT algorithms, and if the claim scope is limited to O(n) FFT algorithms there is no enablement issue.
In effect, you claim interpretation rule (scope limited to enabled embodiments) appears to tautologically define every claim as fully enabled. How can that be reconciled with the enablement requirement
First, not every claim is enabled. “Generate an opinion and order from the CAFC based on submitted papers that applies current US law and will not be overturned by the Scotus.”
Second, you state the typical complaint. That is the reverse doctrine of equivalents or what is discussed in LazardTech. Basically, what if you claim something but someone comes up with a new solution that your claim shouldn’t cover. This is the one and only problem with this approach and this problem is easily dealt with. There were cases that were forming when the CAFC was a real circuit and not an outlet of SV.
And, you stated this in a way to try and make it seem like a person of ordinary skill would have a hard time understanding the scope of the claims. They would not. Nor would they have a hard time telling if something wasn’t enabled. Nor would they have a hard time telling if something was a new solution that came after the patent application filing and shouldn’t be included in the claim scope.
“Second, you state the typical complaint. That is the reverse doctrine of equivalents or what is discussed in LazardTech. Basically, what if you claim something but someone comes up with a new solution that your claim shouldn’t cover.”
This is really not my point. I’m not crying on behalf of the O(log log n) FFT algorithm inventor here, I’m asking how does your interpretation comport with the law when you tautologically define claims to be enabled.
“First, not every claim is enabled.”
How can a claim not be enabled when it’s scope is limited to the enabled embodiments? A claim to a faster-than-light drive currently has a null claim scope, because there are no enabled embodiments. How do you show a claim with null scope to be untenable?
“And, you stated this in a way to try and make it seem like a person of ordinary skill would have a hard time understanding the scope of the claims.”
This wasn’t something I was trying to get at, but I would say it’s a problem. Isn’t enablement a question of law? Do you really think its reasonable for people to have to consider the Wands factors to determine claim scope? Or are you using some different short-cut enablement analysis to determine claim scope?
This is what I said>>> The claims are functional statements that capture a set of solutions to the functional problem claimed. That is pretty much how every technology has to be claimed now.
You seem to be making a big deal out of the fact that it is possible that there are no solutions to the functional statements in which case obviously the claims are not enabled and invalid.
Or, if you want, a functional claim captures all the known solutions for a person of ordinary skill in the art in view of the specification. Special case: if there are no enabled solutions then the claims are invalid.
Talk about making a mountain out of a mole hill.
>>This wasn’t something I was trying to get at, but I would say >it’s a problem. Isn’t enablement a question of law? Do you really >think its reasonable for people to have to consider the Wands >factors to determine claim scope? Or are you using some >different short-cut enablement analysis to determine claim >scope?
This is already the case in every instance. Still, the reality of practice is that people of ordinary skill in the art know exactly the set of solutions that a functional claim includes. Guess what? We spent many years learning this in engineering college.
“This is already the case in every instance.”
Can you point me to one CAFC or CCPA case that uses the Wands factors for claim construction?
I think Ben that you are going for edge cases to try to make this seem more complicated than it is.
There are edge embodiments where it may not be clear if they are covered by a functional claim but this is always the case.
I would reverse the question on you. Tell me why you think the approach I outlined which is already the approach that is used in LizardTech and other opinions and which is the approach that is explicitly stated in engineering textbooks is hard or different than whatever you think is being done now?
When I say the enabled claims I think this is pretty much the same as saying how one of ordinary skill in the art would interpret the claims and I have quoted from engineering textbooks from MIT to back this up.
I see this approach as Phillips v. AWH Corp. I was using enabled in the sense of a person of ordinary skill seeing the functional claim as enabled. As in what functions would perform those functions that the inventor would know. The other edge cases cause trouble no matter what you do.
You know reality is important here. I recently worked on a system for hydraulics and part of the invention performed a function on the hydraulic system. The inventor then started listing all the ways that it could be performed and then said, yeah there are bunch more ways too. We’d be here all day if I had to explain everyone to you.
Another time I was writing a patent for an electronic device. The inventor a ph.d. in physics said, yeah, well you could perform this function in many ways and then spent time telling me some of the major ways.
This is typical. Not special.
The fact is that the method I outline was used in LizardTech and was what the CAFC used before the Obama judges came.
“I would reverse the question on you.”
I’m sure you’d like too, but that’s not how logic works. You don’t defend your preposition by demanding evidence of your interlocutor.
Look, I think the system you describe sounds good (at least better than what we currently have). I think it’d be a plausible way to limit functional claiming, and if you were arguing perscriptively rather than descriptively I’d probably be an ally. I just don’t think this is the current system we have.
If you could put up a case where the Wands factors were used in claim construction, it’d be strong evidence of your preposition. That’s not trying to make things complicated. That’s trying to find decisive evidence.
Apparently I am out of responses so you get the last word such as it is Ben.
NW – the count filter resets after 24 hours.
I’ve replied to Ben twice and both were caught in the filter. Ben raises no issue and is not responding to what I wrote.
Ah, so it is more likely that your post was swallowed by the George Carlin filter….
That can be tougher to figure out, as that filter has different levels, and those like you and I that have challenged the “desired narratives” have higher levels.
This I do know, as I have traced direct “tripping” of the George Carlin filter to words not of my own, but a direct cut-n-paste of someone that I was responding to. Their use tripped no filter, but I post the exact same words, and BOOM.
Interesting anon.
I have noticed similar things and am not sure what to make of it.
There is a simply thing to “make” of it (think Occam):
“Rules” here – as such as they are – are NOT applied objectively.
Fun fact—the MO Eastern District Judge here is cousin to the late Rush Limbaugh. The DJ’s dad was an MO federal judge too, but in the Western District instead.
Please allow me to vent.
In Alice step one, we are asked to determine if the claim is “directed to” an abstract idea. In doing that, the actual language of the claims is improperly ignored or given too little weight.
But decades of binding Supreme Court precedent tell us that the claims are the sole measure of the rights granted by a patent.
Teva Pharm. USA, Inc. v. Sandoz, Inc., 574 U.S. 318, 342 (2015) (“The scope of a patent holder’s monopoly right is defined by claims legally actualized through the procedures established by Congress pursuant to its patent power.”);
Aro Mfg. Co. v. Convertible Top Replacement Co., 365 U.S. 336, 339 (1961) (“Since the patentees never claimed the fabric or its shape as their invention, and the claims made in the patent are the sole measure of the grant . . . .”);
Universal Oil Prod. Co. v. Globe Oil & Ref. Co., 322 U.S. 471, 484 (1944) (“The claim is the measure of the grant.”);
Milcor Steel Co. v. George A. Fuller Co., 316 U.S. 143, 145–46 (1942) (“As the courts below pointed out, it is these claims, not the specifications, that afford the measure of the grant to the patentee.”).
The entire “directed to” inquiry ignores the primacy of patent claims in defining the nature and scope of the patentee’s rights.
A patentee cannot ignore the claims when proving infringement. An alleged infringer – generally – cannot ignore the claims when proving invalidity under 102 or 103.
So why do we have this silly, subjective, non-sensical approach to analyzing for invalidity under 101?!?!
And why won’t the Supreme Court or Congress fix it?!?!
“So why do we have this silly, subjective, non-sensical approach to analyzing for invalidity under 101?!?!”
Because the 1952 law was not up to the task of preventing detritus from being patented, and applicants took advantage of that fact.
Ends do not justify the Means — especially when it simply is NOT up to the Supreme Court to impose their feelings of what is and what is not “detritus.”
The entire “directed to” inquiry ignores the primacy of patent claims in defining the nature and scope of the patentee’s rights.
The Alice test considers all of the claim language, so I don’t know what the complaint is.
In doing that, the actual language of the claims is improperly ignored or given too little weight.
No limitations are ignored. Shouldn’t context limitations be given “less” weight? (And to be clear, “less” weight means it is considered under Step Two rather than Step One)
In this case, claim 1, together with the specification, makes clear that the focus of the claimed advance is simply the abstract idea of automating the AVcaptioning process. That process, consisting of converting audio to text and associating the text with corresponding video, is not itself asserted to be an advance over the prior art. The focus is not any specific improved computer techniques for performing those functions—functions intrinsic
to the concept of AV captioning—but simply the use of computers to “conserve human resources” by automating work otherwise performed through human labor
So what the CLAIM SCOPE is is nothing more than “we should have a processor do what the human traditionally did.” It may be that the specification discloses an eligible invention (for example, an algorithm that actually accomplishes that end). But that is not what was claimed. What was claimed was the abstract functionality of a computer doing what a human used to do.
And there lies the rub, eh? Every claim drafter strives to craft a scope wide enough to embrace any conceivable “design around”. But over-striving results in claiming nothing more than the problem to be solved, as opposed to the invention, namely the enabled and described solution to that problem.
I have a minor quibble about an otherwise spot-on title to this post. What a patent application is “directed to” is defined by the claim which, as somebody once said is indeed “the name of the game”. If it can be correlated to “a technical solution to a technical problem” that ought to be enough to render nugatory any suggestion that it fails to satisfy 101 eligibility. Not only the solution but also the problem has to be “technical” , if you want not to be held back by 101 challenges.
Are you seeing that “I claim the problem to be solved.” here?
What I’m seeing is the court declaring that the level of generality (abstraction, if you will) at which the claim is expressed corresponds to nothing more or less than “automating the AV-captioning process”. Anybody else who automates will infringe.
Now, I know very little about how to automate the AV-captioning process but I imagine it’s not a piece of cake. One of the claim features is to “train” the AI. So enablement might be an issue, even though the patent owner waves their hands and declares that it’s obvious for the skilled person how to “train” the machine to a level high enough to generate spot-on (rather than laughable) captions.
But the thing is, automation, as such, is obvious. Just saying “Do it” doesn’t cut the mustard.
So there you go: abstract, yet obvious. Oh, and not enabled either.
Actually, the squeeze between obviousness and enablement is , these days, endemic. A pandemic of such claims, one might suggest. Radical measures are needed, to contain it, some might think.
You’re sad about it, you say. I’m not.
I am sad about her willingness to apply a “whatever” Ends justify the Means approach.
And if you had any sense of appreciation of law and the Rule of Law, you would too.
“About her” ==> “about the”
Damm autocorrect…
I replied to your “sadness” lark, but you never circled back to answer my direct question to you, so let’s try again, MaxDrie:
Are you seeing that “I claim the problem to be solved.” here?
Your misstatement of 101 and “technical” is noted.
Sadly – and very much part of the problem – is that the broken score board does seem to reflect that improper (and too constrained per the actual US Sovereign choice) of Useful Arts.
An alleged infringer – generally – cannot ignore the claims when proving invalidity under 102 or 103.
So why do we have this silly, subjective, non-sensical approach to analyzing for invalidity under 101?!?!</I
102 and 103 specifically use the term "claimed invention."
101 uses the term invention (or rather, "invents or discovers.".
The "invention" and the "claimed invention" are two different things under USC100.
If you are going to go for precision, can 35 US 101 even be used for invalidity?
Re: “decades of binding Supreme Court precedent tell us that the claims are the sole measure of the rights granted by a patent.”
Well, yes, but we also have lots of other patent law decisions that say that “claims are read in light of the specification.” [To confuse the issue?]
Plus, there are are at least two other complicating case law issues re claim scope [not just 102, 103 or Alice-101]. Namely, 112 (enablement, insufficient enablement examples, 112(f), etc.) and/or “preemption.” [Neither adequately handled in PTO examination.] Which seems to make a statutory fix not that simple, even if there were not also directly conflicting lobbying interests.
I could not parse that Paul.
Can you tell me how many points you tried to squeeze in there?
A third complication in non-literal claim scope interpretation is “the doctrine of equivalents,” which seems to be increasingly asserted in litigation again these days.
CW5: “And why won’t the Supreme Court or Congress fix it?!?!
Supreme Court: As patent Gods, we know better than Congress and the Constitution.
Congress: Big Tech Big Money.
“On appeal the court gave no deference to the PTAB’s interpretation of eligibility doctrine”
Ex Parte Hannun incorporates the Iancu test. Why would anyone expect the CAFC to give deference to such a decision?
Better yet, the CAFC itself does not give deference from one panel to the next, thus exhibiting the Gordian Knot that I have LONG referenced.
One can also look for my prior references of “ping” and “pong” as the panel cases bounced back and forth.
And why would any inventor expect to rely on a PTO claim allowance in any PTO proceeding based on the Iancu Guidelines for Alice 101, after the Fed. Cir. has clearly twice said those guidelines do NOT apply to real [litigated] Alice 101 decisions, which they get to make [under vague Sup. Ct. requirements]? [Even if not fully consistent between different Fed. Cir. panels and cases.]
On second thought, my comment is not fair to an inventor and/or assignee who had a patent attorney not paying adequate attention to Alice 101 decisions and/or not adequately counseling the inventor re potentially subject claims of this particular claim enforcement risk, as I suspect may be too often the case?
“Not paying attention”….
Like you? Not paying attention to the actual inconsistencies and the fact that the score board is broken?
For shame.
There is NO such thing as “fully consistent” when you have the ACTUAL bolus of CAFC cases, Paul.
That’s a comment that Malcolm tried to make way back when.
It did not work for him either. It is entirely inte11ectually dishonest.
The Iancu test on its face sought to promote consistency by making less-than-all possible 101 rejections, and it’s being applied by non-lawyers who have only minimal training in caselaw (because the test does not want examiners applying caselaw) so why would its use have any nexus to what the CAFC does?
Because there is a huge difference between applying case law (or in most instances by examiners, misapplying) and still doing your job – you know, examining under the law.
Here, the court looked to the claims and the specification and concluded that the focus “is simply the abstract idea of automating the AV-captioning process.”
Automating an AV-captioning process is and will never be an abstract idea. It is a technological process that has required and will always require the use of technology in order to be implemented.
On appeal the court gave no deference to the PTAB’s interpretation of eligibility doctrine
If an agency with over 10,000 employees cannot accurately determine patent eligibility based upon the case law, how can an inventor or patent professional?
Here is claim 1 in its entirety:
1. A method for providing captioning in an AV signal, the method comprising:
selecting a number of lines of caption data which can be displayed at one time;
determining a type of a caption encoder being used with a speech-to-text processing system;
retrieving settings for the speech-to-text processing system to communicate with the caption encoder based on the identification of the caption encoder;
automatically identifying a voice and speech pattern in an audio signal from a plurality of voice and speech patterns with the speech-to-text processing system;
training the speech-to-text processing system to learn one or more new words in the audio signal;
directly translating the audio signal in the AV signal to caption data automatically with the speech-to-text processing system, wherein the direct translation is adjusted by the speech-to-text processing system based on the training and the identification of the voice and speech pattern;
associating the caption data with the AV signal at a time substantially corresponding with the converted audio signal in the AV signal from which the caption data was directly translated with the speech-to-text processing system, wherein the associating further comprises synchronizing the caption data with one or more cues in the AV signal; and
displaying the AV signal with the caption data at the time substantially corresponding with the converted audio signal in the AV signal, wherein the number of lines of caption data which is displayed is based on the selection.
This claim would appear to easily pass the 2019 Patent Eligibility Guidelines. How can an inventor reasonably rely upon the US Patent System to protect their technology when it is next to impossible for anybody to determine whether the Federal Circuit will subsequently invalidate any issued patent?
By comparison, this is the claim for DDR Holdings:
13. An e-commerce outsourcing system comprising:
a) a data store including a look and feel description associated with a host web page having a link correlated with a commerce object; and
b) a computer processor coupled to the data store and in communication through the Internet with the host web page and programmed, upon receiving an indication that the link has been activated by a visitor computer in Internet communication with the host web page, to serve a composite web page to the visitor computer wit[h] a look and feel based on the look and feel description in the data store and with content based on the commerce object associated wit[h] the link.
This claim is patent eligible but the other one is not?
Under Article I Section 8 | Clause 8, Congress shall have power “To promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries.” Decisions such as these are the antithesis of promoting the progress of science and the useful arts.
Why spend spend tens of thousands of dollars obtaining a patent and hundreds of thousands of dollars enforcing a patent when the validity of the patent is a complete unknown? This is the question that the anti-patent, efficient infringer crowd wants every inventory and small business to ask prior to applying for a patent. The Federal Circuit appears to be doing everything in its power to frustrate the goals of the US Constitution.
They are but simians in a cage, trained by fire hose by the Supreme Court.
They don’t know why one must not climb that ladder in the middle of the cage and reach for the hanging bananas, they only know that such must not be allowed to happen.
Automating an AV-captioning process is and will never be an abstract idea
Don’t look now but that itself is an expressed abstract idea. You didn’t actually do anything that resulted in the captioning process being automated right there, did you? Do you personally even know how to automate the captioning process?
As a matter of fact, if you added the words “by using a computer” to the end of your sentence, you would have improved the art just as much as this claim does.
Is it possible to have concrete embodiments of this abstraction? Sure. Were they disclosed? Possibly. But that’s not the claim scope claimed. THIS claim is an abstraction.
If an agency with over 10,000 employees cannot accurately determine patent eligibility based upon the case law, how can an inventor or patent professional?…This claim would appear to easily pass the 2019 Patent Eligibility Guidelines.
I agree it would easily pass. Doesn’t that suggest that regardless of how many employees you have, if you give them inaccurate guidelines, they won’t reach the right answer?
Here’s a helpful test though: If you have a bunch of conventional contextual limitations followed by a limitation that just states the utility or functionality of your invention, you’re going to fail eligibility all the time. And you’re right to. That amounts to a claim on the function itself. This guy didn’t invent EVERY way of coding a computer that automates closed captioning. At most he invented one way (the way consistent with any disclosed algorithm in the spec). So the claims are clearly overbroad and abstract. Most people simply don’t get that what the CAFC would call a Written Description violation the Supreme Court calls either a failure to point out and particularly claim or a claim to abstraction. The Supreme Court has never adopted the Written Description test, it’s a child of the CAFC and has only been around about 20 years. The Supreme Court polices an overbroad claim of possession as a claim to abstraction.
Nobody is being denied anything they didn’t invent. This guy could have generated a claim to the algorithm he posited, and then other people who used that algorithm would be infringing. This guy didn’t invent every way of automating the closed captioning process, it’s unlikely this defendant is using an equivalent algorithm, and the defendant SHOULD be free from suit. You’re not liable for infringement just because your means has some basic level of similar RESULTS to someone else’s totally different means. The patent grant is a grant on means, not a grant on ends.
Do you personally even know how to automate the captioning process?
Forget ‘automating’ the captioning process. Explain to me how this process is performed without using technology.
Doesn’t that suggest that regardless of how many employees you have, if you give them inaccurate guidelines, they won’t reach the right answer?
Who gives these guidelines? Oh wait, isn’t it the agency itself charged with having expertise in this particular area?
If you have a bunch of conventional contextual limitations followed by a limitation that just states the utility or functionality of your invention, you’re going to fail eligibility all the time.
A goobbledygook test — got it.
This guy didn’t invent EVERY way of coding a computer that automates closed captioning. At most he invented one way (the way consistent with any disclosed algorithm in the spec). So the claims are clearly overbroad and abstract.
Was this statement intended to make sense? Oh after reading what else your wrote, it appears you think all he claimed was automating the close captioning process. No wonder you are traveling down the wrong road. In looking at the cited prior art for this patent, it looks like the automation of the close captioning process has been around for at least 25 years and probably predates the patent by 5-10 years. They weren’t attempting to claim the broad concept of automating close captioning — rather the invention was directed to some specific improvement.
Forget ‘automating’ the captioning process. Explain to me how this process is performed without using technology.
Why? “Performed using technology” isn’t the standard. Abstraction is the standard. A nebulous concept or result absent the particular means that achieve it is abstract.
Oh wait, isn’t it the agency itself charged with having expertise in this particular area?
What do you mean by expertise? A very small percentage of the examiner corp are attorneys, which would be the relevant expertise for the legal question of what constitutes the scope of protectable subject matter. If this was a fact-intensive question (where the office DOES have expertise) they wouldn’t be subject to summary judgements.
A goobbledygook test — got it.
It’s gobbledygook to you to be told you can’t claim a result, only the means that makes that result?
Oh after reading what else your wrote, it appears you think all he claimed was automating the close captioning process…They weren’t attempting to claim the broad concept of automating close captioning — rather the invention was directed to some specific improvement.
You could try reading the specification, or failing that the decision where they quote from the specification:
See ’191 patent, col. 1, lines 24–56 (in describing the asserted advance, focusing on the fact that in prior-art systems, “captions are either typed-in from a script or are typed in real-time by stenographers” or “an individual listens to a recorded audio signal and manually inputs the caption information as text in a computer”)
The specification outright says that the improvement is automating the speech-to-text function that people used to manually type in. That’s what the court held:
In this case, claim 1, together with the specification, makes clear that the focus of the claimed advance is simply the abstract idea of automating the AVcaptioning process. That process, consisting of converting audio to text and associating the text with corresponding
video, is not itself asserted to be an advance over the prior art. The focus is not any specific improved computer techniques for performing those functions—functions intrinsic to the concept of AV captioning—but simply the use of computers to “conserve human resources” by automating work otherwise performed through human labor.
And it’s pretty clear where those features are in the claims:
automatically identifying a voice and speech pattern in an audio signal from a plurality of voice and speech patterns with the speech-to-text processing system;
training the speech-to-text processing system to learn one or more new words in the audio signal;
In other words, it was a training speech-to-text system for a closed captioning context. Now speech-to-text is something that is protectable if you, you know, describe the program with the algorithms. It’s not protectable to claim the CONCEPT of making the computer convert a person’s speech into words, which is all the claim is doing. It’s not limiting it to the particular posited means, it’s claiming ANY SOFTWARE that can train new speech-to-text words.
So no, you’re wrong to say “the invention was directed to some specific improvement.” As near as I can tell from the decision, no improvement whatsoever was argued beyond the fact that having a computer convert the audio signal to text would be faster and require less human resources than if a human did it. That’s not an improvement, that’s the natural result of having the abstract idea performed by the computer rather than the human.
In looking at the cited prior art for this patent, it looks like the automation of the close captioning process has been around for at least 25 years and probably predates the patent by 5-10 years.
The specification specifically mentions Dragon Speech-To-Text as existing software that does just this kind of training, which is why there’s no “solving a technological problem” question – the specification admits that any technological problem was already solved. That’s part of the reason this is such a cut-and-dry 101 case.
I think you’re confusing the fact that the claim as a whole was found non-obvious with some sort of notion that there must have been novel functionality. But it could have been found non-obvious because of a lack of express motivation to combine (this is a pre-KSR patent) despite all of the parts being known. Regardless, being novel and non-obvious isn’t relevant for eligibility – if 103 was sufficient to allay the judicial exception concerns, there’d be no judicial exception analysis.
Abstraction is the standard
No — “abstraction” is not the standard. At some level, “all inventions . . . embody, use, reflect, rest upon, or apply laws of nature, natural phenomena, or abstract ideas.” Claims are defined using words and words themselves are abstractions. We don’t know what is meant by being ‘directed to an abstract idea’ because the Supreme Court chose not to give us a standard.
What do you mean by expertise?
You know … the agency charged with executing the law should have expertise in the law. That is their job. My job is to obtain patents for my client. The Patent Office’s job is to apply the law. Who else should have expertise if it isn’t the Patent Office?
It’s gobbledygook to you to be told you can’t claim a result, only the means that makes that result?
It is a meaningless test untethered from the language of the statute and the case law.
The specification outright says that the improvement is automating the speech-to-text function that people used to manually type in.
Except that it wasn’t. You, like the Federal Circuit, fail to appreciate that the specification is written before all of the possible prior art is known. As such, what may have been considered an “improvement” at the time of the filing is not necessarily the actual distinction that was deemed patentable.
The originally-filed claim was:
1. A method for providing captioning in an AV signal, the method comprising:
converting an audio signal in the AV signal to caption data automatically, and
associating the caption data with the AV signal at a time substantially corresponding to a video signal associated with the converted audio signal in the AV signal.
THAT is a claim to automating the speech-to-text function. However, the claim that was allowed is considerably different than the originally-filed claim.
It’s not protectable to claim the CONCEPT of making the computer convert a person’s speech into words, which is all the claim is doing.
I hope that you are not a patent professional, as your ability to identify what a claim covers needs some serious, serious help.
As near as I can tell from the decision, no improvement whatsoever was argued beyond the fact that having a computer convert the audio signal to text would be faster and require less human resources than if a human did it.
Seriously, you said you read the specification, but did you really? This is from the specification:
More specifically, speech-to-text processing system 20 converts the spoken words in the audio signal of an AV signal into text data utilizing a conventional speech-to-text software application such as such as Dragon Dictate.
Here, the specification admits that speech-to-text software was CONVENTIONAL. OF COURSE IT WASN’T THE IMPROVEMENT. You actually believe that the Federal Circuit would accurately characterize the improvement? LOL, are you that naïve? The Federal Circuit is going to characterize the improvement in a manner that is going to allow them to ignore it. The Federal Circuit wrote “[t]hat process, consisting of converting audio to text and associating the text with corresponding video, is not itself asserted to be an advance over the prior art.” This in no way, shape, or form accurately characterizes the substance of the invention.
I think you’re confusing the fact that the claim as a whole was found non-obvious with some sort of notion that there must have been novel functionality. But it could have been found non-obvious because of a lack of express motivation to combine (this is a pre-KSR patent) despite all of the parts being known.
Shoot from the hip much? It is quite clear that you didn’t bother to look at the prosecution history, which would have given you a better idea as to patent office allowed this application. The invention was hardly directly to anything nearly as generic as both you and the Federal Circuit described.
That the great thing about the current case law (if you are an infringer). The Federal Circuit has no shame in mischaracterizing inventions – they are all about the ends justifying the means. The law is meaningless at this point and they know it. There is little attempt these days to reconcile all of the disparate opinions by the Federal Circuit. Only the naïve attempt to make sense out of the nonsensical.
Which is exactly why the 2019 PEG was written as it was, aiming to give the examiners off-ramps AND taking away from the examiners the notion that they can “play lawyer” (or as Random asserted, be just a country lawyeh).
Being a lawyer is MORE than merely being able to follow the latest judicial cases and being able to understand legislative text.
Being a lawyer (a good one at least), means being able to critically evaluate judicial cases, to recognize when the score board is broken (and why), and to be rightfully upset when the score board is broken.
Attorneys take an oath of office — and that solemnity DOES mean something.
Leastwise, to good attorneys.
You know … the agency charged with executing the law should have expertise in the law.
That’s fine, so long as you’ll pay for lawyer-engineers to examine your applications and not just engineers. I don’t see applicants clamoring for examination to be staffed with lawyers, so I don’t understand what you’re going for here: You want non-lawyers to give lawyer-level of analysis for non-lawyer level cost? That’s unreasonable on at least two levels. Asking the court to give deference to the legal decisions of non-lawyers is a third level of unreasonableness. Asking for the court to give deference when the allowance of this particular case predated the Alice test by almost a decade adds a fourth. But I at least appreciate the sentiment that there should be confidence in the offices allowance here.
It is a meaningless test untethered from the language of the statute and the case law.
There’s caselaw going back hundreds of years that talk about how functional language is problematic. Halliburton Oil, Morse, Perkin’s Glue, Wabash…Other than giving the two-step procedure, the court didn’t raise any concerns in Alice that weren’t in Morse, which was over a hundred years old.
You, like the Federal Circuit, fail to appreciate that the specification is written before all of the possible prior art is known. As such, what may have been considered an “improvement” at the time of the filing is not necessarily the actual distinction that was deemed patentable.
And you fail to appreciate that the patentee had not one but two chances before courts to identify a valid improvement, and you have a chance right now, and none of you are articulating anything. What is more likely – the examiner followed pre-KSR law and allowed an obvious combination because of no express motivation to combine, or that the attorneys were so incompetent before two separate tribunals that they could not articulate (and neither can you) what the patentable improvement is?
However, the claim that was allowed is considerably different than the originally-filed claim.
That’s not an analysis. How are the differences RELEVANT? The additional features (besides the training, as I already noted) deal with encoding and displaying the captions, i.e. actions that would have to occur in any captioning system, so explain why those differences matter. Why does telling the art what it already knew make a claim any more patentable or eligible? Especially when this exact situation is what step two analysis is for, and no argument was made as to unconventionality of these features.
Here, the specification admits that speech-to-text software was CONVENTIONAL. OF COURSE IT WASN’T THE IMPROVEMENT.
The specification admits that text-to-speech software was out there. The specification does not admit (1) training preexisted, and (2) it was employed in the context of a closed captioning system. This is just the kind of truly obvious invention that was getting allowed in a pre-KSR world.
You actually believe that the Federal Circuit would accurately characterize the improvement? LOL, are you that naïve?
I expect the attorneys for the patentee to put forth a competing characterization. And when you attack the federal circuit, it’s your burden to explain what they did wrong. So what is the improvement? Still waiting for you to explain it.
It is quite clear that you didn’t bother to look at the prosecution history, which would have given you a better idea as to patent office allowed this application.
It’s pretty clear why this claim was issued. Examiner cited four pieces of prior art which he asserted taught all of the previously amended claim. He amended in an encoder setting feature. Here is the entire sum of that disclosure in the specification:
Additionally, since various types of encoders 40 may be utilized with speech-to-text processing system 20, a user may cause the AV captioning system 10 to perform a handshake process to recognize a newly added encoder 40 by clicking on initializing encoder button 88, which causes the speech-to-text processing system 20 to recognize which type of encoder 40 is being used and to retrieve its settings so the speech-to-text processing system 20 can communicate with the device.
Recognizing and retrieving settings of an encoder was known, which is why there’s no enablement problem here. If this was a novelty the claim wouldn’t issue because there’d be a 112a violation. Moreover, if the encoder was plausibly inventive, the attorney could have easily pointed to that.
This claim issued because the examiner didn’t want to go to appeals citing four references in the pre-amended version of this claim, so he amended in anything that would have required a fifth reference and let it go, i.e. it issued because it was non-obvious as a whole because of a motivation to combine issue.
Before the examiner amendment, the arguments put forth by the applicant were (1) display of a variable number of text lines (lol), (2) training of the text-to-speech (i.e. what the court identified as the ineligible subject matter) and (3) time syncing of the caption data with the audio (which the prior art showed).
So in short what the CAFC did aligned with the final argument made by applicant during prosecution, the attorneys had a clear path to assert otherwise and apparently did not, and its pretty clear that even if the encoder feature was pointed to that there’s both a lack of enabling disclosure and statements in the specification that it was conventional – both of which suggest it is not novel.
The Federal Circuit has no shame in mischaracterizing inventions – they are all about the ends justifying the means.
The other way to look at it is that when your original thesis for patentability falls down and you’ve written a specification that admits some features are conventional and truncates the description of other features (suggesting the art already knew how to make and use them by relying upon the supplementary rule) you don’t plow ahead blindly merely because the burden is on the office to show unpatentability. That may result in an allowance but it doesn’t necessarily result in claims that stand up to scrutiny. A presumption of validity is just that, a presumption. At some point someone is going to ask you to articulate why the claim has value and what you’ve taught the art and they’re going to weigh that response against what the defendant is saying. You put a lot of fault on CAFC’s analysis here, but you haven’t put forward an alternative analysis. The CAFC didn’t write this spec, nor agree to the examiner’s amendment, nor run the argument on appeal, so if these claims are so eminently worthy of protection where were patentee’s attorneys? It seems like the sum of the appellate work was likening the claims to a PTO decision and pointing out that the PTO refused to grant IPR.
I’m not a huge fan of the CAFC, but if you don’t give an appellate court an analysis that results in you winning for them to adopt, you really shouldn’t be surprised when you don’t win. It was not new law to say that decisions of the PTO don’t bind the CAFC and that obviousness has nothing to do with the eligibility analysis.
That’s fine, so long as you’ll pay for lawyer-engineers to examine your applications and not just engineers.
Maybe you want to refer back to the original comment (i.e., “On appeal the court gave no deference to the PTAB’s interpretation of eligibility doctrine”) that I spoke to. It was the engineer-lawyers (and their application of the 2019 Patent Eligibility Guidelines, also written by lawyers) that were ignored. This comment referred to Ex parte Hunnan, which was an “informative” PTAB decision that the Federal Circuit swatted away.
Other than giving the two-step procedure, the court didn’t raise any concerns in Alice that weren’t in Morse, which was over a hundred years old
LOL. Morse was about lack of enablement that was recast as an eligibility decision by a Justice (in Benson) who thought trees should have standing.
Why does telling the art what it already knew make a claim any more patentable or eligible?
It looks like you are unfamiliar with Bascom (“[t]he inventive concept inquiry requires more than recognizing that each claim element, by itself, was known in the art”) and FN5 of DDR Holdings (“Modern inventions very often consist merely of a new combination of old elements or devices, where nothing is or can be claimed except the new combination”).
And you fail to appreciate that the patentee had not one but two chances before courts to identify a valid improvement, and you have a chance right now, and none of you are articulating anything.
Again, relying upon the Federal Circuit to say what the Patentee did or didn’t argue regarding any improvement. Naivety stacked on top of naivety.
The specification does not admit (1) training preexisted, and (2) it was employed in the context of a closed captioning system. This is just the kind of truly obvious invention that was getting allowed in a pre-KSR world.
You really think that would have been allowed in the pre-KSR world? You really are grasping at straws.
I expect the attorneys for the patentee to put forth a competing characterization. And when you attack the federal circuit, it’s your burden to explain what they did wrong.
I did. They mischaracterized the invention as “automating the AV-captioning process,” which they also mischaracterized as being an abstract idea.
So what is the improvement? Still waiting for you to explain it.
Read pages 7-10 of the amendment filed on September 23, 2005.
He amended in an encoder setting feature
Oh, so it is something different than how you and the Federal Circuit described the invention.
This claim issued because the examiner didn’t want to go to appeals citing four references in the pre-amended version of this claim
Maybe because the invention wasn’t obvious? Ever thought of that? Perhaps the Examiner didn’t think his combination would hold up so he offered up an amendment in the hopes that applicant would take it. This is Standard Operating Procedure (SOP) at the USPTO for decades now.
So in short what the CAFC did aligned with the final argument made by applicant during prosecution, the attorneys had a clear path to assert otherwise and apparently did not, and its pretty clear that even if the encoder feature was pointed to that there’s both a lack of enabling disclosure and statements in the specification that it was conventional – both of which suggest it is not novel.
I suspect my legal writing professor would look at this and write: “nice try, throw it out and write something that is understandable next time.”
The other way to look at it is … on the office to show unpatentability.
My legal writing professor would probably draw a line from the last comment to this comment.
At some point someone is going to ask you to articulate why the claim has value and what you’ve taught the art and they’re going to weigh that response against what the defendant is saying
“value” is not part of any of section 101, 102, 103, or 112 of 35 U.S.C. – however, keep making stuff up, this is quite amusing.
You put a lot of fault on CAFC’s analysis here, but you haven’t put forward an alternative analysis
Not directed to an abstract idea. Happy now?
I’m not a huge fan of the CAFC
They don’t invalidate enough patents for your liking? You have pretty high standards because they invalidate a large percentage of the patents that hit their docket.
It was not new law to say that decisions of the PTO don’t bind the CAFC and that obviousness has nothing to do with the eligibility analysis.
I never said that the PTO’s guidance binds the CAFC and nothing I wrote involved how obviousness impacts the eligibility analysis. However, what I did write is that the USPTO is the agency that is tasked with applying the law yet the USPTO (in all of its institutional expertise) cannot figure out how the courts are applying 101 – I’m not just talking about the examiners either. I’m talking about the USPTO’s lawyers (who write the guidelines) and the APJs who issue the informative opinions. Moreover, as I have noted many times, I take issue with the Federal Circuit time and time again misrepresenting what is really inventive about a particular invention
It was the engineer-lawyers (and their application of the 2019 Patent Eligibility Guidelines, also written by lawyers) that were ignored.
The office did not analyze this case under the judicial exceptions EVER, not during original examination and not since. Hunnan was not ignored, it was distinguished. The 2019 PEG is not involved in this case at all, but even if it were the goals of that PEG are not to make all of the possible 101 invalidations that could be made, so why would that be evidence that this claim is not ineligible?
Morse was about lack of enablement that was recast as an eligibility decision by a Justice (in Benson) who thought trees should have standing.
There was no evidence that the scope was not enabled. The only thing the Morse court uses is the threat that there could be future inventions that fall within the scope that are currently unenabled. But speculation has never been sufficient to overcome the presumption of validity. If you’re asserting that the failure to foreclose possible future inventions at the time of filing constituted an enablement violation, that simply is not the enablement test. But I agree that if you conceive of the test in that manner it’s not far off from the judicial exceptions.
It looks like you are unfamiliar with Bascom (“[t]he inventive concept inquiry requires more than recognizing that each claim element, by itself, was known in the art”) and FN5 of DDR Holdings (“Modern inventions very often consist merely of a new combination of old elements or devices, where nothing is or can be claimed except the new combination”).
It seems like you’re conflating the fact that the second step of the alice test must be performed with the notion that conventional subject matter is sufficient to render a claim eligible. I don’t need to cite the CAFC – the Supreme Court already told me that merely conventional subject matter isn’t sufficient under Alice Step 2. The whole point of Alice Step 2 is that telling the art what it already knows doesn’t make an ineligible claim eligible. Eligibility is “not like a nose of wax” and “does not depend simply on the draftsman’s art” as the Supremes say.
Again, relying upon the Federal Circuit to say what the Patentee did or didn’t argue regarding any improvement. Naivety stacked on top of naivety.
Surely the en banc or the Supremes will correct such a blatant failure of basic appellate function when it’s appealed to them, right? I mean according to you these guys got no appelate argument at all, the panel just ignored their arguments altogether. I await the en banc reversal.
Read pages 7-10 of the amendment filed on September 23, 2005.
Those claims weren’t allowed. If the office was persuaded by that argument those claims would have been allowed. You have an assertion by the applicant that the office disagreed with, which you can see because the examiner required additional language and then cited art in the notice of allowance that taught the limitations the remarks argued about.
Oh, so it is something different than how you and the Federal Circuit described the invention.
It is, too bad the specification admits encoders were known and didn’t teach anything about the encoder determination feature. This is at least something that could have been argued and the correct invalidation would at least have been enablement rather than 101.
My legal writing professor would probably draw a line from the last comment to this comment.
Is this the person who taught you that complaining about the CAFC doing what the supreme court told them to do was a viable tactic for prevailing in court, or did you come up with that on your own?
Not directed to an abstract idea. Happy now?
*my* legal writing teacher would call this conclusory.
However, what I did write is that the USPTO is the agency that is tasked with applying the law yet the USPTO (in all of its institutional expertise) cannot figure out how the courts are applying 101 – I’m not just talking about the examiners either. I’m talking about the USPTO’s lawyers (who write the guidelines) and the APJs who issue the informative opinions.
This statement assumes the only goal of the office is to figure out how the courts are applying 101 and that it can’t do it. You, like anon, emphasize that one goal of the office is to accurately apply the law while ignoring that the office has other goals such as timely examination and sufficiently low cost to allow individuals to contribute.
The office knows it doesn’t employ lawyers. The office knows how expensive it is to its mission to require examiners to do an analysis, and how expensive it is to update examiners every time new cases come out, and the office knows the federal circuit is not attempting to make an overarching test. The office has never given examiners any additional time to perform the Alice analysis. (Relatedly, the office likes to crow about how average time to first action and total pending applications are consistently falling year over year.) Ask yourself how it is that the Alice analysis could possibly be done properly (even assuming one was perfectly trained) when no additional time is given to even write it up let alone consider it?
The office used to have rules that required examiners to liken their cases to other cases (i.e. exactly what the patentee tried to do with Hunnan here) and it was too expensive to do, because no applicant wanted to accept an examiner’s non-legal opinion that was entitled to no deference from the PTAB.
So faced with the prospect of having to give more time to force ill-suited examiners to write rejections which would result in more appeals to PTAB judges which may or may not uphold them based upon guidance that changes monthly, the office solved the issue by simply promulgating rules that relieved examiners from making the rejections in the vast majority of cases. The office largely relies upon 103 to control claim scope, largely ignoring 112 and 101. Which frankly is fair because its only using people whose expertise is in doing a 103-style analysis and only very basic 101/112 inquiries.
It’s specifically not the intent of the rules to simulate the court’s analysis, which is why it is unsurprising that once claims ARE finally subjected to the analysis, they often fail. It’s not that the PTO’s lawyers are incapable of distilling judicial opinions (the office mind as well hire Crouch for that), it’s that the guidance does not represent the goal of accurate examining – the guidance represents the goals of timely and cheap examining, which was probably the right choice given that the office doesn’t have the workforce to substantially achieve accurate examining in this area anyway. But simply because the office doesn’t comport with the CAFC doesn’t mean that comporting with the CAFC is unknowable or unachievable, especially when one has the tool of dependent claims that allow for multiple levels of abstraction and therefore multiple bites at the abstraction apple.
I don’t blame this patent’s prosecutor because these claims issued in a pre-Alice world, but for an entire claimset to fail in a post-Alice world is really an indictment of the person writing the spec or the person (and sometimes this includes the examiner) generating the claims. It is neither difficult to create nor unforseeable to require multiple layers of abstraction. For entire claimsets to be tossed out years and years after Alice came down is clearly due more to the stubbornness and ego of attorneys than the lack of guidance and clarity of the federal circuit at this point. The majority of specifications relate at least one eligible scope, the first 20 claims are flat rate, one should not lose an entire claimset at trial.
“You, like anon, emphasize that one goal of the office is to accurately apply the law while ignoring that the office has other goals such as timely examination and sufficiently low cost to allow individuals to contribute.”
I ign0re no such other goals.
I — however — do know that those other goals cannot take the place of the most important goal.
You seem to have missed that lesson.
Random,
Yet again, you simply do not get the Ladders of Abstraction, and to apply what you state would result in the only valid (or even eligible) claims to be exacting picture claims.
Throw in your odd sense of scope (and the fact that an improvement patent would necessarily by your logic vitiate any proceeding claim that was NOT “picture claim enough”), and you end up with the collision that your logic necessitates: only excessive picture claims that cannot possibly be the subject of improvement patents are “real” inventions.
Yet again, you simply do not get the Ladders of Abstraction, and to apply what you state would result in the only valid (or even eligible) claims to be exacting picture claims.
Well the CAFC invalidated the claims, so I guess they don’t “get” ladders of abstraction as much as I do either. A lot of the final words on a lot of cases seem to agree with my ungotten view of the ladders. Maybe the Supremes will take this and reverse to show me how wrong I am. But I think not.
only excessive picture claims that cannot possibly be the subject of improvement patents are “real” inventions
So what is the appropriate “rung” of the ladder here anon? Should Dragon Speech-to-Text get to keep all competition out of the speech-to-text field, regardless of how bad it’s algorithm is? Do consumers get only one option for speech-to-text because the most abstract, functional rung is appropriate?
I think, and maybe I’m just spitballing here, that maybe dragon spech-to-text gets a patent on the speech-to-text algorithm it uses, and someone else who has a potentially better way of converting audio signals into text is free to get protect and market their algorithm to compete with Dragon speech-to-text. Maybe that system encourages more and most importantly better algorithms be disclosed to the public to, you know, “advance the arts” which may be a tangential goal of the patent system or something.
You are deeper into the weeds than you know.
Just spitballing indeed (your attempt to be flip actually reflects a far worse state of your “understanding”).
Should Dragon Speech-to-Text get to keep all competition out of the speech-to-text field, regardless of how bad it’s algorithm is
Speech recognition supposed first existed in the 50s in Bell Laboratories. It was until the early 80s that Dragon Systems launched.
It is difficult to make good arguments about what the law is and/or should be when you have a poor understanding of the real facts — as opposed to the shaded “facts” presented by others. The problem with your hypothetical is that it doesn’t reflect reality. These broad ‘abstract ideas’ supposedly being claimed in patents don’t exist in the real world. However, their supposed identification (by unsympathetic judges) makes for good ex post facto rationalizations as to why certain claims are deemed patent ineligible.
These broad ‘abstract ideas’ supposedly being claimed in patents don’t exist in the real world.
I love that you keep telling me what the facts are as if I’m not an expert in this area and see an order of magnitude more claims than you. This happens routinely in software claims.
Let’s assume you’re the first person to come up with the idea (and it is an abstract idea) of using a computer to evaluate different possible driving paths from a given point A to point B, to tell you which is the best path, and to estimate the amount of time it would take to get there. An algorithm that considers 10,000 randomly selected options will get you a response. But it won’t be consistent, there’s no guarantee it will return the top path, and it will take a noticeable amount of time to work. Similarly, when estimating the amount of travel time, I suppose one could assume that you average 25mph when travelling by car and simply divide the distance by that speed. That algorithm will also not be very accurate. Google has a far superior algorithm, both in route choice and trip time estimation.
And yet Google was not the first to the office. But the first person to the office will routinely generate a claim that uses broad functional language to overclaim what is invented. The terrible algorithm described above will become “an embodiment” of the functionality, the specification will have boilerplate language that other algorithms could be used, and the claim will be a handful of contextual limitations surrounded by novel functional language:
Claim 1: A method comprising: receiving input comprising a desired destination; determining the current position of a device; calculating, based upon the current position and the desired destination, a substantially optimal driving route between the current position and the desired destination; estimating a travel time for the substantially optimal driving route; and displaying the substantially optimal driving route and the travel time on the device.
Notice how all of the terrible fails of my actually-posited algorithm are swept away in functional language and my scope forecloses the unposited and vastly superior Google algorithm. Because all analysis requires some statement of where one is and where one is going, this abstract claim sweeps all possible competitive algorithms into the monopoly. This claim is ineligible as soon as you see the words “calculating” and “estimating” without limitations that explicitly define a particular calculation.
This happens daily. It famously happened in what was arguably the first judicial exception case – Morse v. O’Reilly, where Morse set forth a claim scope that would read on the modern fax machine, something he obviously didn’t invent and which shares only a tangential abstract functional-similarity relationship with the telegraph that Morse did invent.
Speech recognition supposed first existed in the 50s in Bell Laboratories. It was until the early 80s that Dragon Systems launched. It is difficult to make good arguments about what the law is and/or should be when you have a poor understanding of the real facts — as opposed to the shaded “facts” presented by others.
My suggestion was not that Dragon invented speech to text. My suggestion was that Dragon, like everyone else, cannot describe the means that they posit using such vague and abstract terms that it reads upon everything that will be invented in the field that isn’t already foreclosed by prior art. anon views “ladders of abstraction” as a right of the patent drafter to claim however they see fit. Obviously a drafter may claim however they wish subject to extremely limited rules such as a dependent claim must depend from a previous claim. The question is whether such a category of claiming is valid.
However, their supposed identification (by unsympathetic judges) makes for good ex post facto rationalizations as to why certain claims are deemed patent ineligible.
One may discover string tension and invent the bow. One should get a valid claim that covers the bow. If the claim also covers the catapault, the claim is likely overbroad. If the claim also covers the gun, the claim is certainly overbroad. The only way that language that covers the bow could also reach to the gun would be to use such vague and abstract terms so untethered from the actual structure of the bow that they do not accurately describe the contribution to the art. The bow may perform the function of firing a projectile, or it may have the utility of striking an opponent from a distance, but when those become claim terms that operate to bring the gun within the scope of the claim, the claim is invalid. Some people would view the claim to the function as a failure to “point out and particularly claim” or to make the scope “reasonably certain” under 112b. That ignores that the intent of the applicant is to actually claim the functionality itself, and views the bow as only one embodiment. Some people would argue that enablement or written description is violated. That ignores that they are analyzed at the time of filing, and at that time string tension is the only known mechanism for achieving the result, which suggests the full scope is enabled or described. The only mechanism to prevent the unjust enrichment and the failure to promote the arts that comes with the overbroad claim would either be to reconfigure the 112a or b tests, or to generate a test that essentially makes the act of claiming “the function not limited by particular means” ineligible. That’s the judicial exception test of Alice. If you like, we can just fudge the 112b or WD tests and solve the problem that way. But either way, the system cannot tolerate the person who invents the bow being given exclusory rights over the gun. The gun will never be researched or disclosed if they can’t monetize their invention because of the overbroad bow claim.
The same is just as true for functionally claimed software. If this applicant had an improvement to the closed captioning process, he could have disclosed the algorithm and claimed the algorithm. But the claim is nothing more than basic contextual limitations that operate in the sphere of closed captioning together with functional language. It frankly does not matter what was invented (and the reason you have such difficulty pointing to an improvement is that in reality nothing was invented here) – what was claimed was obviously overbroad of that. There is nothing ex post facto about it – any improving algorithm was clearly amenable to description (after all, the computer understood it) but the applicant didn’t describe the invented algorithm, they described the result that occurs when the invented algorithm is used. That is not an invention. One doesn’t invent functionality or utility, one only invents the means which brings the utility about.
“I love that you keep telling me what the facts are as if I’m not an expert in this area and see an order of magnitude more claims than you. This happens routinely in software claims.”
Random, your volume (much like the diarrhea of words in your posts) do NOT make you into ANY type of expert.
The problem with using eligibility to police claiming functions not limited by particular means is that those concepts are not related to any particular subject matter, and the point of section 101 is to demarcate subject matters which may be patented; doing to by naming four specific categories of subject matter. 112 (b) is explicitly about claims, not subject matter, and should most certainly be the aegis of the Alice inquiry.
Three of those categories almost never present a problem. The Bilski court failed to resolve the question of what may or may not be an eligible process. The problem is information. Your mapping example, in my scheme?
Is it a method? Yes. Is the result of the method only an item of information? Yes. Is the utility of the method achieved by human use of the information? Yes. Ineligible.
If the invention were part of an autonomous vehicle claim, where the information was consumed by the vehicle for the purpose of navigation, it would then be eligible. Once eligible, Alice should then apply in exactly the same way, but in a 112(b) or 103 context, where it belongs.
If your weak algorithm was described in the claims such that PHOSITA could make it, and it passed Alice , it could issue, and would not preempt Google’s later and better algo- although the question of the doctrine of equivalents would eventually lead to some kind of subjective decision anyway, because eventually, nearly all
seriously litigated patents come down to one form or another of a subjective decision.
That’s the nature of the business. Eligibility as currently practices conflates 101 and 112(b) / 103 because the former is simple and a threshold question, saving untold millions of dollars. The solution is a procedure to construe inventions early, as claims are, to knock the easy ones out cheaply. Just move Alice where she belongs, next to Markman.
You edge closer Marty, but still come up far short (enamored as you are with your pet theory and substantial miss on the legal term of art of utility).
This would be eligible under my scheme. The method’s result is an item of information. The information is consumed by the device’s rendering system, and the utility of the consumed information is independent of any specific human meaning or use.
Is it obvious or could PHOSITA practice the invention as described?
The Alice test may or may not help answer that. But this is another clear example of a case where the invention required judicial construction. There should be a standard, adversarial process for doing that as a matter of law, just like Markman does for claims.
? There already is now [by default?] such a standard judicial claim construction for 101 analysis – a FRCP 12(b)(6) motion hearing, as here.